Has anybody had any experience using Palm's new WDBAS wireless application
server?
There's not much info available on it (other than marketing hype). I sent
an email asking for a price on it, but no response. I guess they're too
busy raking in the bux from all those sales? Whatever ... If they
Well, heck ... that was easy, wasn't it?
I don't remember clicking on the release version. Must have been before I
had my caffeine that day.
Thanks.
> It sounds like you're using the Release version of PalmSim. There is
> also a Debug version of DR11, available as a separate download (look
>
According to: http://www.palmos.com/dev/tech/palmos5/testcases.html , we're
supposed to "Open up the NTFullDbg_enUS.rom in the enUS subdirectory".
That doesn't exist in dr11. It existed in dr6 and dr8. In dr11, it's
"NTFull_enUS.rom".
I presume we use the one in the dr11 enUS subdirectory? Th
Keith: Sounds cool!
Are there any plans (or, better yet, does a feature already exist) to
implement a "record stuff during debug" feature that will allow playback to
recreate bugs? This could also be used to record/playback a self-running
demo of our apps. Assuming, of course, that we could pro
I ran into something similar many moons ago. If I recall correctly, it had
something to do with optimization of the scroll bar processing in Palm OS.
It only draws the arrowheads when it thinks it needs to. By drawing a
bitmap over the arrowhead, you confused the optimization routine.
To get ar
Welcome to the club of frustrated noFrame users ...
I faked it out. Create the button as a repeatingButton, which by definition
has no frame, then treat each repeat-tap as a single tap.
Works like a champ. Hokey, but ...
Regards,
Terry
--
For information on using the Palm Developer Foru
This is a basic "C" problem and has nothing to do with Palm OS, or with the
size of your application..
A 2 byte unsigned integer can hold a maximum value of 65535.
Obviously, you're trying to stuff 10 pounds of a number into a 5 pound sack.
Either change your definition from a 2 byte integer to
Interesting. I'm not sure which release fixed that PrvFixUpPtrs glitch.
How "corrupted" are your lists? Can you provide an example of before &
after?
--
For information on using the Palm Developer Forums, or to unsubscribe, please see
http://www.palmos.com/dev/tech/support/forums/
Depending on what version of Palm OS you're running, you could be tripping
on the infamous "PrvFixUpPtrs" bug. Any time a new object was added to a
form, the form grew and was possibly relocated. If Palm OS had to relocate
the form, it also "adjusted" pointers within the form structure that poin
Ooops ... if you were using gdb previously, then you might not have
CodeWarrior.
Sorry about that.
--
For information on using the Palm Developer Forums, or to unsubscribe, please see
http://www.palmos.com/dev/tech/support/forums/
I'm not sure how you're starting up the debugger. Here's how I do it ...
1) Open the project I want to debug in CW's IDE
2) click on CW's IDE Palm/LaunchEmulator (if necessary)
3) click on CW's IDE Project/Debug
That will automatically download the app to POSE, and stop at the entry
point.
Goo
Check "FldGetAttributes", then see if it's "editable".
--
For information on using the Palm Developer Forums, or to unsubscribe, please see
http://www.palmos.com/dev/tech/support/forums/
I'll presume you're a beginner at this.
Not sure which "menus" you're talking about that don't respond to the mouse.
The buttons on the device?
The "app windows" should be blank at startup, assuming the debugger stops at
the entry point.
Have you tried clicking on the Debugger's "Run" green arr
Oh, boy ... another member of our club ...
As of the current moment in time ... sadly, the answer is "no, there are no
glue accessors provided".
It appears that those of us who braved the new world of Palm OS in the early
days and figured out ways to provide UI functionality that wasn't there in
re: the table is a container for other UI elements
Right ... it's a "wrapper" around UI elements within its own little
"window", so you can treat them as an "object". Essentially, you're trying
to make a gadget do what a table already does. Except that a gadget
requires you to handle more of th
re: Could anyone help or advise me what to use for tapping and drawing the
gadget? I am quite new to using the gadget. I am suppose to write a
application that which is quite similar to the week schedule of the
datebook. Generally my screen will just display the table where the rows are
the times(
re: Why is it that whenever I insert the pendownEvent, I could not get a
respond from the either 4 of the buttons i.e when the selected button is
not highlighted?
case pendownEvent :
handled = true; break;
Because by setting "handled" to true, you're telling Palm OS that you
handled the penDown
re: When we load an app onto a device for debugging, we lock the database so
it won't move and affect the mapping of memory addresses to functions.
Makes sense ... and, I lock the chunk I need to modify.
re: We also have the DB open for writing, since breakpoints are implemented
by changing the
I posted the following message on 2/21/2002 (subject: PalmSim / 5.0 ??)
START REPOST
This code has been running perfectly fine on Palm OS 3.1 through 4.0 in POSE
3.4.
// Open the resource database for read/write
tmpDBH = DmOpenDatabase(cardNo, dbID, dmModeReadWrite);
if (tmpDBH == NULL
re: I'm still not following what it is you are actually trying to accomplish
... why do you need the information about which group a control belongs to?
If you want to know something about the group selection, use
FrmGetControlGroupSelection(). Why are you starting with the control and
working yo
John: re: FrmHideObject / FrmShowObject
Note that SDK 4.0 Upd 1 disallows access to internal record structure
descriptions, so you'll have to set up a controlled pre-4.0 Upd 1
environment..
The following is what I do in an isolated "MyAppFrmHideObject", which I call
in place of "FrmHideObject".
This code has been running perfectly fine on Palm OS 3.1 through 4.0 in POSE
3.4.
// Open the resource database for read/write
tmpDBH = DmOpenDatabase(cardNo, dbID, dmModeReadWrite);
if (tmpDBH == NULL)
{
err = DmGetLastErr();
AbandonUpdates();
return (err);
}
In PalmSim with Palm
Mark: Check out a msg I posted on this forum on 02/20/2002 @ 1:49pm,
subject = SDK 4.0 Upd1
I ran into a similar situation with "unusable", and documented how I intend
to get around it there.
Regards,
Terry
--
For information on using the Palm Developer Forums, or to unsubscribe, please se
FrmReturnToForm doesn't immediately exit to the previous form. All it does
is erase the current form, then set the active form to the previous one.
Your application continues to execute until you return to the Palm OS event
handler.
I wouldn't be surprised to find out that your code mistakenly e
I think you have more problems here than just handling non-alphabet
characters.
Are you sure your algorithm is correct?
According to the following statements:
> if ((badgeNum[i] >= 'A') && (badgeNum[i] <= 'Z'))
> badgeNumPhysLo |= 10 + badgeNum[i] -'A';
> else
> {
>
errBuff is defined as an array of 50 pointers to strings, not as an array of
50 strings.
"recText1=(char *)MemHandleLock(recHandle1);" retrieves the ptr to the
memory chunk of the record you retrieved.
"str1=recText1; " assigns the ptr to a temporary variable, which I assume is
defined as "Char *
re: Can you instead iterate over all the form objects, get their IDs, and
discover that way which items are in your form?
Doh !! You could at least have made the quick-fix solution a little
tougher, you know, so it wouldn't look like I was TOTALLY out to lunch.
[gg]
re: I'll look into this one
Ha! I finally figured out how to "legally" create a frameless button.
Well, at least for my client's specific need.
Create a repeatingPushButton (which, by definition has no frame), then treat
all repeat taps as a single tap.
I hate having to go through Kansas City to get to Boston from Miami .
Geoff: If you're just starting out, I can send you some code snippets that
will probably do what you want. If you're looking at getting VERY dynamic
and controlling everything from a set of control tables, I can't really send
snippets of what I developed here for client use because it's proprieta
re: The following partial code (edited for clarity) does not compile for me
under the 4.0r1 SDK, but was fine under 4.0:
I ran into a similar situation with code that I had to implement early on
(like 3.2 ??) that fixed a bug in 3.1's FrmHideObject logic. It dealt with
setting the useable attri
re: How about creating a button w/ a symbol character to trick constructor
into giving you a frameless button, then changing the character at runtime?
Great idea !!
But, the form is created completely dynamically (meaning, Constructor isn't
involved at all). The number, placement, and content o
According to the documentation, FrmGetObjectIndex ...
"Returns the index of the specified object (the index of the first object is
0), or frmInvalidObjectId if the supplied object ID is invalid."
Works great. Except that, in POSE (3.4, not sure about others), anytime the
Object corresponding to
Check out the series of posts on this thread from Monday, February 18, 2002
8:07 AM.
In short, you can't get there from here ... through internal control blocks.
Regards,
Terry
--
For information on using the Palm Developer Forums, or to unsubscribe, please see
http://www.palmos.com/dev/te
re: Is there a reason you can't make your application multi-segment?
Not really. I'm just trying to keep it as small as possible, and therefore
didn't want to add the overhead of the segment manager library routines.
After I posted my question, I determined what was causing the call to
"_lmul_"
Robert: Again, thanks for your response.
re: (2) a (hidden) gadget you have in your form (the resource cost is
minimal and the UInt32 you give it is actually hiding within the form's
information, hence is dynamic and associates only with the current form; but
if you are re-using the same form ma
Robert: Thanks for the reply.
The "store allocated chunks in globals" concept certainly has merit. The
trouble with "globals" is that they are non-reentrant.
I built a single generic form driver system (driven by totally user-defined
tables) that does everything dynamically, thus taking the vas
Thanks for the response ...
> Yes -- on Palm OS 5, the FormType pointer doesn't actually point to a
> useful structure. It instead is a handle to a shadow structure
> allocated by ARM code. It probably isn't a chunk pointer, which means
> that MemPtrSize won't work on it.
I suspected something
While executing the following statement in a "frmClose" routine:
tempInt = MemPtrSize(frmP);
I got the following error message:
The instruction at "0x00a4d22f" referenced memory at "0x01176858". The
memory could not be "read".
Is there something illegal about that in 5.0? It has worked
To stay within the 32K jump limits of Small Model applications, I
implemented a "Jump table" to all the static runtime library routines that
need to be accessed by application code. And, I positioned my library at
the end of my application Segments so more application code will be able to
fit wit
Thomas: I just spent a lovely weekend tearing up perfectly functional code
in a generic form driver to "fix" this frmPopupObj problem.
The reason for "needing" access to that "virtual" object was to dynamically
update the label of the PopTrigger from the value of a user controlled
dynamically po
Sounds like you're trying to access an internal table structure. That's a
no-no because your app might not run on future releases of Palm OS.
Check out "TblGetNumberofColumns" if you're on Palm OS 4.0 or above. Or,
check out TblGlue.h for a corresponding function that will support earlier
relea
Never mind ... I finally found the "Generate Link Map" option, which is
under the "Debugger Info" section in the Settings panel (why there?).
The link map shows that the final application size is 61KB.
So, it appears that the Project window displays an incorrect "code" size.
Maybe not incorrect
According to the Palm OS Programmer's API Reference in the PalmOSGlueLib
chapter ...
Note: PalmOsGlue is a link library, not a shared library. Linking with
this library increases your application's code size. The amount by which
your code size increases varies depending on the number of librar
Sorry, I worded that wrong. Constructor doesn't have an explicit way to
define a frameless button, but the text that accompanies the "Frame"
attribute implies that buttons can be built with no frames ... which is why
I started poking around way back when to figure out how to do that and
stumbled
There's a difference between fixing bugs and rewriting code. The
suggestion was to switch to a gadget. That's not a simple bug fix. A
gadget would require much more manipulation by my set of library routines.
Essentially what I built was an "object" (using OOP terminology) that uses
standard
re: I believe that FrmHideObject works correctly in Palm OS 3.1 for label
objects, which is why Poser is not letting you access the object directly
yourself.
Looks like you're right. I commented out that label attribute code, and the
labels still hide and show at the appropriate times on 3.1.
T
re: switch from buttons to gadgets.
Well, yeah. That's a way to do it. But, the existing code, which was
written way back when Palm OS was a pup, is working very smoothly ... other
than the fact that now I can't change the frame attribute without getting my
wrist slapped. And, it might not wor
> I think he means "4.0 SDK Update 1 DR 1" in the seeding area. DR2 was
> up for a while, but disappeared. This was supposed to go public last
> week, but it seems to be in the same web posting black hole that POSE
> 3.4 was in.
Right. My bad. The name of the download was sdk40upd1dr1.zip
Keith: Here's the error message I get ..
MyApp just wrote to location (0xWhatever) which is in the "attr" field of
the frmLabelObj .. blah blah blah
It occurs in this case:
case frmLabelObj:
obj.label->attr.usable = false;
break;
I haven't found any accessors or Glue for La
Keith: There's a blurb in the FrmHideObject function that says: "Prior to OS
version 3.2, this function did not set the usable bit of the object
attribute data to false. On an OS version prior to 3.2 you can work around
this bug by directly setting this bit to false yourself."
I cloned some code
Thanks for the response.
Re: Your application should not access those fields directly, but instead
should go through the glue code and let it do it for you.
Right. Understood and agree wholeheartedly. The bad news is that there
AREN"T any glue routines for accessing (view or update) the frame
re: This blurb from the "_news.txt" file ...
* Allow read access to "attr" fields of Control, List, Bitmap, Label,
and Scrollbar objects if the "accessor" function is not available --
FrmGlueGetObjectUsable accesses those fields. FrmGlueGetObjectUsable
also accesses Gadget and Table objects, b
Well, boo hiss ...
After submitting a bug report (twice!) via standard Palm methods
(http://www.palmos.com/dev/support/bugs/) about 3.3 POSE in combination
with the 4.1 SDK Beta, and not receiving a response either time, I see that
the new 3.4 POSE contains the same bug.
That being
"MyApp
> Have you updated to the 7.1 patch? This has some debugger changes.
I also could not get the Simulator to function with MW 7.0, and I had
already applied the 7.1 patch.
However, I had updated to 7.1 prior to having the Palm 4.0 SDK installed.
Since then, I have installed the 4.0 SDK.
Accordin
re: FrmAddSpaceForObject does work like a charm, and doesn't even seem all
that hackish
I sure hope it's not "hackish" ... the concept was cloned from the Palm 3.5
source where they add other dynamic UI objects. That's why it works so
well, and why it "shouldn't" be that big of a deal for Palm t
Sajul: re: The code you have given, since it accesses Palm OS internal
structures won't it fail on higher end devices like O.S. 5? Or is there any
new provision for the new OS?
Yes, the code I provided uses an "illegal access" and might not work on Palm
OS 5 and future releases. But, Dan mention
Dan: Welcome to the small, but exclusive, club whose members try to make use
of Palm's Dynamic UI functionality.
Here's the code snippet I built that works hunkey-dorey fine ... well,
except for the fact that it uses the "system use only" function called
"FrmAddSpaceForObject". You're welcome to
re: the doc writer(s) weren't thinking about scrollbars when they wrote the
text you quoted
In my neck of the professional Software Development world, the tech writers
take direction from the engineering department on which major subjects and
associated details need to go into the manuals, then u
According to the Palm OS Programmers' Companion documentation ...
**
Some applications, such as forms packages, must create their displays at
runtime-it is for applications such as these that the Dynamic UI API is
provided. If you're not absolutely sure that you need to change your UI
dynamic
re: When this comes up, usually at least one of the responses is: Please
demonstrate a sane UI that needs dynamic addition of scroll bars.
OK, wiseguy ... check out the prior message I sent on a completely
dynamically created User driven User Interface, posted to this newsgroup on
7/20/2001.
Wh
I posted a message on 1/11/2002, but just now realized the text got
truncated to 1,000 characters because the newsgroup server uses an
evaluation copy of the "Newsgroups Interface". A shortened version follows.
We can add forms and specific types of objects, but there doesn't seem to be
a way to
In prior communications here, I provided several accessor method
possibilities that were "accepted" by the Palm staff as "yeah, we can do
that" ... but I forgot asking about dynamically adding a ScrollBar.
POSE 3.3's excellent sensitivity to an application touching stuff it
shouldn't (internal re
> >4) I can't "clone" a button...
>
> I'm not sure we would add such a routine, if instead we had ways to get or
> set each of the supported attributes individually. I presume that a
cloning
> routine wouldn't be necessary in that scenario?
>
Yes. Access to the attributes would give me what I ne
You'd think that by starting at the Palm Developer's site and following the
links to the "recommended" text based resource editor that I would find a
reference to prc2pilrc. Most likely, that reference would be on the PilRC
page itself, because it seems like a very useful and collaborative utilit
Switching to PilRC was my initial knee-jerk reaction.
But, I didn't see anything in the PilRC doc that would let me salvage the
work (ie: convert it to text) I've already done in Constructor for 40+ forms
and a ton of UI objects. I didn't realize the plugin was a different (yet,
similar) tool.
I submitted this request a week or so ago, but as far as I know did not
receive any feedback from the Palm contingent.
I have added another request at the end. That additional request also
discussed in a few messages several weeks ago, with no comment from the Palm
folks.
** [Begin
Is there an easy way to share the resource files used in Constructor so
multiple people can be updating them while working on different forms?
Looks live Version Control System only locks it when checked out. How do we
merge our changes? Other than manually, of course, which is labor
intensive,
I commented several weeks ago on the enhanced Emulator UI Mgr Data Access
error messages which are welcomed restrictions, but are causing problems
because of code implemented prior to those restrictions. For the time
being, I have been forced to turn off the setting for UI Mgr Data Access.
Unfort
That's a whole bunch of processing to accomplish a relatively simple
function.
First of all, you can't write to a Resource with StrCopy. A Resource is a
PDB, so you need to use the "Dm..." functions.
Personally, I'd ...
1) Pre-scan to determine how long your resultant STL resource needs to be
I ran into that problem months ago. You might look in the archive for posts
on that particular glitch.
Use sclRepeatEvent instead.
Regards,
Terry
***
--
For information on using the Palm Developer Forums, or to unsubscribe, please see
http://www.palmos.com/dev/tech/support/forums/
> I must be missing something. If you have the address of the first
> string in the list (for the tAIS case), that gives you the address of
> the chunk. If you've got the address of the first indexed string in a
> tSTL resource, and you control the contents of the resource (e.g. the
> prefix is al
Keith:
> I think I probably "spoke" too quickly. I'm now thinking that
> setting a control's frame style is not supported for the standard
> controls. Buttons and PushButtons get standard frames, and others
> get no frame. I think that's the standard UI that CtlNewControl is
> trying to enforc
Aaron: Thanks for your response ...
> .. maybe a better solution to these
> types of problems is to actually have a function like:
>
> CtlGetAbstractAttributes()
>
BINGO !!
Palm defined th bits and values being stored/accessed in objects for a
reason. And, if they needed it to control
Keith: Thanks for the response.
> I think that the "official" way to do this is to create multiple
> versions of the button you need with the frame attributes you need,
> and then show only the one you want. Wasteful, awkward, yes, but I
> don't think that setting the frame attribute is supporte
Thanks for the response.
Please add the following to your list of needed glue routines.
ControlAttrType: I can't find a POSE "safe" way to set the frame attribute
for buttons.
ControlType: need to determine which group a control is in.
I guess the "simple" solution is to look at POSE's list of
> If you have a list of chunks, you could do a linear search by checking the
address and the size of each one. If (chunk addr <= ptr) and (chunk addr +
chunk size >= ptr) then the pointer is in that chunk.
Right .. but I don't HAVE the list of chunks. I had no reason to save that
list in the o
Palm OS is tightening down access to internal structures, as evidenced by
the increased error checking in POSE. That's a good thing, and I wish it
had been done earlier.
However, for this "isolated data" concept to work properly, we have to be
provided routines that perform the same functions as
I'd recommend putting an error check after your "DmGetResource".
Also ...
CharPtr unlock=(CharPtr)MemHandleLock(gStringArrayH);
looks like it should be
CharPtr unlock=(CharPtr)MemHandleLock(theListItems);
Or ...
theListItems = (Handle) SysFormPointerArrayToStrings( theRecPtr,
theListCount );
Right ... but I don't have the actual pointer. All I have is an address of
a location (actually a pointer to a string) somewhere within the chunk. If
there was a function call to get the pointer to the chunk by passing the
address of the string, I could use MemHandleRecoverPtr and everything wou
There doesn't appear to be a function that will tell me the Chunk that a
given address in memory belongs to.
I have a dynamic routine that has to deal with potentially hundreds of
addresses that I allocated with potentially dozens of MemHandleNews, then
segregated them according to an algorithm.
Is there such a thing as an API to access PDB's on Windows that parallels
the functionality on the HandHeld?
I'd like to copy a PDB in its entirety to Windows, then access it directly
from there. I'd also like to populate a PDB on Windows and then copy it to
the HandHeld.
The same requirements
Make sure your resultant string is null-terminated. On the Palm V, it might
have residual non-zero data there. You could do a MemSet for the entire
block first.
--
For information on using the Palm Developer Forums, or to unsubscribe, please see
http://www.palmos.com/dev/tech/support/forum
Thanks for responding. It's amazing how much better my text searches work
when I update the icon Properties on my desktop to point to the 4.0
documentation. Doh !!
This is, indeed, aggravating. Now I'll have to go back and change several
processing routines to ignore that stupid "index" entry.
I just converted to SDK4.0.
Now, when I link my application, I get the following message:
PalmRez Warning: Less than 16 categories in list: 'SAIt ID=7200' (for
miscellaneous strings use tSTL resource instead.)
Where does the doc say I can't (or, I shouldn't) use tAIS resources for
anything but
Thanks, Laurie. Very clever workaround
It was nice of them to optimize the LstScrollList logic so it wouldn't go
through the same logic as the LstDrawList logic. Your forcing it to scroll
more lines than are currently displayed was absolutely diabolical !
I think I'll use the "LstGetVisibleIte
Yes, I tried it on 4.0, and it no longer is a problem on that version. It's
also OK on 3.1 and 3.3.
Sorry about the snippet. I didn't think it would be a problem to post
something that small, because I *dare* anybody to make use of that code
without its surrounding content.
--
For informati
With the color support in Palm OS 3.5, lists with greater than a certain
number of entries are susceptible to an intermittent bug where the list
display ends up blank.
The bug results from the following snippet from OS 3.5's List.c at line 510,
where it's trying to determine whether there is any
Doh!
Ben Combee is right (as usual). After his comment that "there is an easy
way", I figured I must be missing something obvious. So, I went back and
played around a bit with my project structure. Sure enough, it works just
like it should.
I don't know where I (and, apparently you) got off o
Chris: I used a naming convention for my libary routines to avoid the
collision with the "native" application code routines. Typically, an
application routine called "UpdateDBField" would do some context sensitive
setup processing then call "xxUpdateDBField" where all the hard work is
done. (whe
I tap on the "Application" button to exit my application without going
through "normal" shutdown procedures. This works fine on all defined "real"
(ie: Constructor built) forms.
When I have a dynamically defined form popped up over a "real" one, tapping
on the "Application" button causes "FrmClo
90 matches
Mail list logo
