Re: Using a variable for an array name

[Sannyasin Brahmanathaswami via use-livecode]

Peter: FYI multi-dimensional arrays are awesome. I use to be "scared" of them, 
but in RevIgniter we have all kinds of things like

galleriaShow["options"]["width"]["805]

and even more… so I had go understand it.

And, on top of this you  can have an array "inside" a key/value (where the 
value is an array)

Once you wrap your head around this you will never look back!

BR




___
use-livecode mailing list
use-livecode@lists.runrev.com
Please visit this url to subscribe, unsubscribe and manage your subscription 
preferences:
http://lists.runrev.com/mailman/listinfo/use-livecode

Drag List Items Up and Down

[Sannyasin Brahmanathaswami via use-livecode]

There is a near universal behavior on mobile, varying flavors of color and 
style but the UX is almost the same everywhere whether it is your contacts, 
book marks in Safari, or songs on play lists in Spotify.

you have a list like this, after each item the common icon is often 2 
horizontal bars,

apples   ==
oranges==
grapes   ==
kiwi   ==

the user can long press, hold down and drag up and down on the little bars on 
the right and move the list item up and down to re-order the list.

Before I go nuts in a deep dive building this from ground zero, does anyone 
have a library that does this already they can share?

Needs to work on mobile.

Mahalo!

BR
___
use-livecode mailing list
use-livecode@lists.runrev.com
Please visit this url to subscribe, unsubscribe and manage your subscription 
preferences:
http://lists.runrev.com/mailman/listinfo/use-livecode

Coding Responsive UI in Livecode

[Sannyasin Brahmanathaswami via use-livecode]

[moved to a new thread]

Sample stacks are probably the way to go… with some lightweight "guide on how 
to hink about it. Like Trevor wrote "slice up your UI into rectangle and 
groups… go from there."


Richard wrote:

> A lot of the "flak" I hear about using LC is "but it's not
> responsive, at that so easy with HTML5"   They gloss over the
> depth of javascript skills required to actually program the UI
> to do anything other than look pretty. However easy it may be
> to make a  and apply a class. The "ease"
> ends there.

Well said.  "Responsive" isn't a technology, but a design decision
(these days, arguably a requirement).  Regardless of the language used,
decisions need to be made, and the expression of those decisions in code
needs to be tested.

Some things are easier in one language than another, and other things
easier in other languages.

But all of them require commitment to providing a good UX, clear
thinking about the layout goals, and good familiarity with whatever
language will be used to express those goals.


> Point: if we had a more robust set of examples, lessons, youTube
> Tutorials on "building responsive UI for LiveCode" it would go
> a long way in helping the product feel that this issue, is a
> non-issue. Right now it is.

Personally, when it comes to programming I prefer textual tutorials, so
I can copy-and-paste code.  But any tutorials folks in our community
want to deliver would be welcome.


Svasti Astu, Be Well
Brahmanathaswami
www.himalayanacademy.com

___
use-livecode mailing list
use-livecode@lists.runrev.com
Please visit this url to subscribe, unsubscribe and manage your subscription 
preferences:
http://lists.runrev.com/mailman/listinfo/use-livecode

Re: Drag List Items Up and Down

[Sannyasin Brahmanathaswami via use-livecode]

Thank you all!  

FYI tmpanel.rev crashes 8.1.3 (at least here on Sierra)

 

On 2/4/17, 11:23 AM, "use-livecode on behalf of AndyP via use-livecode" 
 wrote:

Nice addition .. thanks.



-
Andy Piddock

___
use-livecode mailing list
use-livecode@lists.runrev.com
Please visit this url to subscribe, unsubscribe and manage your subscription 
preferences:
http://lists.runrev.com/mailman/listinfo/use-livecode

Re: JavaScript, Browser Object, VR with Hotspots?

[Sannyasin Brahmanathaswami via use-livecode]

FYI Mike,

Getting those params over to LC is doable… Jonathan  wrote me off line.. I will 
email you a small stack he created that works and perhaps will suffice for your 
needs.

But, beware, you may head into this thinking "Wow, this is so cool. We could do 
a lot with this."  Hold your horses, you will banging against walls in no time. 
The browser widget is *not* in the message path 

We had some long discussions on the integration of the Browser Widget/HTML5 
apps  already

see:  thread: Browser Widget/HTML5/LC Integration

We also have an enhancement request which was 'hibernated' that may interest 
you.

http://quality.livecode.com/show_bug.cgi?id=19184

related:

http://quality.livecode.com/show_bug.cgi?id=19021#c15

I even got a quote from HQ team on the price for implementation, I can pitch in 
$1,000.00 and if anyone else is interested in tighter integration between the 
browser widget and LIvecode message path, contact me off list at 
bra...@hindu.org. I was hoping for a group effort to pay for the development, 
HQ's estimate $  is actually more than reasonable given the scope of the 
requirements…  but not getting any takers so far.

The discussion continues here if you want to add more.

http://forums.livecode.com/viewtopic.php?f=11=28737

BR





 

On 2/5/17, 7:08 AM, "use-livecode on behalf of Mike Bonner via use-livecode" 
 wrote:

Strange.  I've been messing with your project and have some ideas, but I've
run into a blocker.   I can't seem to get javascript params to pass to the
lc handler.  The handler fires as expected, but params are empty. *sigh*

___
use-livecode mailing list
use-livecode@lists.runrev.com
Please visit this url to subscribe, unsubscribe and manage your subscription 
preferences:
http://lists.runrev.com/mailman/listinfo/use-livecode

Re: Converting scripts in stacks to script only stack behaviors

[Sannyasin Brahmanathaswami via use-livecode]

Ali, thanks for chiming in… in the current app we stayed away from any 
resizing, the Design passed down from our team which has years of experience in 
print graphics (but zero in digital UX/UI) would have been a challenge to get 
responsive. It could be done of course, anything can be done in Livecode!  But, 
compared to using CSS for web, writing all the code for making the design I got 
handed, responsive, would have taken more time than we wanted to put in at that 
stage of the project (with "tiny" budget!)… possibly in the future.  

RE: "a good resizeStack Handler" … it would be great is those who have mastered 
this skill--have libraries, examples… were to place them somewhere for us to 
study.

A lot of the "flak" I hear about using LC is "but it's not responsive, at that 
so easy with HTML5"   They gloss over the depth of javascript skills required 
to actually program the UI to do anything other than look pretty. However easy 
it may be to make a  and apply a class. The "ease" ends 
there.  If I have a UI ready to go and ask a JS jockey to program the whole UX, 
click a stop watch and have Jacqueline do it in LiveCode. Jacque will be done 
in half the time. 

Point: if we had a more robust set of examples, lessons, youTube Tutorials on 
"building responsive UI for LiveCode" it would go a long way in helping the 
product feel that this issue, is a non-issue. Right now it is.

BR
 

On 2/5/17, 8:21 AM, "use-livecode on behalf of Ali Lloyd via use-livecode" 
 wrote:

On the subject of building UI from script, I generally think it boils down
to whether you already have a good resizeStack handler implemented.  If you
do, you've done all the hard work already - you can just create all the
objects with their required properties before the card opens and delete
them when it closes, and your resizeStack does the rest of the placement
work.

Obviously this is a bit trickier if you have a lot of fields prepopulated
with text, or groups with more complicated structure where it's not so easy
to tease out the scripts into a behavior.

___
use-livecode mailing list
use-livecode@lists.runrev.com
Please visit this url to subscribe, unsubscribe and manage your subscription 
preferences:
http://lists.runrev.com/mailman/listinfo/use-livecode

Re: JavaScript, Browser Object, VR with Hotspots?

[Sannyasin Brahmanathaswami via use-livecode]

Actually you can get it here:

http://livecodeshare.runrev.com/stack/817/LC-to-JS-Data-Transfer



On 2/5/17, 10:28 AM, "use-livecode on behalf of Sannyasin Brahmanathaswami via 
use-livecode" <use-livecode-boun...@lists.runrev.com on behalf of 
use-livecode@lists.runrev.com> wrote:

Getting those params over to LC is doable… Jonathan  wrote me off line.. I 
will email you a small stack he created that works and perhaps will suffice for 
your needs.



___
use-livecode mailing list
use-livecode@lists.runrev.com
Please visit this url to subscribe, unsubscribe and manage your subscription 
preferences:
http://lists.runrev.com/mailman/listinfo/use-livecode

Re: Browser Widget Appears to be caching data (javascript)?

[Sannyasin Brahmanathaswami via use-livecode]

John: Ditto what Bill said…that, keep us all posted! Possibly get a thread 
going on the multi-media are of the forums?

 William Prothero:

John, this is a very cool project. I hope you get it working. I’d love to 
be able to use your VR framework. One of my colleagues created a VR exploration 
of the East Pacific Rise, with lots of videos, etc. It stopped working when 
Quicktime VR went away. It was written in Director, ages ago. But, it was a 
very cool app and would be really wonderful to be able to resurrect it in 
livecode.



___
use-livecode mailing list
use-livecode@lists.runrev.com
Please visit this url to subscribe, unsubscribe and manage your subscription 
preferences:
http://lists.runrev.com/mailman/listinfo/use-livecode

Re: Converting scripts in stacks to script only stack behaviors

[Sannyasin Brahmanathaswami via use-livecode]

Trevor, thanks for that intro to the new tools in LC9


I'm already using libraries and behaviors extensively, but still find my self 
build LC stack binaries with a full stack script or, quite often,  95% of the 
code in the card script (if the UI can all be done on a single card) 

The other day I was thinking 'gee, why not just move this card script out to a 
behavior and assign it to the card."  been mulling this over for a few days, 
the the work flow for making this transition seemed onerous, but LC9 will make 
it s easy!

Our new app, SivaSiva, is being done with GIT and this is the first time 
working with GIT for something like this… for web and RevIgniter, GIT make 
obvious sense, since the whole environment is text files from the ground up.

But now, with the new app, we have this mix of text only scripts and binary 
*.livecode stacks.

So this leads to two more interesting avenue of adventure/explorations

1) we are already hitting GIT conflicts with the main stack of the project 
because if one developer needs to add some stack files, or wants to change a 
standalone setting for testing a build on a device, the IDE of course requires 
a Save… so now this stack must be either stashed (assuming you don’t need the 
changes) or committed.. and we hit a conflict later that is not so easily 
resolved.

2) building UI from script: every now and the notion passes by that perhaps 
building UI from script has advantages of  the LC WSIWIG model. I was  
intrigued in your to see in the background in sublime text, you had a handler 
that built some of the UI.  perhaps all of it for that dialog box?  
So the question then becomes when, where and why do we make a decision to go 
the route to build UI by script? I guess the one obvious answer is that you 
want to put the UI under git control so that we have, as you put it 
"fine-grained control over changes, edits, history" etc.

I wonder if there are other criteria besides that. Possibly building UI be 
script may be better suited to responsive design, since, in the end, if you 
want responsive geometry, you will end up writing those scripts anyway.  The 
only thing missing being the default init props for the control (which we 
normally would create in the IDE) so one must as well "create button" and set 
the default init buttons in the script. Actually it would be pretty easy to 
create the button in WSIWIG and have a small tool to write out the initial 
props that you could then turn into a script.

But this should probably be different thread.  I'm sure HQ has solved moved of 
these issues, but we may need to wait until LC9 is unveiled before we get the 
full picture from the mother ship, until then… I "muddle along"

Brahmanathaswami



___
use-livecode mailing list
use-livecode@lists.runrev.com
Please visit this url to subscribe, unsubscribe and manage your subscription 
preferences:
http://lists.runrev.com/mailman/listinfo/use-livecode

Re: Browser Widget Appears to be caching data (JavaScript)?

[Sannyasin Brahmanathaswami via use-livecode]

Are you serving this content yourself from your web server? If so there are 
"cb" (cache busting) methods already well worked out.

e.g. 

logo.jpg  # on disk
logo-cb123455678.jpg  # in the html code

and mod-rewrite handles the translation back to the original image. So you can 
update "logo.jpg" without changing it's name, but your content assembly system 
appends the cache busting string to the outgoing html… (Ralf does this in 
RevIgniter)

BR
 

On 2/6/17, 7:07 AM, "use-livecode on behalf of Ralph DiMola via use-livecode" 
 wrote:

To refresh a browser cache I put an argument on the URL that is different
from the last request and the cache will be invalidated. In LC I put
"=12345678" at the end of the URL arguments or if there are no
arguments then I put "?seconds=12345678" where "12345678" is the LC "the
seconds". This only helps if there is no more than 1 request per second. 

___
use-livecode mailing list
use-livecode@lists.runrev.com
Please visit this url to subscribe, unsubscribe and manage your subscription 
preferences:
http://lists.runrev.com/mailman/listinfo/use-livecode

Re: Size of Image in RAM

[Sannyasin Brahmanathaswami via use-livecode]

@ Scott:

That's not what I get if I open this image in Photoshop.  (CC 2017)



575 X 1000 

http://wiki.hindu.org/uploads/img37.jpg


It is a JPEG, but under the mode menu it shows "RGB/8 Bit" and if I look under 
indexed colors it says "256"  definitely not 16bit (in which case we should see 
65536 colors)

So, somehow this IS a JPG that at least reports in PS as 8 bit but shows it is 
taking 2.17 MB in RAM, which it would if it were 16 bit.

Scott Rossi wrote:

When I generate an 8 bit indexed color image in Photoshop and look at the 
Save As menu, JPEG is not an option.  If I instead choose Save for Web which 
allows saving as JPEG, the resulting image re-opens in Photoshop in RGB (24 bit 
color) mode.

But, and this is interesting: if I save as PNG Photoshop offers the option to 
change turn off transparency and save as  bit… the resulting images on disk is 
1/3 bigger I size  (jpeg,:229K, png:353K)  But the former/jpg opens in 
photoshop at 2.17MB and not the 8-bit PNG opens at 739K in RAM  

This tends to confirm your theory and Photoshop is miss reporting the bit depth 
on the jpg (or something!)  … very interesting. the PNG at half the size in RAM 
is visually indistinquishable from the JPG (at least to my eyes)

http://wiki.hindu.org/uploads/img37.png

So we have a trade off: use 8 bit PNG and pack more data than we have room for 
in the mobile app package. or optimize really small JPG's on disk that then 
take up twice the room in RAM!



___
use-livecode mailing list
use-livecode@lists.runrev.com
Please visit this url to subscribe, unsubscribe and manage your subscription 
preferences:
http://lists.runrev.com/mailman/listinfo/use-livecode

Size of Image in RAM

[Sannyasin Brahmanathaswami via use-livecode]

I'm trying to optimize for Mobile.  Photoshop is playing tricks on me

given a 38K  jpg;

rect 3 X 5

552px w
736 px h
72 dpi (irrelevant for screen)

Open in Photoshop: it indicates 1.16M in RAM, but mode is 8 bit… but

but the online calculation sites for file size for that rect/bit-depth should 
make it only take up 398k or so, in RAM.

If I change the online calculator for that rect to "24 bit" it returns the 
exactly size  (1.16M) I'm seeing in photoshop…roughly 3X the size of the 8 bit, 
which is what we would expect.

So

1) Why is Photoshop reporting the files size as if it were 24 bit? and
2) Does LC have a function to check the size of an image in terms of RAM 
consumed? I couldn't find one in the dictionary.

BR  (thinking about affinity these days!)
___
use-livecode mailing list
use-livecode@lists.runrev.com
Please visit this url to subscribe, unsubscribe and manage your subscription 
preferences:
http://lists.runrev.com/mailman/listinfo/use-livecode

Re: Converting scripts in stacks to script only stack behaviors

[Sannyasin Brahmanathaswami via use-livecode]

In line
On 2/5/17, 3:41 PM, "use-livecode on behalf of Trevor DeVore via use-livecode" 
 wrote:

 With Levure adding stack files won't cause any conflicts. Each stack
is an independent file and the file that specifies which stacks to load is
a text file.


BR: I must be missing something. in the video you show how the stack files of 
the URL Dialog stack have been set in the stack properties of the binary 
livecode stack…so if those changed, the binary is changed Y/N?

but here you say 

"and the file that specifies which stacks to load is a text file."

 If I am testing standalone settings then I do my testing and
then reset the file to the last commit. 

BR: Oh… I should learn that one (GIT newbie here, barely crawling)

No need to stash the changes or
commit them.

___
use-livecode mailing list
use-livecode@lists.runrev.com
Please visit this url to subscribe, unsubscribe and manage your subscription 
preferences:
http://lists.runrev.com/mailman/listinfo/use-livecode

Re: JavaScript, Browser Object, VR with Hotspots?

[Sannyasin Brahmanathaswami via use-livecode]

On 2/5/17, 12:08 PM, "use-livecode on behalf of Mike Bonner via use-livecode" 
 wrote:

Thx for the info.  Honestly, all I want is for the example in the
dictionary to work for me.   (and yes, the javascript causes the lc side
handler to fire, but.. No parameters passed.)  I'll read through the stuff
you linked and see what I can learn.  Thanks!


Right… OK, well back to John's problem: 

Passing params to LC

I think you are right:

"Basically, pass the value of me.hotspot.id to livecode as part of the
function call, and add a switch block that gives a message based on the
hotspot id."

then, theoretically just click anywhere on the pano should fail to evoke the 
last call (to be tested) 

If it does, it goes back to: Browser widget is possibly caching the last event 
params?

I have more than a passing interest in this.  could be very useful. Those going 
Richmond's "simple image in native LC control" path may be easier. I don't know 
how far you can take the pano model with a still image before requiring the 
software to tile specialized sliced images that are dynamically stitched for 
the 360.  which is what Pano2VR is doing  (and also Kolor's similar software).

it can be done… a bit of a hack here because the distortion is a bit more than 
what the human eye would actually perceive  but it works, would not require 
expense pano software… and if you had A number of these you could create a 
robust VR with just JPGs. Assumes some front end production skills to create 
them of course. 

http://www.eisklotz.com/photography/360-panorama/360-panorama-jpg-for-free/

single image: 

view-source:http://www.eisklotz.com/wp-content/uploads/2016/01/IMG_2226-Panorama.jpg

he used blender for the stitching… 

only 243K… Adding hotspots and controlling the view with LC should be easy…



 



___
use-livecode mailing list
use-livecode@lists.runrev.com
Please visit this url to subscribe, unsubscribe and manage your subscription 
preferences:
http://lists.runrev.com/mailman/listinfo/use-livecode

Re: JavaScript, Browser Object, VR with Hotspots?

[Sannyasin Brahmanathaswami via use-livecode]

right, you will not be the first one to change  file on his own machine think 
it would magically appear in his browser.

 time for either a nap or a swim.

and "duh" here too… seems the 360 pano in native LC has been a long running 
topic on the forums already since 2007

http://forums.livecode.com/viewtopic.php?f=49=13430=64576=panorama#p64550

I'm not sure if it is as easy as I thought to get that 360 jpg to behave like a 
real pano in LC.  real good VR in LC still seems a bit far away.

I'm pursuing that discussion over at the forums.


 BR

On 2/5/17, 5:04 PM, "use-livecode on behalf of Mike Bonner via use-livecode" 
 wrote:

Still feel silly that I was editing local files.. and expecting it to
change what I was getting from the server.  (DO

___
use-livecode mailing list
use-livecode@lists.runrev.com
Please visit this url to subscribe, unsubscribe and manage your subscription 
preferences:
http://lists.runrev.com/mailman/listinfo/use-livecode

Re: Converting scripts in stacks to script only stack behaviors

[Sannyasin Brahmanathaswami via use-livecode]

@ trevor:  OK, understood. In fact we are doing a similar/ almost same thing 
though calling it "SivaSiva.json"  which is read on start up, 

re the stack files for the components: interesting… 

We have been adding *all* behaviors for all modules in the main stack, since 
many of these are reused by controls in different modules/components, but I can 
see if we externalize the *entire* stack script of a component as a behavior 
then, your layout is obviously the way to do it. Thanks for that! I have been 
thinking about this already this past week and wasn't clear, Levure shows the 
way forward… I’ll download 9 asap.

@ Mike

I think we may not be on the same page. There is certainly no "one file" for 
the entire project. Both in our SivaSiva app we have many "modules" aka 
"components" in Trevor's framework ("views" in a MVC way of looking at it) … 
OTOH  some modules benefit by having one more UI's on the same card, or at 
least different cards in the same stack.

Yeah.. I use to worry about the array of folder and files, but believe me, we 
have it pretty straight forward. Trevor's Levure and our new app are 
surprisingly close. Actually Ralf's RevIgniter was there way ahead of us…

 if you think about it there are only a few classes of objects, so you don't 
have to make it too complex:

assets # images, sound, text files, video, info files.
behaviors # code you may want to assign to controls in many different contexts
config # start up defaults store as plain text/json/yaml/ LC arrays whatever… 
libraries # code you want accessible everywhere all the time.
models  # re-usable code for doing "work/jobs"
modules  (aka components or "views")  # stacks users actually see
SuperApp.livecode # the "loader" stack

I don't think that is too much of a "jumble"  doesn't it add some obvious 
clarity?  If you want to "see" your app, open the components folder that's 
where the UI/views lives.… and this kind of frame work has a strong track 
record in web apps, CodeIgniter (PHP)  so it is well proven for usability.  

Anyway, the ability to diff previous commits and cross reference and "cherry 
pick" from other  branches has already saved me headaches in so many ways I 
can't ever go back.

AS for scary jumble snake pit, just expand some html5 apps, some of them are 
*really* scary. I unwrapped a Ionic app last month done in Angular2; it has 
20,000+ files! Yikes! Of course this is like seeing the entire LC engine IDE 
files set right alongside your own, but with LiveCode we don’t' have to do that.

BR

On 2/5/17, 6:04 PM, "use-livecode on behalf of Trevor DeVore via use-livecode" 
 wrote:

I'm referring to the app.yml file here. The app.yml file is what ties
together all of the files in the application folder into a single
application. It is where you specify which libraries, frontscripts,
backscripts, and other stacks make up the application. If I add a new UI
stack named "Prompt User Prior to Delete" no other stack files will be
modified. The only file that might be updated is the app.yml file and that
is unlikely as well. In the app.yml file you can specify that the contents
of a specific folder are components and Levure will treat all of the
contents of that folder as such.



___
use-livecode mailing list
use-livecode@lists.runrev.com
Please visit this url to subscribe, unsubscribe and manage your subscription 
preferences:
http://lists.runrev.com/mailman/listinfo/use-livecode

8.1.3 (RC1) Invoking Script Editor Failing Everywhere

[Sannyasin Brahmanathaswami via use-livecode]

I need the updates in 8.1.3 (RC1) so took  the leap.

IDE is problematic… deleted my prefs before opening… but now

-- the point ertool icon does not change from browse to pointer…
-- switching to the pointer too and double clicking on a button runs the button 
script instead of just selecting the button.  (it used to open the property 
inspector for the selected group/object.)
-- cmd-opt click will not cause the script editor open;  (yes, I have it set in 
the preferences to do that, but it's not working)
-- from the Project Browser, double clicking on in the icon with the number of 
lines in a script does not invoke that script in the SE
-- opening the PI for an object with an assigne behavior: click "edit script" 
the script for the assigne behavior does not open

Help!

BR



___
use-livecode mailing list
use-livecode@lists.runrev.com
Please visit this url to subscribe, unsubscribe and manage your subscription 
preferences:
http://lists.runrev.com/mailman/listinfo/use-livecode

Re: Delete element from array

[Sannyasin Brahmanathaswami via use-livecode]

Thank. Yes.. it is there in the dictionary but I glossed over it somehow. Not 
sure the new dictionary is an improvement of the one we had all these years… 

@Bernd: is your dictionary plug in working in the latest 8+ versions of LC? 


 

On 1/24/17, 5:04 PM, "use-livecode on behalf of Phil Davis via use-livecode" 
 wrote:

delete variable tMyArray["this key"]



___
use-livecode mailing list
use-livecode@lists.runrev.com
Please visit this url to subscribe, unsubscribe and manage your subscription 
preferences:
http://lists.runrev.com/mailman/listinfo/use-livecode

Run processes/effects simultaneously with a visual effect ?

[Sannyasin Brahmanathaswami via use-livecode]

I'm looking to "dress" up our new app with more elegant visuals.

I was using send in time to a handle that changes the blendlevel of small 
objects… this works well and is none blocking.

But someone asked "Why don't you just use dissolve?"

Duh.. the dictionary doesn't give the platforms that work with dissolve but 
apparently, it runs just fine on iOS and Android.

So,  I did some tests…wondering about the engine's timing work with visual 
effects:

create stack
place image
put two buttons "hide"  "show" and one field "bool"

Btn 1:
on mouseUp
put empty into fld "bool"
hide image burger with visual effect dissolve very slowly
put "false" into fld "bool"
answer "Are U Hidden Yet?" with "Yes" or "No"
end mouseUp

btn 2:
on mouseUp
put empty into fld "bool"
show image burger with visual effect dissolve very slowly
put "true" into fld "bool"
answer "Are U Showing Yet?" with "Yes" or "No"
end mouseUp

that visual effect is blocking. The commands that follow do not run until the 
visual effect is done, effectively visual effect is a "wait"

Also the dictionary says if you put two visual effects one after another they 
run consecutively.

but what if you want 2 visual effects to run simultaneously  and also have 
another process (like type txt or invoke a dialog etc) occur at the same time?

typical example:, let's say I want to create a cross fade between two images. 
The show and hide for each one with dissolve would need to run simultaneously.  
I'm using "very slowly" for testing, in a real app we would probably use very 
fast or fast.

img 1: "burger:
img 2: "shot"

on mouseUp
hide image burger with visual effect dissolve very slowly
show image shot with visual effect dissolve very slowly
end mouseUp


on mouseUp
put empty into fld "bool"
show image burger with visual effect dissolve very slowly
hide image shot with visual effect dissolve very slowly
put "true" into fld "bool"
answer "Are U Showing Yet?" with "Yes" or "No"
end mouseUp

in both cases the first effect runs and then the next one.. I don't get a 
"cross fade"  effect. in the latter, first the hidden image appears (behind the 
one in front) then the front one dissolves, then the fld is filled and then the 
answer dialog opens…

locking and unlocking the screen doesn't help

Are simultaneous visual effects + some process even possible? The dictionary 
does not explicitly declare visual effect to be blocking. But all my tests 
indicate it is.  I hope I am wrong.

BR







___
use-livecode mailing list
use-livecode@lists.runrev.com
Please visit this url to subscribe, unsubscribe and manage your subscription 
preferences:
http://lists.runrev.com/mailman/listinfo/use-livecode

Script editor chops "end handler" to "nd handler"

[Sannyasin Brahmanathaswami via use-livecode]

I have a bug report on this but HQ says they cannot replicate it.

Today I think I have a definitive recipe. Can you try and confirm?

Symptom:

the "e" at the beginning of the last line end of handlers is getting chopped off

on closeStack
if the environment is "mobile" then
scroller_Delete "scrolling lesson field" # not actually used
end if
pass closeStack
end closeStack   # Ok put your cursor right here immediately after closestack 
and hit return and we get this

on closeStack
if the environment is "mobile" then
scroller_Delete "scrolling lesson field" # not actually used
end if
pass closeStack
nd closeStack
# cursor is here now. but the word "end" above has become "nd"

first seen in 8*  (I think)  and persists in 8.1.3


Can anyone else replicate this? (mac Sierre lc 8*

BR

___
use-livecode mailing list
use-livecode@lists.runrev.com
Please visit this url to subscribe, unsubscribe and manage your subscription 
preferences:
http://lists.runrev.com/mailman/listinfo/use-livecode

Snippet to show status bar on mobile

[Sannyasin Brahmanathaswami via use-livecode]

Do we have control over the layer of the native device top status status bar?

is it always on topmost layer?  if so then we have to dynamically hide and show 
it if we want to occupy the whole screen

OR

is it always on the background layer

if so then we simply have some objects on higher top layer go all the way to top

Your experience with this, share here will be helpful

Thanks!


___
use-livecode mailing list
use-livecode@lists.runrev.com
Please visit this url to subscribe, unsubscribe and manage your subscription 
preferences:
http://lists.runrev.com/mailman/listinfo/use-livecode

Re: 8.1.3 (RC1) Invoking Script Editor Failing Everywhere

[Sannyasin Brahmanathaswami via use-livecode]

Turns out it was one of my plug ins… you can see the bug report where the 
problem was attributed to the DataGrid Helper… but I don't use the DGH and I 
had problems with an old old stack I've had for 15 years called 
"Favorites.livecode" build circa 2000 or earlier and migrating forward over 
time. Small toolbox, with launch menu, some buttons for testing "stuff" etc… 

there's really nothing unusual about its properties, it's not lock or 
cantDelete or anything like that. no odd open stack or preopenstack handlers 
that would do weird things, but if it is in my plugins folder 8.1.3 "dies in 
chaos"  take it out, put in others (like your 4thworld ones andsome of Scotts) 
it works just fine.

I sent the stack in with the bug report.  It *is* quite possible that there is 
some horrible legacy code in a button for a test done in 2002 or something… and 
the handler contains lines that are "bad code" from 8.1.3… but I don't think 
the IDE steaps through *all* the code of every plugin… so??

lead me to the question: we know here that some legacy documents that were, for 
example created in Quark Express, then exported to InDesign then used as 
templates over several years.. will gradually become corrupt in mysterious ways 
are indiscernable to the user… presumably at some binary level that does not 
actually cause the program to crash but enough to cause a lot of trouble

could the same be with old, old, old LC stacks?

We'll see it will be interesting to find out what HQ spots….

> -- from the Project Browser, double clicking on in the icon with the
>number of lines in a script does not invoke that script in the SE
> -- opening the PI for an object with an assigne behavior: click "edit
> script" the script for the assigne behavior does not open

I'm unable to reproduce those last two; both work as expected here.

___
use-livecode mailing list
use-livecode@lists.runrev.com
Please visit this url to subscribe, unsubscribe and manage your subscription 
preferences:
http://lists.runrev.com/mailman/listinfo/use-livecode

Re: Delete element from array

[Sannyasin Brahmanathaswami via use-livecode]

Others can please chime in here… I'm not sure exactly why the new dictionary is 
"irritating" but it is. 

let's say I want to see all the properties for buttons. I think the old 
dictionary gave us that option.

Now if I click on "association "button"  and click type "properties" I get the 
button props, but the top window on the right only shows a only a few

there was a view in the older dictionary that had a much longer list… I would 
prefer myself a left column that was really long so I can see at a glance all 
the props that the vertical column height will allow, depending on how the user 
has the window sized… and the definitions on the right take up the whole 
vertical space i.e. split window horizontally.  

IMHO we get: 
1) to see a longer list of dictionary entries with out scrolling and
 2) can read more of the definition without scrolling.  to conserve screen 
space I could narrow the window and the definition area on the right panel 
would narrow responsively.

OK now if I type "htmltext" into the search, it comes up empty… why? well of 
course… I am now "stuck" in

associations: "button"
type: "property"

I'm scratching my head… what is happening?  Let's say I have a misconception 
that htmltext has something to do with buttons (it does not, but newbies will 
not know that) so I'm just stuck… my search gave me nothing.. I'm trying to 
understand buttons.  the right and left arrow don’t do anything… 

In the far left column you need a "clear" button or something… I could click 
that, and then my search for "htmltext" would return results (without having to 
type again.)  Only now as I test did I discover that the way to dismiss the 
filters is to click on the filter label "button" or "properties" to "back out" 
of where you are… that is UI "mysterious hidden behavior" that breaks all the 
rules for any good UI design… at least put a small X  top the right of each 
filter or a "clear all" or something.  I don't mind a few "mystery meat" icons… 
any 8 year old today just clicks on everything.. but you don't even have a 
control to reset the space back to "top.. 

OK so I dismiss the filters on the left (but inadvertently leave my search 
string  "htmlText" in the box top right) click on group filter, database list 
of terms is empty… Oh.. .duh I have to go up and manually delete my search 
string…*now* the list appears again..

Then there's the two little arrows at the top next ot a clock… what do they do? 
hmm, I go to  control structure filter… click on  the left arrow at the top and 
takes me to "revCloseCursor"  Why?  Oh wait, I guess that was my previous view… 


It just feels more like a "burden" now to look for information. Perhaps others 
with more UI/UX experience can offer thoughts on this.  I'm using it more and 
more now that I'm away from server side code and code the UI again (too much 
fun…) Maybe its just me… and I need to "get used to it" like riding a motor 
cycle instead of a bike….

On 1/26/17, 1:32 PM, "use-livecode on behalf of Ali Lloyd via use-livecode" 
 wrote:

Hi Brahma,
Goes without saying perhaps but it would be extremely helpful if you could
articulate *why* you find this information harder to come by in the current
dictionary. Thanks!



___
use-livecode mailing list
use-livecode@lists.runrev.com
Please visit this url to subscribe, unsubscribe and manage your subscription 
preferences:
http://lists.runrev.com/mailman/listinfo/use-livecode

Re: Run processes/effects simultaneously with a visual effect ?

[Sannyasin Brahmanathaswami via use-livecode]

Ha! Klaus.. Jai Ganesha! Awesome this works (yet to test on mobile)


on mouseUp
  
set the effectrate to 500

   lock screen for visual effect

   put the long Id of image "stone" into pHiddenImage
   put the long Id of image "night" into pVisibleImage

   show tHiddenImage 
   hide tVisibleImage 

   unlock screen with visual effect dissolve very slowly

end mouseUp


so we can set up a "command crossFade pImg1, pImg2" in a library and pass the 
long ID's to it

command crossFade pHiddenImage,pVisibleImage

set the effectrate to 500

lock screen for visual effect
show pHiddenImage
hide pVisibleImage

unlock screen with visual effect dissolve very slowly

end crossFade

let me see how much trouble this gets me into ☺  while trying to do other 
things before and after the fade… 

Brahmanathaswami


 
Klaus major- wrote:

Yes, visual effects are in fact blocking unfortunately!

But maybe you can use something like:
...
set the effectrate to 500
lock screen for visual effect
show image "burger"
hide image "shot"
unlock screen with visual effect dissolve very slowly
...
?

___
use-livecode mailing list
use-livecode@lists.runrev.com
Please visit this url to subscribe, unsubscribe and manage your subscription 
preferences:
http://lists.runrev.com/mailman/listinfo/use-livecode

Re: 8.1.3 (RC1) Invoking Script Editor Failing Everywhere

[Sannyasin Brahmanathaswami via use-livecode]

Ooops sorry  I should have added that.

http://quality.livecode.com/show_bug.cgi?id=19120

copy of my stack is there, nothing proprietary in it… it does not do any front 
scripts.

but there is definitely something in it that 8.1.3 does not like (and also 
lcStackBrowser was unhappy with this stack also for equally unknown, mysterious 
reasons.


 

On 1/27/17, 7:30 AM, "use-livecode on behalf of Richard Gaskin via 
use-livecode"  wrote:

Report #/URL?

___
use-livecode mailing list
use-livecode@lists.runrev.com
Please visit this url to subscribe, unsubscribe and manage your subscription 
preferences:
http://lists.runrev.com/mailman/listinfo/use-livecode

Re: Delete element from array

[Sannyasin Brahmanathaswami via use-livecode]

I'm game to play with this… I need it

where are these array files?

Do we want to change the thread subject to "Dictionary Improvements"
 

On 1/27/17, 6:55 AM, "use-livecode on behalf of Richard Gaskin via 
use-livecode"  wrote:

The current Dictionary content is in wonderfully well-separated array 
files, making it easier than ever for third parties to use existing 
internal APIs to deliver the content in any form we choose.

___
use-livecode mailing list
use-livecode@lists.runrev.com
Please visit this url to subscribe, unsubscribe and manage your subscription 
preferences:
http://lists.runrev.com/mailman/listinfo/use-livecode

Re: Script editor chops "end handler" to "nd handler"

[Sannyasin Brahmanathaswami via use-livecode]

FWIW:

http://quality.livecode.com/show_bug.cgi?id=19145



 

On 1/27/17, 4:46 AM, "use-livecode on behalf of panagiotis merakos via 
use-livecode"  wrote:

Hi guys,

Thanks for spending time on this. Indeed, I can now replicate this bug. As
you said, it happens only if the code is not indented. Thus, make sure you
press "Tab" key before editing your un-indented script. I will update the
bug report.

Best,
Panos

___
use-livecode mailing list
use-livecode@lists.runrev.com
Please visit this url to subscribe, unsubscribe and manage your subscription 
preferences:
http://lists.runrev.com/mailman/listinfo/use-livecode

Re: Launch URL with LInk Back to LC App

[Sannyasin Brahmanathaswami via use-livecode]

Ralph

Thanks, hacking the pList definitely works!

I guess Apple doesn't care, right?



 

On 1/30/17, 10:18 AM, "use-livecode on behalf of Ralph DiMola via use-livecode" 
 wrote:

JLM, Thanks for the clarification. BR, Another way to do this without using
the un-supported plist hack is to save the state of your app before you
launch the browser and then restore the app to the saved state when it
restarts.



___
use-livecode mailing list
use-livecode@lists.runrev.com
Please visit this url to subscribe, unsubscribe and manage your subscription 
preferences:
http://lists.runrev.com/mailman/listinfo/use-livecode

Browser Widget/HTML5/LC Integration

[Sannyasin Brahmanathaswami via use-livecode]

How many devs here would be willing to pitch in $ to cover the cost for HQ to 
add "deep integration" between the browser widget and Livecode.

Currently there is a "wall" between LC and the browser widget. Jonathan Lynch, 
myself and Andre have been looking into this very closely. There is no way for 
a javacript to "ping" LC and get an immediate response.

Our use case is simple: the browser widget by default can only store data in 
the web indexDB space.  But if you want to run a SQLLite database in your 
mobile app, and use html5 for either, complete robust small app modules or even 
as light weight "views" … you need to have the browser widget act and behave 
and communicate with the LC back like any other control. Presently if, e.g. you 
issue click down in the browser widget to fire some handler in LC (to store 
some data in the database)  and the LC handler is supposed to store the data.. 
that handler's response never gets back to the Javascript. You have to start 
building up crazy solutions where you set values in LC and then other JS 
handler check those… it turns into "call back hell" with no means to control 
the timing.

I've as HQ to quote on this… I'm wondering if anyone else is interested. It 
expands the horizons for you team, as you could engage some young html5 wizard 
to build stuff, that could be fully integrated into your LC platform.

Brahmanathaswami






___
use-livecode mailing list
use-livecode@lists.runrev.com
Please visit this url to subscribe, unsubscribe and manage your subscription 
preferences:
http://lists.runrev.com/mailman/listinfo/use-livecode

Re: Community Dictionary (was Re: Delete element from array)

[Sannyasin Brahmanathaswami via use-livecode]

Richard Gaskin wrote:

/Documentation/resources/data/api/api.sqlite

Oh boy, gonna have fun with that!

BR

___
use-livecode mailing list
use-livecode@lists.runrev.com
Please visit this url to subscribe, unsubscribe and manage your subscription 
preferences:
http://lists.runrev.com/mailman/listinfo/use-livecode

Re: Browser Widget/HTML5/LC Integration

[Sannyasin Brahmanathaswami via use-livecode]

So as not to burden this list which  "use" list and not a "complain" list  I 
have responded to you at length directly.

in a word:

1) the graphics rendering layer of LC is behind the graphics rendering layer of 
the browser

2) an IDE that provides a tool box for super elegant 21st century UI  -- 
smooth, responsive, easy to change the Look/Feel
even my attempts with AnimationEngine  which is, I believe as good as it's 
going to ever get with LC's engine, have a clunky, jerky feel
e.g. try a ken burns: move the image while simultaneously increasing is rect in 
proportion. even rotating a ball around a circular path… has this subtle "1990" 
feel to it vs the same thing done with CSS animation in the browser. Maybe not 
an issue for all those whose use case is business/systems/tools etc. but in our 
shop it has to look fabulous or the horse is not considered a viable vehicle 
for the content delivery. Period end of discussion.

I don't know why or the technical issues behind that… I would be interested to 
know to understand it

3) robust surport for SVG (not just a single color) on top of 1 and 2 above

even shorter summary

vector perfect, smooth as glass, motion graphics

Isn’t the dictionary html in a widget? Why? There's one answer to your question 
from HQ itself.

If a player serves to provide the space to show content develop on another 
platform (video/Premiere)

Why not the browser widget to provide a space to show HTML5?

I believe it has tremendous potential for engagement across a wide spectrum of 
developers, going both ways.

More in the email I sent you

BR



 

On 1/27/17, 12:21 PM, "use-livecode on behalf of Richard Gaskin via 
use-livecode"  wrote:

There seems to be another, perhaps bigger, question implied here:

What needs to be added in LC which would make it easier to develop GUIs 
in than HTML/CSS/JavaScript, so that you wouldn't need to split your 
work between such very different paradigms?

___
use-livecode mailing list
use-livecode@lists.runrev.com
Please visit this url to subscribe, unsubscribe and manage your subscription 
preferences:
http://lists.runrev.com/mailman/listinfo/use-livecode

Re: Snippet to show status bar on mobile

[Sannyasin Brahmanathaswami via use-livecode]

found it!

mobileHideStatusBar

mobileShowStatusBar

Now I just need to "slide" everything (the entire card/content) down… 

tks
 

On 1/27/17, 10:19 AM, "use-livecode on behalf of J. Landman Gay via 
use-livecode"  wrote:

It is part of the OS and so overlays the stack. You can hide or show it, 
but it isn't otherwise under LC control. Think of it like the MacOS 
dock, menubar, or any other system element.

___
use-livecode mailing list
use-livecode@lists.runrev.com
Please visit this url to subscribe, unsubscribe and manage your subscription 
preferences:
http://lists.runrev.com/mailman/listinfo/use-livecode

Re: Browser Widget/HTML5/LC Integration

[Sannyasin Brahmanathaswami via use-livecode]

Sure.. try this:

go stack url "http://wiki.hindu.org/uploads/SivasHawaiiRevels_v001.livecode;

it's so "bad" as it is "just below par" enough to cause others to blink at it.

I realize this seems like a challenge but take my word for it.  Here in our 
shop  we have several Keynote presentation masters that can make a keynote look 
almost like a good short flick-movie done in Hollywood.

 One html5 man who work in Angular and soon React. We also develop videos, 
movies on a monthly basis…and we have hired out some HTML5 development, and I 
get to see the iterations on development of those media modules in any browser; 
so it's not like this is just opinionated griping.

We are working in a very real world media context where *everything*  is about 
what you see and how it moves and how easy/hard it is to develop the front end 
(what you see)

And that is the other issue -- developing the "views"  Again this not opinion 
but based very real hard experience.

Take the new home screen for our new app (I'll let everyone download from GIT 
soon if you are interested)

http://wiki.hindu.org/uploads/home-screen-1.jpg 
http://wiki.hindu.org/uploads/home-screen-2.jpg

This is in LC with a simple panel set of a group that has to scroll up and down.

You have no idea how much time was spend ($) getting this to work easily in LC, 
including the ability to change things very quickly across all parts of the 
whole group.

This same "view" could have been done Angular or React in a few hours. And, 
once done, be responsive, scrollable and any attribute that stake holder 
thought should be changed -- done on the spot with the minor adjustment of a 
single integer value in an CSS rule.

Put another way: a competent html5 dev can build gorgeous, responsive, easily 
tweakable GUI in 1/10 the time it takes to achieve the same level of production 
values in Livecode.

OTOH when it comes to coding the controllers/ models (libraries and scripts) my 
experience is that even html5 dev with a lot of experience in JS will still 
take 5 times longer to code the JS than Jacqueline who knows only LIvecode. And 
in the end Jacque's code is actually shorter and readable.

So: that's my point:  the ability to create/use HTML5 for "views" with 
lightweight JS that can talk to a robust LC back end, like any other LC 
control,  that would be a total snake pit if it were all done in JS.. it would 
require a level of expertise (and the money to pay for expertise hours) 

We even have a small project done for "peanuts" by a very low priced HTML5 man 
in Belaruz. The UI is so simple but gorgeous animation , but cost for just a 
tiny framework, even at low rates is very high relative to the output. An LC 
developer code out the module in 1/10 the time/cost  . 

I would past the URL to see what he has done, but he changed the framework on 
the back end so that the page would load faster  -- another "cost" of the 
JS/Python/Node code requirements to put up a simple animated splash screen. 

Point: html + CSS + a small bit of JS  = Marvelous UI

+

LC on the backend

dynamite combo! 



 

On 1/28/17, 12:46 AM, "use-livecode on behalf of hh via use-livecode" 
 wrote:

BR wrote:
"Even my attempts with AnimationEngine which is, I believe as good as it's
going to ever get with LC's engine, have a clunky, jerky feel."

I tried to generate such "bad" graphics in LC without any success. So:
Could we see (at least one of) such attempts? Don't forget the source code.



___
use-livecode mailing list
use-livecode@lists.runrev.com
Please visit this url to subscribe, unsubscribe and manage your subscription 
preferences:
http://lists.runrev.com/mailman/listinfo/use-livecode

Re: Browser Widget/HTML5/LC Integration

[Sannyasin Brahmanathaswami via use-livecode]

thanks for these tips… I will try again.



Svasti Astu, Be Well 
Brahmanathaswami

www.himalayanacademy.com

 

On 1/28/17, 6:20 PM, "use-livecode on behalf of hh via use-livecode" 
 wrote:

It's not the LC graphics it's the "jagged" and "uncomplete" scripting.
LC can do much more.

Some basic examples related to the starting animation:
= You use "wait " for animations (!!!) and moreover
  +++ with  far below the refreshrate of the app. 
  +++ with  independent of the device.
= You animate along a random (jagged) line of points ignoring the
  existence of smooth curves.
= You use an image that doesn't even have an antialiased border.



___
use-livecode mailing list
use-livecode@lists.runrev.com
Please visit this url to subscribe, unsubscribe and manage your subscription 
preferences:
http://lists.runrev.com/mailman/listinfo/use-livecode

Re: Browser Widget/HTML5/LC Integration

[Sannyasin Brahmanathaswami via use-livecode]

I forget to give the URL for a complete web app that I was able to put into a 
LC browser widget  (too heavy for delivery becuase of all the audio content, 
but good proof of concept for using the brower widget to drive local html5 
content in the same package: 

www.himalayanacademy.com/apps/spiritual-workout

  - This is a recent app we released, HTML5 done in Ionic/Angular2

___
use-livecode mailing list
use-livecode@lists.runrev.com
Please visit this url to subscribe, unsubscribe and manage your subscription 
preferences:
http://lists.runrev.com/mailman/listinfo/use-livecode

Re: Browser Widget/HTML5/LC Integration

[Sannyasin Brahmanathaswami via use-livecode]

@ Richard.

OK, one more SVG example (I believe that's what you were asking for)

now you have three

http://wiki.hindu.org/uploads/floor-plan.svg # not responsive, but could be

http://wiki.hindu.org/uploads/SivaSivaDiagram.svg  
   # not responsive, but could be (though that's not wanted in this scenario  
   # imagine a "map world" you could explore…so it stays full size, user zooms 
in and out and swipes
   
http://www.himalayanacademy.com/assets/img/khm-logo-full-horizontal.svg

So is SKIA in LC's future?

Again, I'm not clear about graphics engines issues.

BR

 

On 1/29/17, 8:29 AM, "Richmond Mathewson via use-livecode" 
 wrote:

If that is meant to be a series of mobile phone outlines in a sort of 
flowchart the thing is
viewable, but unlike the Himalayan Academy logo is doesn't scale at all.

Richmond.



___
use-livecode mailing list
use-livecode@lists.runrev.com
Please visit this url to subscribe, unsubscribe and manage your subscription 
preferences:
http://lists.runrev.com/mailman/listinfo/use-livecode

Re: Browser Widget/HTML5/LC Integration

[Sannyasin Brahmanathaswami via use-livecode]

oops sorry, I was trying to make is responsive and broke something

try this now

http://wiki.hindu.org/uploads/SivaSivaDiagram.svg


 

On 1/29/17, 8:14 AM, "use-livecode on behalf of Richmond Mathewson via 
use-livecode"  wrote:

http://wiki.hindu.org/uploads/SivaSivaDiagram.svg

___
use-livecode mailing list
use-livecode@lists.runrev.com
Please visit this url to subscribe, unsubscribe and manage your subscription 
preferences:
http://lists.runrev.com/mailman/listinfo/use-livecode

Re: Browser Widget/HTML5/LC Integration

[Sannyasin Brahmanathaswami via use-livecode]

as mentioned… it's not responsive, that was just a quick export, would require 
some tweaking.

and the use case envisioned for larger ones is different: use pinch and zoom 
and swipe to move around and go in on the "map", this means we actually do not 
want it to scale down to fit….



On 1/29/17, 8:29 AM, "Richmond Mathewson via use-livecode" 
 wrote:

If that is meant to be a series of mobile phone outlines in a sort of 
flowchart the thing is
viewable, but unlike the Himalayan Academy logo is doesn't scale at all.

Richmond.



___
use-livecode mailing list
use-livecode@lists.runrev.com
Please visit this url to subscribe, unsubscribe and manage your subscription 
preferences:
http://lists.runrev.com/mailman/listinfo/use-livecode

Re: Browser Widget/HTML5/LC Integration

[Sannyasin Brahmanathaswami via use-livecode]

Until we hear that HQ is listening or want to know more I'll keep it short

1) SVG requirements

rendering this would be a minimum. If here is a way to do it now in LC let me 
know

http://www.himalayanacademy.com/assets/img/khm-logo-full-horizontal.svg

resize your browser, it remains perfect!

More robust example… not currently set to resize, (but could be)  

http://wiki.hindu.org/uploads/SivaSivaDiagram.svg  

2) Yes I am using the browser widget to show pages on our web site. Even that 
has lots of issues, but we are working with it. and HQ has been responsive on 
the issue of thecontent rendering in odd rects.. though I am still seeing this 
in 8.1.3  need to do more testing in the next few days… we made be forced out 
to use launch URL and send the user off the LC app into the browser app on the 
phone.. (rather not have to do that)

3) HTML Integration has two use cases (in my current purview, there could be 
more)
   
  i. simple views

this can be created in less than 5 minutes the view consists of  a few lines


 Mangos
 Mangos
 Papayas
 Chiku


touch of CSS and you get this:

http://wiki.hindu.org/uploads/gui-eg.html

responsive, easily tweaked if the content people want another four rows in the 
panel, cut and paste, changing props, sure, it's a text file, but so what… 
change one integer value and all rows change.

This simply cannot be done that easily created in Livecode. (Please don't 
respond with "yes we can make this in LC" I know you can, but to scale change 
and make it responsive change colors on all objects in the group at once etc.. 
you need to be an old xTalk wizard with toolboxes)

  Now if touch events, JS handlers, for a small HTML5 widget/view could talk 
back to LC controllers and models and actually call a function and get the 
return value synchronously then we have a power house  combination… 

build simple gui views in html… robust LC backend.
  
  ii. full web app integration
   - This is a recent app we released, HTML5 done in Ionic/Angular2
   - I can actually copy the core pages/js/css assets and drop them in a 
folder in an LC app and point the browser widget (URL) to the index.html file 
and the app runs perfectly, right inside the browser widget. 
  -- trouble is I can't really get a robust handshaking going on. LC can 
"poke" the app (but get nothing in return) and the web app can "poke" LC (but 
get nothing in return)  and we need more for true integration
   This is more like Jonathan's use case: It's not only about creating views, 
it's also about the possibility of wrapping existing complex web apps in an LC 
framework… but we need the browser widget's javascript handlers to be, 
basically, in the message path up and down.

HQ gave me a quote on doing this, but it's out our budgets, I fishing for 
anyone else who might be interested to pitch in.


 back to work here..

On 1/28/17, 3:45 PM, "Richard Gaskin via use-livecode" 
 wrote:

Sannyasin Brahmanathaswami wrote:

> So as not to burden this list which  "use" list and not a "complain"
> list  I have responded to you at length directly.

...ands yet here we are, so we might as well enjoy the discussion.

> 1) the graphics rendering layer of LC is behind the graphics
> rendering layer of the browser

Likely, but since both Chrome and LC use Skia we'd have to look a little 
deeper.

Specifics are helpful.


> 2) an IDE that provides a tool box for super elegant 21st century UI
>  -- smooth, responsive, easy to change the Look/Feel even my attempts
> with AnimationEngine  which is, I believe as good as it's going to
> ever get with LC's engine, have a clunky, jerky feel e.g. try a ken
> burns: move the image while simultaneously increasing is rect in
> proportion. even rotating a ball around a circular path… has this
> subtle "1990" feel to it vs the same thing done with CSS animation
> in the browser. Maybe not an issue for all those whose use case is
> business/systems/tools etc. but in our shop it has to look fabulous
> or the horse is not considered a viable vehicle for the content
> delivery. Period end of discussion.

That's a good use case.  UIs are increasingly animated, so animation 
performance is increasingly useful.

Anyone on the team listening in and can offer some insight into this?


> 3) robust surport for SVG (not just a single color) on top of 1 and 2
> above

What limitations currently exist in the SVG widget?

Let's identify the remaining specifics and see if we can get a plan 
together to handle those.


> Isn’t the dictionary html in a widget? Why? There's one answer to
> your question from HQ itself.

I believe the Dictionary is indeed using the Browser widget.

My question is why, since that content doesn't require 

Re: Browser Widget/HTML5/LC Integration

[Sannyasin Brahmanathaswami via use-livecode]

@ Bernd  It's good to hear there is some progress! thanks for testing.  I will 
look into Peter's files but also hope for the day when this is in the product. 

@ Richard, No, as you can see they will not display in the widget.  But here 
appears to be hope. I was not aware of the big shift over to SKIA (nor do I 
really know the scope of what that means in terms of future possibilities)

I'll do a small test loading these into a browser widget..

 

On 1/29/17, 11:41 AM, "BNig via use-livecode"  
wrote:

Sannyasin Brahmanathaswami wrote:

 > OK, one more SVG example (I believe that's what you were asking for)
 >
 > now you have three
 >
 > http://wiki.hindu.org/uploads/floor-plan.svg # not responsive, but
 > could be
 >
 > http://wiki.hindu.org/uploads/SivaSivaDiagram.svg
 ># not responsive, but could be (though that's not wanted in this
 > scenario
 ># imagine a "map world" you could explore…so it stays full size,
 > user zooms in and out and swipes
 >
 > http://www.himalayanacademy.com/assets/img/khm-logo-full-horizontal.svg

Do those render in LC's SVG widget?


 > So is SKIA in LC's future?

Skia replaced LC's older graphics subsystem many versions ago (5.x?). 
They did such a good job that most folks never noticed.  But where you 
see new capabilities like the SVG widget, those are among the many 
upsides of having switched to this popular graphics system.

-- 
  Richard Gaskin


Peter Thirkell posted a -lcb file that has to be compiled using the
Extension Builder that he uses to display simple color SVGs. No gradients
and no advanced features of SVG.
http://forums.livecode.com/viewtopic.php?f=93=27811=150208#p150208
He parses svg export from Affinity Designer.

That said I scraped two of the above mentioned SVGs, after exporting them
from Affinity Designer I was able to feed the logo to Peter's hacked SVG
Color widget and it displayed fully. Rescalable and everything.

The floor plan lost the annotational text and some fills (because of
gradients) but otherwise rendered ok.

I did not attempt to use the SivaSivaDiagramm because the source code was
full of things Peter's widget can not display.

I recommend anybody interested in widgets to look at Peter's .lcb file and
especially appreciate his parsing handlers.

That all said what the svg widget is lacking is a more advanced svg parser
beyond black and white and transparency. Mark Waddingham had started on
implementing an open source parser but apparently Infinite Livecode kept him
from finishing it yet.
https://github.com/livecode/livecode/pull/3089

Kind regards
Bernd



--
View this message in context: 
http://runtime-revolution.278305.n4.nabble.com/Browser-Widget-HTML5-LC-Integration-tp4712003p4712034.html
Sent from the Revolution - User mailing list archive at Nabble.com.

___
use-livecode mailing list
use-livecode@lists.runrev.com
Please visit this url to subscribe, unsubscribe and manage your 
subscription preferences:
http://lists.runrev.com/mailman/listinfo/use-livecode

___
use-livecode mailing list
use-livecode@lists.runrev.com
Please visit this url to subscribe, unsubscribe and manage your subscription 
preferences:
http://lists.runrev.com/mailman/listinfo/use-livecode

Launch URL with LInk Back to LC App

[Sannyasin Brahmanathaswami via use-livecode]

Use case is: multiple apps and the local  mobile browser, all in a navigable 
ecosystem on small devices.

on iOS, you can make an app that does this

1) Open to home screen
2) go to card "Virtual Tours"
3) click on "Gardens"
4) go to another card name "gardens" (just an example)
5) click a button that has a script like

On  mouseup
launch url http://www.kauai-botanical-gardens.com
end mouseup

Ok build, load and run on the iPhone…

if you click through in the app to card gardens and click on the link and it 
takes you to Safari on iPHone and Chrome on Android.

in iOS you will get a small "< MyApp" in the upper left corner of Safari. 
Android of course has it's default back button.

on iOS if you click that button  "

Re: Browser Widget/HTML5/LC Integration

[Sannyasin Brahmanathaswami via use-livecode]

Hmm, interesting

try this

http://wiki.hindu.org/uploads/SivaSivaDiagram.svg  # in firefox; chrome; LC 
browser widget

result:

 Firefox renders the gradients.
Chrome and the LC browser widget render some place holder in the same space.


On 1/29/17, 11:41 AM, "BNig via use-livecode"  
wrote:

http://wiki.hindu.org/uploads/SivaSivaDiagram.svg

___
use-livecode mailing list
use-livecode@lists.runrev.com
Please visit this url to subscribe, unsubscribe and manage your subscription 
preferences:
http://lists.runrev.com/mailman/listinfo/use-livecode

Re: Storing and object reference in a var

[Sannyasin Brahmanathaswami via use-livecode]

Exactly

I'm studying scripts by some "pros" (Jacque, David, Malte-animation engine)

in order to get a group on creating custom visual effect that I can use "all 
over the place" but creating an external script/text stack

I've been doing server side script for so many years, the whole UI coding is 
all new but a lot more fun!

so If you want "safe" generic handlers… and you don’t' want to worry about the 
name space collisions you need to use the long ID

to drive things like this

put the long id of  grp "lessonTextControl"  into pEffectsTarget
send "object_FadeOut pEffectsTarget,0, 2, 20" to me 
wait for 2 seconds with messages


#then in your library:

command object_FadeIn  pEffectsTarget, newBlendLevel, someIncrement, someDelay
   set the blendLevel of  pEffectsTarget  to newBlendLevel
   if newBlendLevel > 0 then
  subtract someIncrement from newBlendLevel
  send "object_FadeIn pEffectsTarget, 
newBlendLevel,someIncrement,someDelay" to me in someDelay milliseconds
   else
  set the blendLevel of pEffectsTarget to 0
   end if
end object_FadeIn

command object_FadeOut  pEffectsTarget, newBlendLevel, someIncrement, someDelay
   set the blendLevel of  pEffectsTarget  to newBlendLevel
   if newBlendLevel <100 then
  add someIncrement to newBlendLevel
  send "object_FadeOut pEffectsTarget, 
newBlendLevel,someIncrement,someDelay" to me in someDelay milliseconds
   else
  set the blendLevel of pEffectsTarget to 100
   end if
end object_FadeOut

Disclaimer: I am not a real programmer and I don't really know what I am doing, 
but because it is LiveCode I can still reach my goals ☺

BR




 

On 1/25/17, 5:44 AM, "use-livecode on behalf of Bob Sneidar via use-livecode" 
 wrote:

I want to create a variable called tTargetField so I use

put "Hello" into tTargetField

Now I want to put something into a field so I create a variable called 
tTargetField2 and
put the long id of field "foo" into tTargetfield2

Then to put something into the field I
put "Hello" into tTargetField2

Now do you see? All I am doing here is replacing the value of tTargetField2 
that used to contain the long id of field "foo" with the word "Hello". How does 
the engine know what you mean? It's Ambiguous.

Bob S



___
use-livecode mailing list
use-livecode@lists.runrev.com
Please visit this url to subscribe, unsubscribe and manage your subscription 
preferences:
http://lists.runrev.com/mailman/listinfo/use-livecode

Search and replace Raw unicode html-hex-dec characters

[Sannyasin Brahmanathaswami via use-livecode]

I have *.html files that were extracted from epub documents.

At our instruction, the company preparing these epubs was requested to use 
Unicode throughout and avoid all ANSI chars (i.e. us the Unicode for mdash and 
not — from the mac keyboard)

If I drop these on any browser they are perfectly rendered.

(see on line example here:

http://www.himalayanacademy.com/media/books/living-with-siva/web/part1_06.html)



When I open these as UTF-8 unicode files in BBEdit.  I see the characters like 
this:

The knowledge of how to realize this oneness and not create 
unwanted experiences along the way. The peerless path is 
following the way of our spiritual forefathers, discovering the mystical 
meaning of the scriptures.

To make life interesting, some files were exported with the decimal form and 
other with the hex form () some of the files mix the hex style notation 
along side the decimal.. in the same paragraph block… browser doesn't seem to 
care about that, so we are gettingand then later in the same 
paragraph: "


which is also exactly what you see if you look at source in the browser for the 
page.

I need to be able to target, for search and replace, a subset of these 
characters. In particular, we have troublesome

1) Pilcrow (Paragraph Sign)#old "run on" paragraph mark… which we 
want to turn into block paragraph form: just have an extra blank line.

2) discretionary hyphens coming through from InDesign all the way into the HTML 
files which are ignore in browsers, but appear as dashes/hyphens in Livecode 
fields after textDecode…   or  sometimes appear after I run scripts as: 
""


3) the old ligature double characters, in particule fi   "fi" tied together. 
 In Livecode, this just disappeard… instead "we find" I see "we nd" in 
the field.  I suspect the character is actually there, but the font can't 
render it,so the LC field is just white space in that location.

fulllment


Try as I might, I am unable to target these characters viaLC script…it's as if 
what I see in BBEdit and in the source of the page in the browser is not what 
gets imported using "put url " … in any form… in LC.  using

put url ("file:/" & "part1_ch3.html") into tText ## no search and replace I 
attempt will work
using binfile or textDecode  before insertion into the variable also doesn’t 
work.

Here is my script (latest iteration on many variants all of which don't get the 
job done)



global gBookFilesLocation
on mouseup
palette this stack
set the defaultstack to the topstack
put fld "TOC" into tToc
repeat for each line x in tToc
set the itemdel to "."
put item 1 of x & "-clean.txt" into tFilename
put the uBookFilesLocation of this stack into tPath
#put url ("file:" & tPath & "/ops/xhtml/" & x ) into tText
put url ("binfile:" & tPath & "/ops/xhtml/" & x ) into tText
put textDecode(tText,"UTF-8") into tNewText
# check for discretionary hyphens
if tNewText contains "" then
answer "Yes got unicode soft hyphen string" with "OK"
exit to top # this never fires ...
end if

## Attempts to replace here:
replace "¶" with (cr & cr) in tNewText  # never happens;may need to use 
""&""
replace "" with (cr) in tNewText # Pilcrow sign for the above char: 
U+00B6 old name "Paragraph Sign" --we want an extra blank line.
replace "" with "" in tNewText # soft hyphen
replace "" with "" in tNewText # Soft hyphen - replace with nothing
replace "" with " " in tNewText
replace "" with " " in tNewText
replace "" with " " in tNewText
replace " " with " " in tNewText # remove double spaces create by breaks 
following a space.

put textDecode(tNewText,"UTF-8") into tNewTextOut
set the htmlText of fld "CurrentChapterText" to tNewTextOut
put textEncode(fld "CurrentChapterText","UTF-8") into url ("binfile:" & 
gBookFilesLocation &"/_cleanExport1/" & tFilename)
end repeat
end mouseup

I tried, in another text processing stack I have, just to go direct from file 
into script/vars and then out to file without going thru a field… that's no 
better.

Any clues on how to approach this?

Brahmanathaswami



___
use-livecode mailing list
use-livecode@lists.runrev.com
Please visit this url to subscribe, unsubscribe and manage your subscription 
preferences:
http://lists.runrev.com/mailman/listinfo/use-livecode

Storing and object reference in a var

[Sannyasin Brahmanathaswami via use-livecode]

I have scripts supplied by others, that seem to work where the long id of an 
object is stored in a variable and then you can refer to that object with a 
single var-name

But I can't get this to work

put the long id of fld "daily-lesson-text" into tTargetField
put "Hello" into tTargetField

what am I doing wrong?

BR
___
use-livecode mailing list
use-livecode@lists.runrev.com
Please visit this url to subscribe, unsubscribe and manage your subscription 
preferences:
http://lists.runrev.com/mailman/listinfo/use-livecode

Re: Storing and object reference in a var

[Sannyasin Brahmanathaswami via use-livecode]

Edit;

get a grip on creating custom visual effect that I can use "all over the place" 
by creating an external script/text stack


-
Exactly

I'm studying scripts by some "pros" (Jacque, David, Malte-animation engine)

in order to get a group on creating custom visual effect that I can use 
"all over the place" but creating an external script/text stack

I've been doing server side script for so many years, the whole UI coding 
is all new but a lot more fun!

so If you want "safe" generic handlers… and you don’t' want to worry about 
the name space collisions you need to use the long ID

to drive things like this

put the long id of  grp "lessonTextControl"  into pEffectsTarget
send "object_FadeOut pEffectsTarget,0, 2, 20" to me 
wait for 2 seconds with messages


#then in your library:

command object_FadeIn  pEffectsTarget, newBlendLevel, someIncrement, 
someDelay
   set the blendLevel of  pEffectsTarget  to newBlendLevel
   if newBlendLevel > 0 then
  subtract someIncrement from newBlendLevel
  send "object_FadeIn pEffectsTarget, 
newBlendLevel,someIncrement,someDelay" to me in someDelay milliseconds
   else
  set the blendLevel of pEffectsTarget to 0
   end if
end object_FadeIn

command object_FadeOut  pEffectsTarget, newBlendLevel, someIncrement, 
someDelay
   set the blendLevel of  pEffectsTarget  to newBlendLevel
   if newBlendLevel <100 then
  add someIncrement to newBlendLevel
  send "object_FadeOut pEffectsTarget, 
newBlendLevel,someIncrement,someDelay" to me in someDelay milliseconds
   else
  set the blendLevel of pEffectsTarget to 100
   end if
end object_FadeOut

Disclaimer: I am not a real programmer and I don't really know what I am 
doing, but because it is LiveCode I can still reach my goals ☺

___
use-livecode mailing list
use-livecode@lists.runrev.com
Please visit this url to subscribe, unsubscribe and manage your subscription 
preferences:
http://lists.runrev.com/mailman/listinfo/use-livecode

Delete element from array

[Sannyasin Brahmanathaswami via use-livecode]

how do you remove a key from an array?

___
use-livecode mailing list
use-livecode@lists.runrev.com
Please visit this url to subscribe, unsubscribe and manage your subscription 
preferences:
http://lists.runrev.com/mailman/listinfo/use-livecode

Re: Storing and object reference in a var

[Sannyasin Brahmanathaswami via use-livecode]

ha!  duh! I need to go body surfing for a day and get away from the code

… I'm "losing it" 

Put "hello" into tTargetField#  replace the long id in the var with "hello" 

sheesh…

this get set prop thing has a lot of power, one day I'll wrap my head around it.



 

On 1/24/17, 12:57 PM, "use-livecode on behalf of Monte Goulding via 
use-livecode"  wrote:

You are writing ambiguous code and the engine can’t read your mind ;-)

Try:

set the text of tTargetField to “Hello”

___
use-livecode mailing list
use-livecode@lists.runrev.com
Please visit this url to subscribe, unsubscribe and manage your subscription 
preferences:
http://lists.runrev.com/mailman/listinfo/use-livecode

Reveal in Find on Windows

[Sannyasin Brahmanathaswami via use-livecode]

for in house production tools I use an apples script or shell script to reveal 
a file in the finder

how do we do this on windows?

BR

___
use-livecode mailing list
use-livecode@lists.runrev.com
Please visit this url to subscribe, unsubscribe and manage your subscription 
preferences:
http://lists.runrev.com/mailman/listinfo/use-livecode

Sets props of multiple objects from msg box

[Sannyasin Brahmanathaswami via use-livecode]

select three fields

enter in msg box:

set the margins of the selectedObjects to 10,10,10,10

Expected: margins of all three fields should change

observed: only the first one registers a change in margins
the other two do not.

Is this a bug?

___
use-livecode mailing list
use-livecode@lists.runrev.com
Please visit this url to subscribe, unsubscribe and manage your subscription 
preferences:
http://lists.runrev.com/mailman/listinfo/use-livecode

Re: Reveal in Find on Windows

[Sannyasin Brahmanathaswami via use-livecode]

HH wrote:
Yes, you are right, the handler only opens the enclosing folder
(or any path item of such a folder path).

Actually that's my main use case…

But now we have all options… thanks to all.

nice to find that I can use launch for both Windows and Mac

But has to have a slash on the end for Windows?

@Mike, right that was a typo sRootFolder is what it should be.  (script local 
for future use in another session)

BR

___
use-livecode mailing list
use-livecode@lists.runrev.com
Please visit this url to subscribe, unsubscribe and manage your subscription 
preferences:
http://lists.runrev.com/mailman/listinfo/use-livecode

Re: [ANN] Release 9.0.0 DP-5

[Sannyasin Brahmanathaswami via use-livecode]

@ Bob (who wrote)

   " Do you suggest that all development and improvement for version 8 and that 
we all wait for certain bugs to be addressed in the next major release? That 
would be unfortunate. Already since the earlier versions of 8, there have been 
marked improvements and bug fixes, such as much faster search times and an 
order of magnitude faster Build times."

Not at all. But for those of us who are

a) making a major move to script only (stacks, libs, behaviors) 
b) struggling with de-bugging on mobile

9 DP5 has us, as Mike says "drooling"

@Richard who wrote: 

"for deployment -- most recent stable…"
"for dev -- most recent build, know current state of engine as pertains to your 
work."

How logical!  thanks… assuming there is no change in file format and stacks are 
backward compatible, that makes good sense.

I have backups everywhere, so I'm off with 9DP5

Congrats to the team at HQ

BR

___
use-livecode mailing list
use-livecode@lists.runrev.com
Please visit this url to subscribe, unsubscribe and manage your subscription 
preferences:
http://lists.runrev.com/mailman/listinfo/use-livecode

Re: [ANN] Release 9.0.0 DP-5

[Sannyasin Brahmanathaswami via use-livecode]

I'm behind here in the overview:

Why even continue in 8.*   ?  

What issues are there with 9.* that would make it advisable not to try it for 
production?

BR


On 2/23/17, 10:41 AM, "use-livecode on behalf of panagiotis merakos via 
use-livecode"  wrote:

Dear list members,

We are pleased to announce the release of LiveCode 9.0.0 DP-5.


Developer Preview Release
=
Warning: this is not a stable release.  Please ensure that you back up your
stacks before testing them.

Getting the Release
===
You can get the release at https://downloads.livecode.com/livecode/ or via
the automatic updater.

Release Contents

LiveCode 9.0.0 DP-5 contains exciting new features, bug fixes, stability
and security improvements.

In total, 21 bugs have been fixed since the last development preview
release (LiveCode 9.0.0 DP-4).

Moreover, 36 additional bugs that have been fixed for 8.1.4 RC-1 (which is
not released yet), are included in LiveCode 9.0.0 DP-5.

LiveCode 9.0.0 DP-5 contains, among others, the following exciting new
features:

- Implementation of NFC tag reading on Android (Community, Indy and
Business):
New functions/commands/messages to read data from NFC tags.

- Implementation of a script library for common MIME functions (Community,
Indy and Business):
A new Multipurpose Internet Mail Extensions (MIME) library has been
implemented to provide support for common data formats such as multipart
email message bodies.

- Implementation of a remote debugger (Business only):
Run a standalone for the local platform from the IDE and debug it whilst it
runs locally
Run a standalone for iOS or Android from the IDE and debug it whilst it
runs on a directly connected Android or iOS device

- Implementation of a script profiler (Business only):
Use the new script profiler to identify the most costly lines of code in a
stack.


The full release notes are available from:

http://downloads.livecode.com/livecode/9_0_0/LiveCodeNotes-9_0_0_dp_5.pdf

Feedback

Please report any bugs encountered on our BugZilla at
http://quality.livecode.com/

We have a forum available for discussing LiveCode Builder at
http://forums.livecode.com/viewforum.php?f=93

Have fun!

The LiveCode Team
--
___
use-livecode mailing list
use-livecode@lists.runrev.com
Please visit this url to subscribe, unsubscribe and manage your 
subscription preferences:
http://lists.runrev.com/mailman/listinfo/use-livecode


___
use-livecode mailing list
use-livecode@lists.runrev.com
Please visit this url to subscribe, unsubscribe and manage your subscription 
preferences:
http://lists.runrev.com/mailman/listinfo/use-livecode

Parsing XML Nodes w/Same Tag

[Sannyasin Brahmanathaswami via use-livecode]

My longest running in house production app is an audio transcriber. Very 
successful little gadget, running in xTalk since 2001

We have over 1,000 XML files from an audio archive of transcripts.

Now I'm digging in and getting the data out.

I'm not facile with xml routines but did my best with the help of Bernd new, 
actually useable, dictionary.

But ran into a bug  in 9 DP5  (I think… ) OR I am doing something wrong

given transcripts formatted with nodes like this:




  CAS0886_radio-pilot_Inspired-Talks.A.mp3
  1980-01-03
  Gurudeva
  Three Words of Existence
  God and Lords of Dharma
  18 min, 36 secs
  San Francisco
  Brahmanathaswami
  
  Subtopic: three worlds: 0:3:56
  Subtopic: temple: 0:4:7
  

  
   
 [Radio Announcer: Ravi Peruman introduces Gurudeva]
   
   
 Gurudeva says ..
   
   
 More content here
   
   
 Subtopic: three worlds: 0:3:56
   
   
 All about temple
   
   
 Subtopic: temple: 0:4:7
   
  


My script looks like this

put revXMLChildContents(pTree, "/audio_transcript/header",tab,return,false,4) 
into fld "productionNotes"  # this works… I get all the contents
put revXMLNodeContents(pTree,"/audio_transcript/transcript_text/p") into tText 
# this works but we only get the first  content

# so I presume (like I said… parsing xml is new to me) we need to loop/iterate 
over the sibling  tags..
put revXMLNumberOfChildren(pTree,"/audio_transcript/transcript_text/","p",4) # 
return "6

# the following line should provide us what we need, I think, to set up a 
repeat loop  using the indexed node function
# and this is a) according to the dictionary b) and the script will compile:

put revXMLChildNames(pTree,"/audio_transcript/transcript_text/", 
return,"p",true)

I get a "green" OK in the script editor, but when I run it. we get this output, 
which is expected

p[1]
p[2]
p[3]
p[4]
p[5]
p[6]

and presumably I can use that list to now fetch the contents of all those nodes 
(haven't figured that out yet)

but the engine fires an error msg (even though the script compiled without 
complaining)  when we run it..

button "Load Transcript": execution error at line 22 (Handler: can't find 
handler) near "", char 89

it is breaking on the end of this line

put revXMLChildNames(pTree,"/audio_transcript/transcript_text/", 
return,"p",true)


even though the script compiles… isn't this a bug? If it a) is what the 
dictionary says it should be and b) compiles, why the error?

if not, what am I doing wrong?

The full button script is below… and you see my "fumbling" to fetch the content 
of all the "p" nodes. There seems to be some oddity relating to multiples nodes 
all having the same tag. 

global theTape
on mouseUp
put theTape into tTranscript
set the itemdel to "."
put "xml" into item -1 of tTranscript
if there is a file tTranscript then
put url ("file:/" & tTranscript) into tTranscriptXML
else
answer "Sorry, there is no transcript in the same folder as the audio" with "OK"
exit to top
end if
put revXMLCreateTree(tTranscriptXML,false, true,true) into pTree
if pTree is not an integer then
answer "Problem with the XML. Open in a text editor" with "OK"
end if
put revXMLChildContents(pTree, "/audio_transcript/header",tab,return,false,4) 
into fld "productionNotes"
put revXMLNodeContents(pTree,"/audio_transcript/transcript_text/p") into tText
put revXMLNumberOfChildren(pTree,"/audio_transcript/transcript_text/","p",4)
put revXMLChildNames(pTree,"/audio_transcript/transcript_text/", 
return,"p",true) 

#this script complies, but breaks on the above line when run
--put revXMLNextSibling(pTree,"/audio_transcript/transcript_text/p") into 
nextSibling
--put revXMLNodeContents(pTree,nextSibling) after tText # feeble attempt fails, 
need to do some loop but don't know how.
# no robust examples to follow, any help appreciated!
--put revXMLNodeContents(pTree, "audio_transcript/header/duration") into 
tTranscriptHTML # works for single node (of course)
--set the htmltext of fld "transcript" of stack "Audio_transcriber" to 
tTranscriptHTML

end mouseUp




___
use-livecode mailing list
use-livecode@lists.runrev.com
Please visit this url to subscribe, unsubscribe and manage your subscription 
preferences:
http://lists.runrev.com/mailman/listinfo/use-livecode

Re: Script Editor fixable? (was: Configuring a Sublime Text project to notify LiveCode IDE about updates to script only stacks)

[Sannyasin Brahmanathaswami via use-livecode]

going a bit OT here… I've dabbled withj Atom, Visual Studio and also long time 
ago Sublime… any reason you prefer Sublime?

I still keep dropping out back to BBEdit.. but if I could find a nother editor 
that had comparable tools I would use it. So far, atom and Visual Studio keep 
falling short…


 

On 2/22/17, 5:57 AM, "use-livecode on behalf of Trevor DeVore via use-livecode" 
 wrote:

They don’t have enough resources to create an editor that
matches a product like Sublime Text. Sublime Text is an app dedicated
entirely to editing. There are thousands of people contributing tools to
Sublime Text.

___
use-livecode mailing list
use-livecode@lists.runrev.com
Please visit this url to subscribe, unsubscribe and manage your subscription 
preferences:
http://lists.runrev.com/mailman/listinfo/use-livecode

Re: Reveal in Find on Windows

[Sannyasin Brahmanathaswami via use-livecode]

Thanks Roger!  

 

On 2/22/17, 1:54 AM, "use-livecode on behalf of Roger Eller via use-livecode" 
 wrote:

get shell (”explorer.exe /select,c:\windows\system32\selected_file.txt")

Source:

http://stackoverflow.com/questions/29970897/cmd-file-tasks-open-directory-and-select-file

~Roger



___
use-livecode mailing list
use-livecode@lists.runrev.com
Please visit this url to subscribe, unsubscribe and manage your subscription 
preferences:
http://lists.runrev.com/mailman/listinfo/use-livecode

Re: Reveal in Find on Windows

[Sannyasin Brahmanathaswami via use-livecode]

OK  I studied the doc here:  

https://support.microsoft.com/en-us/help/130510/command-line-switches-for-windows-explorer

and actually the use cases are more often to open a folder

but I don't have windows to test

does this work to open a folder window on Win32?

get shell (”explorer.exe /n,/root,c:\windows\system32\my-project-folder")


BR


 

On 2/22/17, 1:54 AM, "use-livecode on behalf of Roger Eller via use-livecode" 
 wrote:

get shell (”explorer.exe /select,c:\windows\system32\selected_file.txt")

Source:

http://stackoverflow.com/questions/29970897/cmd-file-tasks-open-directory-and-select-file

~Roger

___
use-livecode mailing list
use-livecode@lists.runrev.com
Please visit this url to subscribe, unsubscribe and manage your subscription 
preferences:
http://lists.runrev.com/mailman/listinfo/use-livecode

Re: Reveal in Find on Windows

[Sannyasin Brahmanathaswami via use-livecode]

this is what I have so far…

command openWorkingFolderInFinder pFolderFile
set the itemdel to "/"
put pFolderFile into sRootFolder
delete item -1 of sRootFolder
switch platform()
case "MacOS"
put ("open " & quote & sRootFolder & quote) into tShell
break
case "Win32"
put ("explorer.exe /n/root," & tRootfolder) into tShell
break   
end switch
get shell(tShell)
end openWorkingFolderInFinder

Can anyone test if this opens a folder on Windows?

tks


On 2/22/17, 4:59 PM, "use-livecode on behalf of Sannyasin Brahmanathaswami via 
use-livecode" <use-livecode-boun...@lists.runrev.com on behalf of 
use-livecode@lists.runrev.com> wrote:

OK  I studied the doc here:  


https://support.microsoft.com/en-us/help/130510/command-line-switches-for-windows-explorer

and actually the use cases are more often to open a folder

but I don't have windows to test

does this work to open a folder window on Win32?

get shell (”explorer.exe /n,/root,c:\windows\system32\my-project-folder")


BR



On 2/22/17, 1:54 AM, "use-livecode on behalf of Roger Eller via 
use-livecode" <use-livecode-boun...@lists.runrev.com on behalf of 
use-livecode@lists.runrev.com> wrote:

get shell (”explorer.exe /select,c:\windows\system32\selected_file.txt")

Source:

http://stackoverflow.com/questions/29970897/cmd-file-tasks-open-directory-and-select-file

~Roger



___
use-livecode mailing list
use-livecode@lists.runrev.com
Please visit this url to subscribe, unsubscribe and manage your subscription 
preferences:
http://lists.runrev.com/mailman/listinfo/use-livecode

Deleting stacks on the fly

[Sannyasin Brahmanathaswami via use-livecode]

I'm trying to optimize for minimal RAM usage on mobile. The app is modular with 
many stacks that are opened from the mainstack.

I had this before in a global library:

command portal_GoStack stackName
-- goes to the new stack
-- closes the current stack # But we want to delete it to clear RAM
-- safely ignores if we are alrady in the new stack
put the short name of this stack into oStackName
if oStackName = stackName then return oStackName
go to stack stackName
--close stack oStackName
-- test deleting the stacks to save on RAM
delete stack oStackName
return oStackName
end portal_GoStack

This is called from a button on a navigation bar of the stack we want to close 
and delete, as it opens a different stack.

This works if we just use "close stack oStackName"

but if I change this to "delete stack oStackName"

we get an error: stack "lib_SivaSivaPortal": execution error at line n/a 
(Object: stack locked, or object's script is executing) near "widget 
"go-home-portal" of group "homeScreensBottomNav" of card "images-quotes" of 
stack 
"/Users/Brahmanathaswami/Documents/_Siva-Siva-App/modules/gems/gems.livecode""

Obviously this means that the stack "gem.livecode" which has the button which 
is trying to "navigate" back to the home/main stack, cannot be deleted because 
the button script that calls the library handler is considered to be executing.

Closing it first before deleting does not help.

so I tried using "send in 10 milliseconds"  And this *does* work.. I'm just 
trying to see if this is best practice. both stacks will occupy RAM briefly, 
but  there is no other way that I know of to avoid this without a "flash" of 
some empty screen.

command portal_GoStack stackName
-- goes to the new stack
-- closes the current stack
-- safely ignores if we are alrady in the new stack
put the short name of this stack into oStackName
if oStackName = stackName then return oStackName
go to stack stackName
send "removeStack oStackName" to me in 10 milliseconds
return oStackName
end portal_GoStack

command removeStack stackName
delete stack stackName
end removeStack


Stack A opens
Stack B opens on top
Stack B want to close itself, be deleted (removed from memory) and open stack A

Anyone has gone through this loop already and discovered best practices to 
accomplish this?

BR

___
use-livecode mailing list
use-livecode@lists.runrev.com
Please visit this url to subscribe, unsubscribe and manage your subscription 
preferences:
http://lists.runrev.com/mailman/listinfo/use-livecode

Re: Deleting stacks on the fly

[Sannyasin Brahmanathaswami via use-livecode]

FWIW, tested on an android device with plenty of memory

Deleting stacks isn't helping. My Nexus 5 has 1 GB free RAM even with all of 
Google's widgets and my app open and going from Stack B back to stack A then 
deleting StackcB in 10 milliseconds  is "nothing" in terms of memory 
requirement, but crashes in Android every time… 

I have a report in to HQ…

 

On 2/11/17, 8:38 AM, "use-livecode on behalf of Sannyasin Brahmanathaswami via 
use-livecode" <use-livecode-boun...@lists.runrev.com on behalf of 
use-livecode@lists.runrev.com> wrote:

I'm trying to optimize for minimal RAM usage on mobile. The app is modular 
with many stacks that are opened from the mainstack.

I had this before in a global library:

command portal_GoStack stackName
-- goes to the new stack
-- closes the current stack # But we want to delete it to clear RAM
-- safely ignores if we are alrady in the new stack
put the short name of this stack into oStackName
if oStackName = stackName then return oStackName
go to stack stackName
--close stack oStackName
-- test deleting the stacks to save on RAM
delete stack oStackName
return oStackName
end portal_GoStack

This is called from a button on a navigation bar of the stack we want to 
close and delete, as it opens a different stack.

This works if we just use "close stack oStackName"

but if I change this to "delete stack oStackName"

we get an error: stack "lib_SivaSivaPortal": execution error at line n/a 
(Object: stack locked, or object's script is executing) near "widget 
"go-home-portal" of group "homeScreensBottomNav" of card "images-quotes" of 
stack 
"/Users/Brahmanathaswami/Documents/_Siva-Siva-App/modules/gems/gems.livecode""

Obviously this means that the stack "gem.livecode" which has the button 
which is trying to "navigate" back to the home/main stack, cannot be deleted 
because the button script that calls the library handler is considered to be 
executing.

Closing it first before deleting does not help.

so I tried using "send in 10 milliseconds"  And this *does* work.. I'm just 
trying to see if this is best practice. both stacks will occupy RAM briefly, 
but  there is no other way that I know of to avoid this without a "flash" of 
some empty screen.

command portal_GoStack stackName
-- goes to the new stack
-- closes the current stack
-- safely ignores if we are alrady in the new stack
put the short name of this stack into oStackName
if oStackName = stackName then return oStackName
go to stack stackName
send "removeStack oStackName" to me in 10 milliseconds
return oStackName
end portal_GoStack

command removeStack stackName
delete stack stackName
end removeStack


Stack A opens
Stack B opens on top
Stack B want to close itself, be deleted (removed from memory) and open 
stack A

Anyone has gone through this loop already and discovered best practices to 
accomplish this?

BR

___
use-livecode mailing list
use-livecode@lists.runrev.com
Please visit this url to subscribe, unsubscribe and manage your 
subscription preferences:
http://lists.runrev.com/mailman/listinfo/use-livecode


___
use-livecode mailing list
use-livecode@lists.runrev.com
Please visit this url to subscribe, unsubscribe and manage your subscription 
preferences:
http://lists.runrev.com/mailman/listinfo/use-livecode

"Deleting" Stacks in Memory - What About Behaviors?

[Sannyasin Brahmanathaswami via use-livecode]

When we delete a main stack, the main stack and all its substacks are removed 
from memory.

But if we delete a stack that has behaviors set from external 
*_behavior.livecodescript  stacks for controls in the "main" (parent?) stack, 
those behaviors are still in memory.

Does it make sense to file an enhancement request, to at least allow the dev to 
set a preference

__ Delete behavior stacks from memory when stack using them are deleted [YES/NO]

This is in line with trying to manage memory usage on small devices.

Yes, someone will no doubt respond "but they are so small why are you worried?"

But small android phone with limited RAM, really do need help keeping the heap 
as low as possible.  I am just looking for all possible means to clear ram, 
then set a "policy" in app to do all possible house keeping along the way: set 
images to empty, delete stacks not in use are the main two "tricks" we need to 
have that are obvious;

But that leaves libraries and behaviors that are not in use. So if we *could* 
clean them up… why not do it.  Some times 200K means the diff between running 
and crashing on these "weak" devices.

Obviouslyl if one is using Libraries with Start Using, the intent is probably 
that these are serving as globally accessible handlers. But this is not the 
case with behaviors attached to controls in a stack that may be deleted. So 
does it not make sense they are treated the same way as sub-stacks when their 
"parent" stack is deleted?

Before filing an enhancement request I want to check here with everyone. What 
do you think?

BR
___
use-livecode mailing list
use-livecode@lists.runrev.com
Please visit this url to subscribe, unsubscribe and manage your subscription 
preferences:
http://lists.runrev.com/mailman/listinfo/use-livecode

Re: Deleting stacks on the fly

[Sannyasin Brahmanathaswami via use-livecode]

@ Jacque: I don't think it is ram after all.

FYI: on my Nexus 5 I have 1.8 GB RAM and our new app + All of Googles "widgets" 
that come preloaded take up a mere .8 GB and I have a full GB of RAM free. 
after proceeding (as we discussed off list with our team) to delete stacks when 
we move from one to another, so that only 1 or 2 are ever open in memory… I am 
pretty certain I'm not have a memory issue on this Android Device: but it still 
crashes.

I took this to HQ… I guess you all know about plugging in your android phone 
and running ./adb logcat  and then pass that over to LC support.

If you are interested  you can view it here:

http://wiki.hindu.org/uploads/siva-siva-app_adb-log.txt  # 13MB file

if you search 

Activity Manager: Process com.himalayan

You will run into all the instances of 

ActivityManager: Process com.himalayanacademy.sivasiva (pid 25473) has died

(crash/exit)


There appear to be three different "flavors" of "exited due to signal (11)  in 
that log.  I won't detail them here as I don't grok the log well enough to talk 
about it. 

Hopefully HQ will figure it out… building from 8.1.3 (RC2)

does 9 dp4 change the file format? Safe?  I could make back ups of course.  and 
test… but need to keep moving on content dev… 

 I have no recipe or rather, we do, but it seems harmless.

we "Go" from one stack to another, then after 10 milliseconds, delete the first 
one

here is the handler where the app exits on signal 11, you can  see where we 
were fiddling with 

1) just closing it, 
2) opening stack B and deleting stack A
3) Opening stack B and send a "Delete" stack A in 10 milliseconds.

All three scenarios cause android to crash and we have *lots* of open RAM. 

command portal_GoStack stackName
   -- goes to the new stack
   -- closes the current stack
   -- safely ignores if we are alrady in the new stack
   
   put the short name of this stack into oStackName
   if oStackName = stackName then return oStackName
   
   go to stack stackName
  --close stack oStackName
   -- test deleting the stacks to save on RAM
  send "portal_RemoveStack oStackName" to me in 10 milliseconds  
--   delete stack oStackName
   return oStackName
end portal_GoStack

command portal_RemoveStack stackName
if stackName <> "Siva-Siva-App" then
delete stack stackName
end if
end portal_RemoveStack


# Works perfectly well on iOS








Svasti Astu, Be Well 
Brahmanathaswami

www.himalayanacademy.com

 

On 2/14/17, 6:48 AM, "use-livecode on behalf of J. Landman Gay via 
use-livecode"  wrote:

I frequently delete stacks when memory is an issue, which is the problem 
with Swami's stack suite. I also create almost all new stacks with 
destroystack set to true for the same reason.



___
use-livecode mailing list
use-livecode@lists.runrev.com
Please visit this url to subscribe, unsubscribe and manage your subscription 
preferences:
http://lists.runrev.com/mailman/listinfo/use-livecode

Re: Script Editor fixable? (was: Configuring a Sublime Text project to notify LiveCode IDE about updates to script only stacks)

[Sannyasin Brahmanathaswami via use-livecode]

Thanks for the thorough reply Trevor. 

I left Sublime Text when it was "2" years ago…thinking it was an EOL'd tool but 
here you say "Sublime Text 3"

Oh… sweet, my Sublime Text 2 license works!

OK I'll try it..since Atom and Visual Studio still don't quite gel for my work 
flow… 


On 2/23/17, 4:33 AM, "use-livecode on behalf of Trevor DeVore via use-livecode" 
 wrote:

Overall I feel like Sublime Text reduces friction while I’m working. I feel
like I can work much faster.

___
use-livecode mailing list
use-livecode@lists.runrev.com
Please visit this url to subscribe, unsubscribe and manage your subscription 
preferences:
http://lists.runrev.com/mailman/listinfo/use-livecode

Re: Dragging widgets

[Sannyasin Brahmanathaswami via use-livecode]

You are not wrong. 

SVG *is* an image and ideally, should behave like any graphic object.  

Also, for the sake of LC future: These are the kind of things that make newbies 
"run away" from the product.. because they expect things to work, but they 
don't, the frustration level can be severe an image is an image is an image, 
whether it is png, bitmap, jpg or svg, the fact that is is a "widget" is 
secondary to the use case.

Enter an enhancement request.

 

On 2/28/17, 2:41 AM, "use-livecode on behalf of Graham Samuel via use-livecode" 
 wrote:

I think that some people (like me and very possible Richmond) want to use 
SVG graphics as effortlessly-resized images. In this context, we don’t really 
care to think of them as widgets at all (I hope Richmond agrees!). What we want 
is a class of images which are implemented as SVGs but which function in the 
same way as other images with respect to mouse messages etc.

Does this make sense or have I entirely got the wrong end of the stick 
(wouldn’t surprise me)?



___
use-livecode mailing list
use-livecode@lists.runrev.com
Please visit this url to subscribe, unsubscribe and manage your subscription 
preferences:
http://lists.runrev.com/mailman/listinfo/use-livecode

Re: Script Editor fixable? (was: Configuring a Sublime Text project to notify LiveCode IDE about updates to script only stacks)

[Sannyasin Brahmanathaswami via use-livecode]

Atom is nice, true, but I had a license for ST2 from 2014 and it worked!  plus 
"multiple cursors" and related multi-line handling  is awesome, I not seeing 
that in 
Atom (maybe there is a package for that…)  but, keep us posted on a syntax 
package updates.


On 2/28/17, 4:18 AM, "use-livecode on behalf of Mike Kerner via use-livecode" 
 wrote:

ST3 is "annoy ware" - it will bug you to register and pay, but you don't
have to.
I've been working on the indent rules, and have discovered that Atom is
almost like a prettier version of ST3.  After I have this working in ST3,
I'm going to go back to Atom and fix them over there.  Even the python
script that Trevor wrote to connect ST3-saves to LC seems to be portable.
I haven't had a look at Notepad++ or BBE, yet, to see what can/has to be
done to make all of this work in those environments.



___
use-livecode mailing list
use-livecode@lists.runrev.com
Please visit this url to subscribe, unsubscribe and manage your subscription 
preferences:
http://lists.runrev.com/mailman/listinfo/use-livecode

Re: Dragging widgets

[Sannyasin Brahmanathaswami via use-livecode]

@ Bob:

Right: not meant to criticize, but hopefully constructive observation about a 
specific issue… that's how we evolve, right?

 "just sayin"  that to the eyes of the user  (as Quentin said) a widget is a 
visual object/control on screen, like any other, the expectation that they can 
enter

on mousedown; grab me; end mousedown

will be very high. 

So this is a "gotcha" for newbies -- and I don't mean "kids" but experienced 
graphic designers who work all day in a visual environment: (a target market 
that IMHO opinion could be huge were LC ramp up on the visual side)  and now, 
in the new IDE(s), like HYPE from Tumult (just installed yesterday.. amazing…)

… Solutions could be simple as adding to the dictionary a disclaimer "SVG 
objects are drawn arithmetically; as such they do not behave like a block of 
rasterized pixels. In order to drag an SVG widget you need to group it first 
and then add the grab me to the group."  Or like Mike said  "fix it" 

Some things are fundamental in the UI/UX and dragging objects *any* object is 
one of them, and has been from the day we mouse down on an icon on the desktop 
and dragged it into the trash.

Just my two rain drops from Kauai where a storm is moving in, could be as much 
as 10 inches in 10 hours tonite. Yikes!

On 2/28/17, 5:32 AM, "use-livecode on behalf of Bob Sneidar via use-livecode" 
 wrote:

oh contraire. Not all images are the same. Support for each kind of image 
has to be implemented. If the industry developed a new kind of image, LC would 
be faced with incorporating that format. 

Bob S



___
use-livecode mailing list
use-livecode@lists.runrev.com
Please visit this url to subscribe, unsubscribe and manage your subscription 
preferences:
http://lists.runrev.com/mailman/listinfo/use-livecode

Re: Dragging widgets

[Sannyasin Brahmanathaswami via use-livecode]

Yes, of course, Hype is nowhere near the scope of LC. my interest is just what 
you describe: an environment for  smooth easy-to-build animation and, if it can 
also be deployed on a web site without having to load the entire Emscriptem , 
so much the better. 

A full graphic mini-novel-story (6 minutes user driven) I would prefer to do it 
in LC, because the scope of the back end is so much bigger, but it's just too 
hard compared to something like HYPE to do a) easily b) look really good.  So 
if we could build it in HTML5, if we could only fully integrate the browser 
widget in the message path we would have our Dosai and Sambar too. Plus you get 
the bonus of being able to deploy on web without having to wait for the entire 
Emscripten engine to load.

and, back on point, being able to touch-drag any object is simply the default 
expectation for a newbie coming from that GUI/Front End/UI design space over to 
check out LC…

On 3/1/17, 11:07 AM, "use-livecode on behalf of Tore Nilsen via use-livecode" 
 wrote:

If animated content is what you want to build, then, yes. But it is nowhere 
near the functionality we expect and get from LC. From time to time I use Hype 
to make animated and interactive content for my students, but when I need to 
make some serious applications that rely on variables, arrays, writing to and 
reading from external files, manipulate files and folders on the host system, 
use controls like radio buttons or check boxes, use menus or editable fields, 
there is no way this can be done in Hype 3.0 without writing html, css and java 
script manually.

The kind of actions you are able to use are limited to mouse/touch actions. 
Personally I am no big fan of menu driven scripting. I prefer to write my own 
code and find that this is very often quicker than navigating meny different 
menus. I like the way they have managed to implement html/css/javascript in a 
seamless way though, and I had hoped that this was the kind of implementation 
we would get in LC.

Regards
Tore Nilsen

___
use-livecode mailing list
use-livecode@lists.runrev.com
Please visit this url to subscribe, unsubscribe and manage your subscription 
preferences:
http://lists.runrev.com/mailman/listinfo/use-livecode

CompositorType Android Performance

[Sannyasin Brahmanathaswami via use-livecode]

we are having some issues with performance of our new app on Android, moving 
from one card/stack to another or dynamically updated the interface on the same 
card…) without navigating to another card or stack) can take 3-5 seconds,

Turning on the accelerated rendering of a stack causees it to crash when 
navigating from one stack (B)  to another (A) , when we issue a command to 
dynamically update the interface of stack A on returning to that card. ( Bug 
confirmed with HQ but no resolution yet.)

So I went hunting and found an old thread on the forums, which also discusses 
Android performance related to compostorType settings.

There's been no more discussion since  Apr '16

http://forums.livecode.com/search.php?sid=192d835ccba0ebcb501a069f71a7a11c

I'm working on Mac and when I query the stack with msg box I get a 
compositorType setting of "coreGraphics" but I'm not setting that.

there is earlier discussion on the forum thread about explicitly setting 
compositor to openGL, tilesize and cache limit etc.

another post there sounds exactly like what we are experiencing

"I'm not sure if this is related - but on Android I was recently experiencing a 
big delay in re-drawing a new card (with associated animations) when switching 
between cards. After a lot of experimentation I was able to solve this by 
turning Accelerated Rendering OFF. If I really needed AR (it turns out that I 
don't in my current project) I'd experiment with turning it off on close card, 
then on again after open card."

but no resolution or robust recommendations after that…

Any updates on this sphere, learned best practices and experiences for Android?

App requirements are

1) dynamically redrawing sub-groups of a main group on click  of one subgroup 
(whole group is rebuilt-card redrawn)
2) moving from one stack  (b) back to another (a) where stack (a) has it's main 
scrolling group rebuilt.

This all works fine on iOS and also Nexus5, but we may be deployed on smaller 
android devices…so looking to optimize as best we can.

BR







___
use-livecode mailing list
use-livecode@lists.runrev.com
Please visit this url to subscribe, unsubscribe and manage your subscription 
preferences:
http://lists.runrev.com/mailman/listinfo/use-livecode

setProp does not fire break?

[Sannyasin Brahmanathaswami via use-livecode]

I'm trying to understand some code done for us that was not documented.

We are using a lot of text only *.livecodescripts  and working with a  frame 
work that is more of an MVC model where we try to keep our binary stacks as 
free of code as possible i.e. they serve as views and then we call behaviors 
and start using libs: This seems to be the direction that the whole LC frame 
work has moved toward. And it certainly does make for cool collaboration on Git 
with text only stacks being easily diff'ed and conflict resolved (Yes! I 
actually have learned to just GIT  just a bit)  But it comes with many caveats

Normally I can start to understand things pretty well by just setting break 
points here and there and studying the variables/values being instantiated all 
the way.. hitting cmd-I to step thru and that's my learning curve on this 
undocumented code.

We have this behavior… which I know for sure is being called on start up...

setprop portal_RowNames [headerName] rowNames
answer "Hello" with "OK"
break
replace comma with CR in rowNames

### more code here

end portal_RowNames

If I comment out this set prop handler… the expected views fail. ergo we know 
for sure that it is being called. because it is assigned the group of objects 
(the "view") on card 1 of the stack

But if I add the "answer" and "break" to the handler, these never fire. 
Everything works, apparently and the setProp is called and does it's job

Why not?

BR




___
use-livecode mailing list
use-livecode@lists.runrev.com
Please visit this url to subscribe, unsubscribe and manage your subscription 
preferences:
http://lists.runrev.com/mailman/listinfo/use-livecode

Re: Documentation on Dispatch

[Sannyasin Brahmanathaswami via use-livecode]

But I found this in a script from an associate:

setprop portal_RowNames [headerName] rowNames

and it compiles… the first param is a variable inside square braces.

Not sure it this is some LC magic or just this programmer's "style"
-
Bob S:  That can only be an element of an array. Without the preceding array 
name, it will not compile.

BR:[var]  # value inside square braces…

that I see in some scripts… what is that about? A custom thing? or is this 
also some jewel expression like @ that you will never learn about unless you 
fall onto it?

___
use-livecode mailing list
use-livecode@lists.runrev.com
Please visit this url to subscribe, unsubscribe and manage your subscription 
preferences:
http://lists.runrev.com/mailman/listinfo/use-livecode

Documentation on Dispatch

[Sannyasin Brahmanathaswami via use-livecode]

for the documentarians among us.

Sometimes you want to do a "remote" call to a function and get data back. At 
first this appear no possible

send | dispatch | do | call

BUT: FYI you can wrap a function in Dispatch and the value is returned in the 
result.

this is not documented.



___
use-livecode mailing list
use-livecode@lists.runrev.com
Please visit this url to subscribe, unsubscribe and manage your subscription 
preferences:
http://lists.runrev.com/mailman/listinfo/use-livecode

Re: Documentation on Dispatch

[Sannyasin Brahmanathaswami via use-livecode]

@ pass by reference

Wow! I'm only now learning about this? ….sheesh, this would have solved so many 
issues I faced in the past!

and now pass arrays too. Fantastic.

@ Richard: yes "wrap" was the wrong way to state this… "call a function with 
dispatch" is was I should have said

BR

PS there is also some convention for using   

[var]  # value inside square braces… 

 that I see in some scripts… what is that about? A custom thing? or is this 
also some jewel expression like @ that you will never learn about unless you 
fall onto it?






 

On 1/10/17, 1:54 PM, "use-livecode on behalf of Bleiler, Timothy via 
use-livecode"  wrote:

e variables passed as call by reference can also be arrays, 

___
use-livecode mailing list
use-livecode@lists.runrev.com
Please visit this url to subscribe, unsubscribe and manage your subscription 
preferences:
http://lists.runrev.com/mailman/listinfo/use-livecode

Buildling Standalone for Mac on 8.1.2 - Can't Copy Icons/Resources/Inclusions

[Sannyasin Brahmanathaswami via use-livecode]

We are facing an old bug in various stacks. Some are very old stacks that have 
moved forward over time from as far back as 5.*  with standalone builder (SAB) 
settings revised along the way until we are up to Sierra and LC 8.1.2 where 
attempts to make a standalone fail due to an old bug where the standalone 
builder cannot copy icons.

to make matters more challenging SAB's  final error message cannot be save for 
pasting here.

we get a dialog with errors looking like this

241.1405,1 revCopyIcons
also errors relation to revCopyResources,revStandaloneCopyIncludes, 
revSaveAsMacStandalone

etc.

I'm only assuming that it is a problem with icons. Deleting the internal SA 
custom prop from the stack and re-saving does not help.

Meanwhile some other simple stacks *do* build without issues.  to make things 
even more mysterious. We passed a copy from a Windows machine where the SA 
build was done for Mac… on a Surface machine… andpull this down to a Mac… 
Switching to 64 bit "experimental " worked earlier the morning, and then later 
in the day, failed.

There is *no* way to debug this… any ideas?

Svasti Astu, Be Well
Brahmanathaswami
www.himalayanacademy.com

___
use-livecode mailing list
use-livecode@lists.runrev.com
Please visit this url to subscribe, unsubscribe and manage your subscription 
preferences:
http://lists.runrev.com/mailman/listinfo/use-livecode

Re: Object Selection Handles

[Sannyasin Brahmanathaswami via use-livecode]

Ditto what Dan, Richard, Craig said

(letting off just a little steam here too)

Ali: "provide examples" 

 Adobe illustrator see:

http://wiki.hindu.org/uploads/square-small-selection-handles.jpg

Ali: "justification"

-- I've been ranting for 2 decades about how LC IDE needs to be optimized for 
21st century elegant GUI design. 

-- Graphic designers don't switch to console and start typing "set the rect of 
some object to X,X,X,X"   "never gonna happen"

-- Good design is precise design, a miniscule adjustment in the scale of an 
object, left-edge, can make the difference in production values between looking 
"hokey" vs looking "awesome". If you take that ability away from the user, you 
shoot your product in the foot.

-- Please, always optimize the IDE for quality design production values and not 
necessarily ease of use for your kid sister. Yes, your kid sister might find a 
large handle useful to just drag out an object in a use case that is as vague 
as "I just want it bigger." But that is not how "real" design is done. We need 
quality tools for the whole front end UI builder market… and each of us here 
switches hats (programmer/designer) 20 times a day… 

-- point: we don't mind if the selection handles are small, zoom is always an 
option; if the mouse can pick up them up easily in a "nearby" area, that's also 
OK, but they must be centered on the border-edge/corner… because if I'm working 
to move the edge, scale up-down an object, I will make the visual 
"subconscious" assumption that the center of the handle = the X,Y or X or Y 
position of [whatever]

Anything else is just "wrong"  

BR

 

On 3/27/17, 10:41 PM, "use-livecode on behalf of Ali Lloyd via use-livecode" 
 wrote:

Instead of declaring that this is how it should be, it's much better to
provide examples and justification. I just checked, and every application I
have on this computer does selection handles with the center of the handle
on the boundary of the object.

___
use-livecode mailing list
use-livecode@lists.runrev.com
Please visit this url to subscribe, unsubscribe and manage your subscription 
preferences:
http://lists.runrev.com/mailman/listinfo/use-livecode

Re: Extra Title Bar in All Stacks

[Sannyasin Brahmanathaswami via use-livecode]

Now, that's obscure

I was in "In Full screen Only"

I set to sys prefs → dock → tabs→ "Manually"

But second tab title bar still appears on resizeable LC windows. even after 
rebooting LC



 Paul Hibbert via use-livecode"  wrote:

Look at the fifth item down the window and you’ll see:

   Prefer tabs when opening documents: Always | In Full Screen Only | 
Manually

If you set this to “Always” you will see the Tab Bars appear in resizable 
LC windows.
If it’s set to "In Full Screen Only”, then the tabs will appear if you 
create a new stack while your current stack or window is open in full screen 
mode.
If it’s set to “Manually” then the Tab bars don’t normally appear.

These settings are obviously local to your machine, so I doubt it would 
(easily) be successful to try developing an app that relies on this, so to me 
it makes most sense to leave it set to “Manually”, but no doubt someone could 
take advantage of this if they are willing to put in the extra effort needed, 
but don’t take my word for that!

___
use-livecode mailing list
use-livecode@lists.runrev.com
Please visit this url to subscribe, unsubscribe and manage your subscription 
preferences:
http://lists.runrev.com/mailman/listinfo/use-livecode

Re: Extra Title Bar in All Stacks

[Sannyasin Brahmanathaswami via use-livecode]

OK… see this:

http://quality.livecode.com/show_bug.cgi?id=19276

the problem was, (not documented anywhere that I can find.) I must have hit 
cmd-shift-t inadvertently… I'm not seeing this in the Mac short cuts… so I 
guess it is a LC short cut?  

once you hit that inadvertently, then thereafter stacks open with two title 
bars. Hit this combo again and they go away, quit livecode and they are still 
gone. 

Mysterious. One wonders where the 'memory" of this setting resides, if not in 
the OS shortcuts and not in LC settings..

roblem solved for now.



 

On 3/26/17, 3:21 PM, "use-livecode on behalf of Sannyasin Brahmanathaswami via 
use-livecode" <use-livecode-boun...@lists.runrev.com on behalf of 
use-livecode@lists.runrev.com> wrote:

Now, that's obscure

I was in "In Full screen Only"

I set to sys prefs → dock → tabs→ "Manually"

But second tab title bar still appears on resizeable LC windows. even after 
rebooting LC



Paul Hibbert via use-livecode" <use-livecode-boun...@lists.runrev.com on 
behalf of use-livecode@lists.runrev.com> wrote:

Look at the fifth item down the window and you’ll see:

   Prefer tabs when opening documents: Always | In Full Screen Only | 
Manually

If you set this to “Always” you will see the Tab Bars appear in 
resizable LC windows.
If it’s set to "In Full Screen Only”, then the tabs will appear if you 
create a new stack while your current stack or window is open in full screen 
mode.
If it’s set to “Manually” then the Tab bars don’t normally appear.

These settings are obviously local to your machine, so I doubt it would 
(easily) be successful to try developing an app that relies on this, so to me 
it makes most sense to leave it set to “Manually”, but no doubt someone could 
take advantage of this if they are willing to put in the extra effort needed, 
but don’t take my word for that!



___
use-livecode mailing list
use-livecode@lists.runrev.com
Please visit this url to subscribe, unsubscribe and manage your subscription 
preferences:
http://lists.runrev.com/mailman/listinfo/use-livecode

Re: LiveCode's handling of Unicode glyphs being dependent on the underlying OS

[Sannyasin Brahmanathaswami via use-livecode]

Two mangos on these issues from Hawaii. As one of the users of Richmond's 
program: He and I have been 'at it" for nearly 8 year with his DevaWrite Pro, 
because the "state of the art" for rendering Sanskrit in the world of fonts is 
pretty abysmal with respect to some small but mission critical unicode glyphs 
that are simply pretty much unavailable via any standard keyboard input… if 
Richmond succeeds with this app, he will have a "dream machine" for a niche 
market that, though small, can't get the job done any other way using unicode 
easily now without going back to Type 1 fonts or setting hot lead type!  enough 
back story…

One anomaly that appears to be generated by LC 9dp5 running on Sierra 10.12.3: 
Code point U803 maps in the Unicode standard to the Extended Latin "H with dot 
underneath" character. 

for some bizarre reason, on my machine/system,  Livecode is mapping this 
character to Lucida (I think… possibly Helvetica.)

see: http://wiki.hindu.org/uploads/h-with-dot-lucida-helvetica.jpeg

Not on Richmond's machine, where it maps correctly to his DevaWriter.ttf font 
point correctly.

Livecode exports the RTF code for this character correctly as u803, but does 
not render it in the font assigned to the field.

I can make a LC stack letter sized, print to PDF save the PDF as HTML and I get 
the strange anomaly where Adobe sees the PDF output from Livecode and renders

css 

f1: {font-family:Devawriter]
f2 (font-family:Ludida}

then in the html we have this odd 

[h-with-dot-here[

But… if I set my preferences in Sublime text to default font  is Richmond's 
font: DevaWriter.tff

*then* the h with dot *is* rendered correctly in the Devawriter Font NOT in 
lucida

So this is an issue with the LC engine…

I may relate to other issues…but  nuisance here because I have no other way to 
print or render the text except via  LiveCode card. unless I opent the RTF 
output in Text Edit and manually change each "h-with-dot" back to Devawriter 
font

so why is LC not mapping U803 to the font assigned to the field? 

POINT: this does prove there is issues with rendering on different OS, since 
that character displays properly on Richmond's machine.
  

 

On 3/27/17, 2:39 AM, "use-livecode on behalf of Mark Waddingham via 
use-livecode"  wrote:

No - it is not being too clever. It is doing precisely what it should 
for the purposes of laying out text (and indeed what the CoreText engine 
on MacOS does - as that is what the engine uses). Pretty much everyone 
writing apps does not want to care about the (very complex) details of 
turning text into positioned glyphs, they just want a text string to be 
rendered how 'you would' expect, with regard to the codified rules which 
have been developed over a large number of years for typesetting 
language into a printed representation

___
use-livecode mailing list
use-livecode@lists.runrev.com
Please visit this url to subscribe, unsubscribe and manage your subscription 
preferences:
http://lists.runrev.com/mailman/listinfo/use-livecode

Stack References When Fired in Backscripts/Libraries

[Sannyasin Brahmanathaswami via use-livecode]

I was (happily) surprised today when  in a handler in a library that was put 
into use on init of the app

that a handler in

lib_initMyApp.livecodescript  # now in the msg hierarchy

on bakeBread pDaKin
   put the name of this stack
end bakeBread

Then in a top level binary stack name… e.g.

baking.livecode

if we call  "bakeBread"

we see

stack "baking.livecode"

in the msg box and not

lib_initMyApp.livecodescript

ergo "this stack" in a script only text stack in the message path is not seen 
as "itself"
but actually the target script from whence the command was issued.

So this means we don't have to parse for "the target"  yay!

Am I right? Is this dependable moving forward? Any caveats?

BR

___
use-livecode mailing list
use-livecode@lists.runrev.com
Please visit this url to subscribe, unsubscribe and manage your subscription 
preferences:
http://lists.runrev.com/mailman/listinfo/use-livecode

Re: Extra Title Bar in All Stacks

[Sannyasin Brahmanathaswami via use-livecode]

"System wide"  ?? I tried a dozen different apps, include Apple's own Pages, 
Numbers and Notes etc.

*none* of them respond to cmd-shift-T to make tabbed windows… the only 
"system-wide" is the Finder.

Adding an Apple-Mac Only Finder feature to LC  that OS X does not even 
implement this in it's other applications and where, all other apps make use of 
cmd-shift-T for some other function in that particular app, does it make sense 
to try to emulate the Mac Finder in LC Windows?  

If HQ is trying to decide if this is useful enough to spend time making it 
work; perhaps a better decision is:

disable and forget it, so that we are not "harassed" by double title bars, 
ever.  

Will any Mac User shed a tear here if this future enhancement is crossed off 
the list?

 BR

On 3/26/17, 4:13 PM, "use-livecode on behalf of Paul Hibbert via use-livecode" 
 wrote:

Ah! that’s the bit I forgot to mention, I think it is a system wide 
shortcut (it works in some other apps), I have seen it appear in the LC View 
menu, but not always for some odd reason. If you hit cmd-shift-t with the menu 
open it immediately selects Toolbar-Text, so I wonder if this is clashing 
somehow.

Paul

___
use-livecode mailing list
use-livecode@lists.runrev.com
Please visit this url to subscribe, unsubscribe and manage your subscription 
preferences:
http://lists.runrev.com/mailman/listinfo/use-livecode

Re: Extra Title Bar in All Stacks

[Sannyasin Brahmanathaswami via use-livecode]

Further testing:  This is only occurring on stacks with "resize" set to true.. 
if I uncheck that the second title bar disappears.  My gut says it has 
something to do with the "decorations" property of stacks being corrupted.


On 3/25/17, 3:20 PM, "use-livecode on behalf of Sannyasin Brahmanathaswami via 
use-livecode" <use-livecode-boun...@lists.runrev.com on behalf of 
use-livecode@lists.runrev.com> wrote:

Something bizarre started happening this afternoon

Mac OSx 10.12.3 Sierra
LC 9DP5

Stack are opening with a second title bar underneath the *real* title bar.

see http://wiki.hindu.org/uploads/extra-title-var.jpg

I deleted my settings, removed all plug-ins, deleted settings again, 
rebooted… stack still appear with a second title bar.

any clues to what is wrong?

BR



___
use-livecode mailing list
use-livecode@lists.runrev.com
Please visit this url to subscribe, unsubscribe and manage your subscription 
preferences:
http://lists.runrev.com/mailman/listinfo/use-livecode

Re: Extra Title Bar in All Stacks

[Sannyasin Brahmanathaswami via use-livecode]

confirmed.

Create new stack (leave it untitled) 

… I am getting a second "tab bar"  under the *real* title bar of the new stack

Create second new stack
  
Hmmm,  yes, fascinating, I can drag the title bar slowly onto the first stack 
and let go

Now I have
   Untitled 2   # top of window

untitled 1 *  | Untitled 2 *# two tabs

-- Nothing in the view menu relatedt to tabs.
-- turn off resizeable in untitled 2 and the stack "jumps" out to a new window.

OK, none of my stacks in this project need to be resizeable I can live with it, 
but 
a) why would it appear suddently and 
b) why doesn't it go away after delete settings?

Is there another place where the engine is writing prefs or settings that we 
don't know about?  It wasn't happening yesterday, in stacks that *were* 
resizeable. Something I did today triggered this new behavior.

BR

On 3/25/17, 3:45 PM, "use-livecode on behalf of Colin Holgate via use-livecode" 
<use-livecode-boun...@lists.runrev.com on behalf of 
use-livecode@lists.runrev.com> wrote:

It’s a standard Sierra feature that LiveCode doesn’t fully support yet. The 
View menu is supposed to have a Show Tab Bar menu item, which would hide the 
tab bar part of the window. The option is there sometimes, so if you’re lucky 
enough to see it you can deselect it.

If the tab bar is showing you can drag one stack’s tab bar to another 
stack’s tab bar to dock them together. There should also be a + button in the 
tab bar too, which in LiveCode I suppose would start a new stack as a tab.

Stacks that are docked together are effectively the same card size, so if 
LiveCode does do better support it should be possible to start a new stack as a 
tab, and at the card side of the existing stack, without you having to say what 
size stack you want.


> On Mar 25, 2017, at 6:20 PM, Sannyasin Brahmanathaswami via use-livecode 
<use-livecode@lists.runrev.com> wrote:
> 
> Something bizarre started happening this afternoon
> 
> Mac OSx 10.12.3 Sierra
> LC 9DP5
> 
> Stack are opening with a second title bar underneath the *real* title bar.
> 
> see http://wiki.hindu.org/uploads/extra-title-var.jpg
> 
> I deleted my settings, removed all plug-ins, deleted settings again, 
rebooted… stack still appear with a second title bar.
> 
> any clues to what is wrong?
> 
> BR
> 


___
use-livecode mailing list
use-livecode@lists.runrev.com
Please visit this url to subscribe, unsubscribe and manage your 
subscription preferences:
http://lists.runrev.com/mailman/listinfo/use-livecode

___
use-livecode mailing list
use-livecode@lists.runrev.com
Please visit this url to subscribe, unsubscribe and manage your subscription 
preferences:
http://lists.runrev.com/mailman/listinfo/use-livecode

Re: Determining if orientation locked on Android?

[Sannyasin Brahmanathaswami via use-livecode]

I'm not sure about how to test for lock orientation, but there are a number of 
related but reports, most recently

http://quality.livecode.com/show_bug.cgi?id=19465

Elanor has an interesting test stack there.

I suspect it is the same bug. in my case I want the "game" like screens/stacks 
to *always* go landscape, 

But we are in a modular environment, the main navigations screens are locked to 
portrait and rendered as such.

This is working on Android if you set the mobileSetAllowedOrientations to 
"portrait, portrait upsidedown" So you *could* take control yourself… 

But if you navigated from a portrait stack to a landscape stack using…

Setting the mobileSetAllowedOrientations to "landscape left,landscape right" 

The stack is rendered left to right very small, and only re-renders full screen 
when the user turns their phone.

Again, not exactly your issue/question, but smells like fish in the same barrel.

BR
 

On 3/25/17, 9:19 AM, "use-livecode on behalf of Richard Gaskin via 
use-livecode"  wrote:

On my Android phone I have orientation locked to Portrait, yet the LC 
app I'm building ignores this and rotates the with the device.

I see that I can handle orientations internally, but I don't see how I 
can determine if the user has locked their device to one specific 
orientation.

Any clues?

--

___
use-livecode mailing list
use-livecode@lists.runrev.com
Please visit this url to subscribe, unsubscribe and manage your subscription 
preferences:
http://lists.runrev.com/mailman/listinfo/use-livecode

Re: D3 LiveCode Chart Examples

[Sannyasin Brahmanathaswami via use-livecode]

Ditta what Trevor said.

 Our new app has both binary and livecodescript stacks.  In the beginning I was 
"sloppy" and named some of the script only stacks (libraries, models, 
behaviors) with extension only ".livecode"  in short order there is massive 
confusion

a)  about what can and cannot be opened in a text editor.
b) about what can be diff'ed with the git compare tools

and other "gotchas"  

"huh? How come that stack did not open? no window?  Maybe it's off screen?"

one in fact wonders why this was no enforce in the new model from the get go.

It gets worse if you try to engage collaboration and the new person on the team 
has not been there on the ground floor. "Hmmm, should we maybe document which 
of these are binary and which are livecodescript files"  

So not only for your own LC team's sanity, but for the clarity of interface 
between Livecode and the rest of the languages world, Ditto what Trevor said.

In fact, is there a case to be made to enforce this at the engine level?

BR
 

On 3/24/17, 1:00 PM, "use-livecode on behalf of Trevor DeVore via use-livecode" 
 wrote:

If everyone who posts files to GitHub differentiates binary stacks from
script only stacks using the extension then it will be helpful going
forward. As part of the submission process you have to send a search to
them that shows how often the language extension is used. A search that
doesn't show binary files would need to be submitted so the search would
filter by .livecodescript.

___
use-livecode mailing list
use-livecode@lists.runrev.com
Please visit this url to subscribe, unsubscribe and manage your subscription 
preferences:
http://lists.runrev.com/mailman/listinfo/use-livecode

Re: D3 LiveCode Chart Examples

[Sannyasin Brahmanathaswami via use-livecode]

Do you mean to say

"We do NOT even use pre-built UI cards any more"

??

Wow… that's amazing.

And you are not seeing any performance issues?

BR


 

On 3/25/17, 10:41 AM, "use-livecode on behalf of Todd Fabacher via 
use-livecode"  wrote:

Sure Trevor. Thank you for being the inspiration for us to move almost 100%
to script ONLY code. We will use .livecodescript. We do even use pre
built UI card anymore. Everything is all dynamically built now. We have not
moved to widgets just yet, but we hope soon.



___
use-livecode mailing list
use-livecode@lists.runrev.com
Please visit this url to subscribe, unsubscribe and manage your subscription 
preferences:
http://lists.runrev.com/mailman/listinfo/use-livecode

Extra Title Bar in All Stacks

[Sannyasin Brahmanathaswami via use-livecode]

Something bizarre started happening this afternoon

Mac OSx 10.12.3 Sierra
LC 9DP5

Stack are opening with a second title bar underneath the *real* title bar.

see http://wiki.hindu.org/uploads/extra-title-var.jpg

I deleted my settings, removed all plug-ins, deleted settings again, rebooted… 
stack still appear with a second title bar.

any clues to what is wrong?

BR


___
use-livecode mailing list
use-livecode@lists.runrev.com
Please visit this url to subscribe, unsubscribe and manage your subscription 
preferences:
http://lists.runrev.com/mailman/listinfo/use-livecode

App Standalone Splash Screens/Icon Deployment

[Sannyasin Brahmanathaswami via use-livecode]

I am wondering how other teams working collaboratively using Git  handle their

a) Splash screens
b) App Icon
c) Main Stack Binary



Scenario: you have Devs A, B, C, D

working on the same app. Very modular with lots of text only scripts. But the 
main stack used to build the standalone, call it MyApp.livecode  is a binary 
stack that keep changing everytime on of the team test builds… it contains 
stack files and a behavior that is a text only stack. otherwise almost no code. 
but, though it never *really* changes, Git sees it as modified.

add to this the additional issue that if we provide all the iOS splash screens 
and the Android app icon in the repository… we can put them in a folder 
adjacent to the app, but not actually include them in the package (since the 
standalone builder will do that automatically.)  But… how can you set the paths 
to the splash screens to a relative path.. with the goal in mind: that the 
binary stack never really changes. I don't know how exactly GIT "measures" the 
modification. if the app doesn't really change, even though it has to be saved 
each time you build a standalone.

So bottom line questions are two.

1) Is there some way to avoid having the main stack "change" even though it is 
being save… our solution now is to stash it in GIT and everyone pulls that same 
stack from our "nightly" branch which is the SOA for that binary (if a 
substantive change is made to the MyApp.livecode, you have to push it to 
nightly… otherwise you never push… but only build your tests and then stash 
MyApp.livecode)  is there a better practice?

2) if you pick the splash screens in standalone builder, I can't actually see 
the path, but we assume that they are absolute paths to the files in the local 
user's HD… as soon as the MyApp.livecode moves from dev A (who selected all the 
splash screens and icons for the app in the SA Builder) is there a way to set 
these to a path relative to the MyApp.livecode binary?

BR
___
use-livecode mailing list
use-livecode@lists.runrev.com
Please visit this url to subscribe, unsubscribe and manage your subscription 
preferences:
http://lists.runrev.com/mailman/listinfo/use-livecode

LC's Future in the Browser - Web Assembly?

[Sannyasin Brahmanathaswami via use-livecode]

try this in Firefox 52

https://s3.amazonaws.com/mozilla-games/ZenGarden/EpicZenGarden.html

" Mozilla celebrated with a demo video of the high-resolution graphics of Zen 
Garden, and while right now WebAssembly supports 
compilation from C and C++ (plus some preliminary support for 
Rust), "We expect that, 
as WebAssembly continues to evolve, you'll also be able to use it with 
programming languages often used for mobile apps, like Java, Swift, and C#."

Very interesting…

___
use-livecode mailing list
use-livecode@lists.runrev.com
Please visit this url to subscribe, unsubscribe and manage your subscription 
preferences:
http://lists.runrev.com/mailman/listinfo/use-livecode

Re: Problem with converting time

[Sannyasin Brahmanathaswami via use-livecode]

Hmm, this is odd if you are located in North Carolina…it means your server is 
off shore, out in the Atlantic, because it is running 3 hours ahead of you. 
but, unless HostM has now floatilla ISP data center ships, or secret bunkers on 
the sea floor, something is amiss.

Ditto everything Richard already said Thanks Richard! Your prolific penchant 
for mentoring/documentation is really appreciagted!  I saved all that to my 
reference here…

put 1489755600 into tVar
convert tVar into dateitems
   -- on the mac tVar contains 2017,3,17,9,0,0,6  # North Carolina
   -- on the server tVar contains 2017,3,17,12,0,0,6 # Huh!  three hour ahead ?

Am I daft? this seem impossible as this indicates 9 am in NC = 12 noon on the 
server.

I had issues a long time ago, because the ISP would spin up an instance of 
Linux using a disk image--  some generic OS image. They never bothered to reset 
the time zone for the OS…many ISP tech's  assume the user is savvy -- "Well, 
it's your server, set it up the way you like it.") and will want to do that 
themselves.   

But naïve users don’t give it a second thought and I got these weird results 
until I checked, like Richard suggested, the time and zone on the server itself 
and reset them. There was a time when I thought it would be cool to set the 
time zone of the OS in California to the same time zone as Hawaii (local) but 
later this turned about to be a bad strategy because, really, like Richard 
said, you want everything to be "the same moment" around the global, which of 
course means, the time in each location, for that moment will be different.   
Any other way…  "therein lies madness"  

I had to master the simple "art" of understanding GMT and local conversions.  
So if your host is in Texas, make sure the TZ/Time on the "box" is correct for 
its real, physical locationin Dallas or where ever… and then if you want the 
web page to "pretend" the server is in NC, you would add time to the output. 
(or subtract time from your inputs in your scripts.) 

interesting related resource : https://www.epochconverter.com/

https://www.epochconverter.com/clock

BR


 



___
use-livecode mailing list
use-livecode@lists.runrev.com
Please visit this url to subscribe, unsubscribe and manage your subscription 
preferences:
http://lists.runrev.com/mailman/listinfo/use-livecode

Re: iOS "Invalid Entitlements"

[Sannyasin Brahmanathaswami via use-livecode]

Thank you Panos; back on track here.

IMHO in future might lead with this solution on any query about entitlements. 
Assume user has enough experience to use cmd line until further notified.

but here's my harrowing tale, in case it is helpful…

I went in unusual circles yesterday: I had reverted my whole app to a commit 
made on GIT on the 17th of March, and moved back to using 8.1.3, because I 
thought our "nightly" branch that I was working on in 9dp5 was irrevocably 
corrupted..

that’s where (old commit in 8.1.3)  I was unable to do a build with the latest 
provisioning profile and SA settings applied to the main/binary stack.  (see my 
email to support, which is resolved now).  

using codesign and security cmd for the app that was built on that old commit 
with 8.1.3: did not show any problems with entitlements… still it would not 
load without the invalid entitlements error.

Aside: the stack files of my main stack were quite messed up from cherry 
picking this stack and moving it from branch to branch etc.  it listed files 
that did not exist in the tree; one script I added to the tree was a 
.livecodescript that did not have the proper declaration on line 1.. etc. This 
whole stack files business needs more thought/refinement for a collaborative 
GIT environment.  (anyone solve this? please post as separate thread on 
handling your stack files via GIT.)

So I spent half a day getting nowhere in this old commit "roll-back" branch, 
using 8.1.3 stable

OK I thought,"we have nothing to lose here, it is only getting worse. Let me go 
back to the nightly branch in 9dp5, because I know I was able to build iOS 
standalones with that mainstack in 9dp5…" (despite other issues) 

 So I  commit all my changes to the "roll-back" branch; check out nightly; do a 
thorough examination of my stack files and the working tree. clean up 
everything there. Save, go to SA settings, by now the provisioning profile had 
changed… so I selected the new provisioning profile (same one that was used for 
the build on the "roll-back" branch in 8.1.3)  and built the standalone… got 
the usual first session failure to find resource (can you PLEEEZE fix that?) 
ran it again, got a successful build. cross fingers, go to xCode to add to my 
phone and BOING!@ it works now.  Yay!

OK so I run codesign against that build; and also security cms against that 
same build/embedded.mobileprovision

now, here's the "kicker" :- the output is *exactly* the same as the build that 
was created in 8.1.3 that fails with the invalid entitlements error message!

 But, built now with 9dp5  it works. My "gut" is telling me that either the 
main stack/or the cRevStandalonesettings in the rollback commit is corrupt very 
deep internally OR  8.1.3  is doing something wrong when building the 
standalone.  Really just a hunch, could be totally wrong.

But the good news is: having your stack files not matching the working tree 
seems to have been a big problem… because now, I have 100% fully functional, no 
bugs app on my iPhone. Jai Ganesha! after a full week of spinning… we are back 
on track!

now, off to test on  Android…

Thank God for a GIT work flow!
--
 

On 4/8/17, 3:42 AM, "use-livecode on behalf of panagiotis merakos via 
use-livecode"  wrote:

Hmm ok. Then you could probably check which entitlements are included in
the standalone, and see if there is any that is not included in the
provisioning profile. To do that:

1. Use the codesign tool to check the entitlements on the .app bundle:
Type in a terminal:

codesign -d --entitlements :- /path/to/yourApp.app

2. Use the security tool to check the entitlements of the app's embedded
provisioning profile:
Type in a terminal:

security cms -D -i /path/to/yourApp.app/embedded.mobileprovision

- Then see if there is an entitlement that is present in (1) but not in (2).

Best,
Panos

___
use-livecode mailing list
use-livecode@lists.runrev.com
Please visit this url to subscribe, unsubscribe and manage your subscription 
preferences:
http://lists.runrev.com/mailman/listinfo/use-livecode

Android Build Woes on Mac

[Sannyasin Brahmanathaswami via use-livecode]

last week thinks were flow well in Mac OS X Sierra w LC 9DP5

I could plug in my Nexus 5 via USB, check for Android build in the SB settings 
and ti would appear in the menu as a test target and I could deploy very fast. 
in fact that was so much more convenient than iOS builds that I didn't even 
both to test on iOS since for the most part if I can get it working on Android, 
it always works on my iPhone 6+

But then a few days ago, nothing I can do in LC will make the phone appear in 
the menu. at someone' recommendation I cd's into the Android SDK folder and ran 
these commands

Brahmanathaswami-3:platform-tools Brahmanathaswami$ ./adb kill-server
* server not running *
Brahmanathaswami-3:platform-tools Brahmanathaswami$ ./adb start-server
* daemon not running. starting it now on port 5037 *
* daemon started successfully *
Brahmanathaswami-3:platform-tools Brahmanathaswami$ ./adb devices
List of devices attached
06d7a4fd439cf2c5  device

so apparently, my Mac and the Android SDK see the device. But Livecode doesn't

Think perhaps this is just a new LC IDE quirk I went ahead and tried to build 
to disk..

LC tries but fails to build the Android Standalone  saying

"There was an error while saving the
standalone applications
could not compile application class."

"OK"

Any clues what is wrong? I don't think I changed anything mission critical 
since it was working last week.I'm targetting 4.1 - Jelly Bean

___
use-livecode mailing list
use-livecode@lists.runrev.com
Please visit this url to subscribe, unsubscribe and manage your subscription 
preferences:
http://lists.runrev.com/mailman/listinfo/use-livecode

Re: Android Build Woes on Mac

[Sannyasin Brahmanathaswami via use-livecode]

"Oh duh…"  you were right… I had deleted my settings in an attempt to see if 
some corruption there was causing other issues. For some reason LC finds the 
xCode mobile without going back to the preferences, but not the Android SDK.  
Now working again, thanks

I have it installed in a non-standard location 

/Documents/app-development/android-sdk-macosx

is there a better location that LC will find by default if we delete settings? 
or does one always need to go back to preferences?

 
Panos:

Hi Brahmanathaswami,

This probably means that LC does not find the android SDK. Go to LC
Preferences -> Mobile Support and make sure the path to android SDK is set
and valid.

Best
Panos

---
 BR wrote:

last week things were flowing well in Mac OS X Sierra w LC 9DP5

I could plug in my Nexus 5 via USB, check for Android build in the SB
settings and ti would appear in the menu as a test target and I could
deploy very fast. 


___
use-livecode mailing list
use-livecode@lists.runrev.com
Please visit this url to subscribe, unsubscribe and manage your subscription 
preferences:
http://lists.runrev.com/mailman/listinfo/use-livecode

Animation External in the Inclusions

[Sannyasin Brahmanathaswami via use-livecode]

I can't find any documentation related  to the item referred to as

"Animation"   included in the list of inclusions

Dictionary entries commands like "revPlayAnimation" with association:  
"animation library" all say:

"No longer supported, and included for backwards compatibility only… etc"

so, is the Animation inclusion the same library referred to in the dictionary.

Animation builder disappeared as far back as 2.5…

BR (always interested in animation options)

___
use-livecode mailing list
use-livecode@lists.runrev.com
Please visit this url to subscribe, unsubscribe and manage your subscription 
preferences:
http://lists.runrev.com/mailman/listinfo/use-livecode

Re: LC crashing when trying to save a standalone

[Sannyasin Brahmanathaswami via use-livecode]

interesting I was having exactly this problem this week… slightly different 
scenario, but smells like the same fishy business.

I was working in our branch "nightly" in 9 dp 5  and did something stupid broke 
stuff… I will spare you the story.

I thought "hmm I had this working in 8.1.3"  so I check out a commit from March 
17th, as a new branch and went to work in 8.3.1

but my main stack refused to be accepted by xCode when I tried to put it on my 
phone…invalide entitlements. four days spinning wheels and testing.. changing 
my provisionig profile, checking entitlements from the cmd line… etc.  I could 
find nothing wrong. 

Anyway, while mucking around in the old commit I discover where I had gone 
wrong later… it was not a big thing, so I when back to my "future" branch 
"nightly" in 9dp5.. and now, guess what.. the mainMyApp.livecode stack… I could 
now build the SA and xCode was happy with it..

but, and this is the mysterious part: the entitlements of my app as output by 
codesign an security cms… were exactly the same as the stack from the earlier 
commit that I was deploy with the SA in 8.1.3.  

I could only assume that the myMainStack.livecode was corrupt some how, and 
yes, I went through the whole thing deleting my settings and even manually 
wiped cRevStandaloneSettings on the "bad stack" and rebuilt the SA settings 
from ground zero… nothing helped.  

On 4/9/17, 2:58 PM, "use-livecode on behalf of Jonathan Lynch via use-livecode" 
 wrote:

One more update - an earlier version of the app will save to a standalone.
So, definitely something wrong with the stack. Going to try cloning the 
stack or something like that.

Sent from my iPhone

___
use-livecode mailing list
use-livecode@lists.runrev.com
Please visit this url to subscribe, unsubscribe and manage your subscription 
preferences:
http://lists.runrev.com/mailman/listinfo/use-livecode

Re: LC crashing when trying to save a standalone

[Sannyasin Brahmanathaswami via use-livecode]

is the browser widget fix (or whatever change that fixes it)  currently in 
9dp5? or is that future branch now "behind" 8.1.4rc1?

BR


On 4/10/17, 9:32 AM, "use-livecode on behalf of Jonathan Lynch via 
use-livecode"  wrote:

As both Monte and Tom have suggested, moving to 8.1.4 rc 1 fixed the
problem for me. I am usually wary of using rc versions, but if you are
going to use a browser widget, it is probably a good idea.

___
use-livecode mailing list
use-livecode@lists.runrev.com
Please visit this url to subscribe, unsubscribe and manage your subscription 
preferences:
http://lists.runrev.com/mailman/listinfo/use-livecode

Re: LiveCode Android Demo

[Sannyasin Brahmanathaswami via use-livecode]

I want to see that also! not the apk, but the code

Our current app is "dying" on android.

specifically, a mainstack, opening other stacks or a stack with browser widget  
in it… something that is working perfectly on desktop and iOS is failing 
miserably on Android.

so if anyone has an android app that does this (dynamically opening and closing 
different stacks/windows)  working on android… I would love to see app code… 
or, a least the functions that drive the navigation.

If you are interested… contact me off list and I will give you the link to our 
public git repository and you can try our nightly yourself on android.

BR

On 4/10/17, 6:06 AM, "use-livecode on behalf of Alejandro Tejada via 
use-livecode"  wrote:

Hi All,

Where I could download an .APK that demonstrate (with user interaction)
all LiveCode features available for Android Platform?

For example, check this Interactive OpenCV Demo (for Win32 platforms)
http://www.tsdconseil.fr/log/opencv/demo/index-en.html
This interactive OpenCV demo allows the user to load their own images
to test the algorithms.

Thanks in advance!

Al

___
use-livecode mailing list
use-livecode@lists.runrev.com
Please visit this url to subscribe, unsubscribe and manage your subscription 
preferences:
http://lists.runrev.com/mailman/listinfo/use-livecode

ResumeStack on Mobile

[Sannyasin Brahmanathaswami via use-livecode]

Our new app framework is a "no brainer"

MainStack A  let’s call the " Home " stack… ala the old Hypercard days.

launches, it is the standalone build stack, with the stack fiels and an init 
script that initializes all the other libraries into the msg path.

it serves as an "index" to the app… it has clickable links on card 1 that

   a) poke a URL into a browser widget in a browser view stack; then opens that 
stack on top of Home which is never closed.

OR

   b) Opens other ModuleStack(s)B, C, D etc.  which are destroyed on close so 
that Home appears again in the window

OR

c) just rebuilds a new set of different links on card 1 of Home (i.e. User 
sees it as shifting between indexes of links)

script only stacks open in the background, models, libraries, behaviors are put 
into use or added to stack files to support functions running the binary stacks 
"above". This is all working fine on desktop and on iOS

But we are having major performance issues on Android.

If acceleratedRendering is set for viewBrowserWeb stack or any of the module 
stacks. the window the ModuleStack(s) B, C, D are not properly drawn

If interested you can follow:

bug 19576


Frequently the scrolling group on cd 1 of Home dies when we go back to Home

I wonder if anyone else is doing this kind of thing?

If so how? There is no resumeStack msg sent in mobile.   (sure could us that!)

Is anyone emulating a "resumeStack" on mobile and if so, how are you doing it?  
I expect we need to dispatch a msg to the Home every time we close "top" stacks 
to tell it to rebuild it's scrolling group and mobile scroller? But the timing 
is tricky… if you close stack ModuleStackC…

on closestack
  dispatch "setIndexScroller" to card 1 of stack Home
end close stack

is one option…

on closeStack
  send "setIndexScroller" to card 1 of stack Home in 500 milliseconds  # give 
time for this stack to get out of RAM before re-rendering
end closeStack


how do you do it?

BR


___
use-livecode mailing list
use-livecode@lists.runrev.com
Please visit this url to subscribe, unsubscribe and manage your subscription 
preferences:
http://lists.runrev.com/mailman/listinfo/use-livecode

Re: [Blog] Infinite LiveCode Progress

[Sannyasin Brahmanathaswami via use-livecode]

heather wrote:

" Folks, in case any of you missed it, LC9 dp6 has some great new 
functionality:"

Beware: I have been doing "great work" (by that I means squashing bugs and 
running tests) in 9dp5, but 9dp6 completely trashed the android standalone 
build (on the phone only… my repository scripts and binaries are fine, I am 
filing a report)

I switched back to 9dp5, right away… and ran the test again and it performed 
perfectly, nothing wrong with our scripts.

FYI: odd but true: I'm making better progress now but targeting my android 
phone… it's faster to build and deploy, I don't have to go to Xcode → Devices 
and add the stand alone…  and whenever I do deploy on iPhone, if it is working 
on Android it always works on iPhone so…. go figure.

BR

 



___
use-livecode mailing list
use-livecode@lists.runrev.com
Please visit this url to subscribe, unsubscribe and manage your subscription 
preferences:
http://lists.runrev.com/mailman/listinfo/use-livecode

Composite Set up - PreopenStack "everywhere"

[Sannyasin Brahmanathaswami via use-livecode]

We are using highly modular framework for our new app. 

Nothing like the IDE itself of course, but we are up to close to 50 
*.livecodescript  text only scripts divided into behaviors, libraries, models  
plus about 15 binarie.livecode stacks for modules and "views"…. and growing. 

I'm trying to optimize for small android devices and it seems the 
compositorType, tile size and cache limit are important. 

I have this in an init script that is a behavior on the main stack that opens 
when the app opens, card 1 has the init behavior that loads all the rest of the 
libraries and stack files

FYI the compositor was doing really, really strange things on my mac so I 
turned it off under development. Alsothe documentation and lesson on this is 
really lean, so if anyone has better, more detailed guidelines for best 
practices on this , please let us know:

I have this in the preopenstack handler

on preopenstack
   # some init routines, load and start using libraries
   # then this:

if the environment <> "development" then
 set the acceleratedRendering of this stack to "true"
 if the platform is "iphone" then
 set the compositorType of this stack to "opengl"
 else
 set the compositorType of this stack to "software"
 end if
 set the compositorTileSize of this stack to 32
 set the compositorCacheLimit of this stack to (16 * 1024 * 1024)
end if
 pass preopenstack
end preopenstack

Ok so this is working now and the rendering of the stack on my macbook pro 
doesn't go "kerpluey!" (strange rects, cropped group with black background 
other weird stuff, flickering graphic) and the app runs fine on the iphone and 
Android.

But apparently we need to have this set for every *[binary].livecode stack/view 
we launch, as the compositor setting is not persistent across a session. So the 
questions are two

1) The above is cut and paste from the online lesson on this subject. Please 
fire away with better tips on the best way to declare these settings. 
"software" obviously does not apply to all device besides iPhone, other wise it 
would not fail (intermittently) on my mac book pro.  The lesson says to 
multiply the rect of all your layers by 4… e.g. 10 layers: (10 x(400x400) x 4) 
what exactly does that mean.. ? what defines the rect of a given layer? the 
full geometric boundary around all the objects on that layer? or simply the 
rect of the card times the number of layers?

2) Given that this preopenstack handler is in a behavior on the card for the 
launching stack, it is not global. What is the best thing to do if we want the 
compositor to be set every time any other stack is opened? 

We have one lib_SivaSiva.livecodescript that we start using that serves as our 
app wide generic handlers "backscript". I have set preopenstack handles in 
backscripts before and sometimes it gets you into trouble if you start doing 
stuff that really should not be globally implemented.  But if I include this 
into that library

on preopenstack
 if the environment <> "development" then
 set the acceleratedRendering of this stack to "true"
 If the platform is "iphone" then
 set the compositorType of this stack to "opengl"
 else
 set the compositorType of this stack to "software"
 end if
 set the compositorTileSize of this stack to 32
 set the compositorCacheLimit of this stack to (16 * 1024 * 1024)
 end if
pass preopenstack
end preopenstack

Now it will fire every time any stack is opened. Is there some penalty for this 
method? My gut tells me that perhaps having these preopenstack handlers "lying 
around in the message path" is just not best practice. But maybe that just a 
"superstitious fear" 

Or maybe we should turn this into a discrete handler

command setUpCompositor
   # set the acceleratedRendering and compositor settings here

end setupCompositor

and then, call this discretely in a preopenstack handler in each separate 
stack. 
That somehow makes me feel better, not sure why.  Possibly because almost all 
the binary.lc stack in the framework *do* have preopenstack handlers already….

What do you recommend for 1) and 2)?

BR


___
use-livecode mailing list
use-livecode@lists.runrev.com
Please visit this url to subscribe, unsubscribe and manage your subscription 
preferences:
http://lists.runrev.com/mailman/listinfo/use-livecode

Distinguishing between tap and swipe

[Sannyasin Brahmanathaswami via use-livecode]

Mobile: To distinquish tap from swipe, I'm using the abs of the difference of 
the mouseloc on mousedown and mouseup….

private function _mouseMoved
put abs (the mouseV - item 2 of sStartLoc) into someDistance
put someDistance >= 10 into someBoolean
return someBoolean
end _mouseMoved

But this is quite "sluggish" on the any phone, meanwhile any straight forward

on mouseup
do something
end mouseup

is quite snappy.

Is there a better way to trap for swipe and immediately trigger a tap.

BR


___
use-livecode mailing list
use-livecode@lists.runrev.com
Please visit this url to subscribe, unsubscribe and manage your subscription 
preferences:
http://lists.runrev.com/mailman/listinfo/use-livecode

Re: Remote debugger

[Sannyasin Brahmanathaswami via use-livecode]

I'm very late to this party.

Where is the documentation?

I set the inclusion, run test target my android → nexus 5  and the dialog opens 
"Do you want to start remote debugging" I say yes… click through my app on the 
phone until finally it crashes and quits. Dialog in the IDE opens "Remote 
debugging session ended"

OK, I did not set any break points… but there do we find information? any log 
file.

From the thread here I'm not seeing the original announcement/link to 
documention… sounds marvelous!
 

On 2/24/17, 9:48 AM, "use-livecode on behalf of J. Landman Gay via 
use-livecode"  wrote:

I'd suggest that documentation make it clear what permissions and 
inclusions need to be set up. I'm not sure I'd make the permission 
automatic, since the final standalone may not want that permission set 
and the developer will probably forget to unset it, especially if they 
didn't know it was turned on in the first place.

___
use-livecode mailing list
use-livecode@lists.runrev.com
Please visit this url to subscribe, unsubscribe and manage your subscription 
preferences:
http://lists.runrev.com/mailman/listinfo/use-livecode