Re: [E-devel] Introducing: elm_tsuite - a pixel comparison test suite for EFL.

2011-07-27 Thread The Rasterman
On Tue, 12 Jul 2011 12:11:00 -0300 Gustavo Sverzut Barbieri
barbi...@profusion.mobi said:

 It would be useful to have some references in SVN and also a tool to
 diff these files, producing a hint on what's different other than pure
 cmp a b: files differ at byte X. This is done by WebKit, for
 instance

tom's elm_testsuite.sh is a REALLY simple hack :) its not going to do these
things. i have a more complete test suit. i actually have the start of a pretty
good test suite that has commands to create objects and manipulate them ans
then save the current screenshot for later comparison. it actually lets u
scripot what objects u will use, where to put them, how to modify them etc so
we can test all sorts of combinations of objects and transitions of state. it's
lurking locally. i sent tasn a copy a bit back.

 Another useful idea from WebKit is the virtual dump of the render
 tree. This walks the canvas, layers and objects (recursively) to
 produce a textual output, that is then able to be diff'ed. This is
 simpler to check and diff, also faster to run. Something like:

ensure can do this. in theory. though it's languishing. we could roll this
right into evas itself though.

 Evas #123 100x100:
Layer #456:
   Rectangle #1 @0,0 100x100 # clipper=NONE
 pass_events=TRUE repeat_events=FALSE visible=TRUE
   Image #2 @10,10 50x50 # clipper=#1 pass_events=TRUE
 repeat_events=FALSE visible=TRUE visible=TRUE border=2,2,2,2
 scale_hint=STATIC file=bla key=NONE
   Smart #3 @0,0 100x100 # clipper=NONE ...
  Rectangle #4 ...
  Text #5...
 
 The idea is to create virtual pointers (#1, #2...) so we can reference
 them and diff the results in different runs.
 
 
 -- 
 Gustavo Sverzut Barbieri
 http://profusion.mobi embedded systems
 --
 MSN: barbi...@gmail.com
 Skype: gsbarbieri
 Mobile: +55 (19) 9225-2202
 
 --
 All of the data generated in your IT infrastructure is seriously valuable.
 Why? It contains a definitive record of application performance, security 
 threats, fraudulent activity, and more. Splunk takes this data and makes 
 sense of it. IT sense. And common sense.
 http://p.sf.net/sfu/splunk-d2d-c2
 ___
 enlightenment-devel mailing list
 enlightenment-devel@lists.sourceforge.net
 https://lists.sourceforge.net/lists/listinfo/enlightenment-devel
 


-- 
- Codito, ergo sum - I code, therefore I am --
The Rasterman (Carsten Haitzler)ras...@rasterman.com


--
Got Input?   Slashdot Needs You.
Take our quick survey online.  Come on, we don't ask for help often.
Plus, you'll get a chance to win $100 to spend on ThinkGeek.
http://p.sf.net/sfu/slashdot-survey
___
enlightenment-devel mailing list
enlightenment-devel@lists.sourceforge.net
https://lists.sourceforge.net/lists/listinfo/enlightenment-devel


Re: [E-devel] Introducing: elm_tsuite - a pixel comparison test suite for EFL.

2011-07-27 Thread Tom Hacohen
On 27/07/11 14:02, Carsten Haitzler (The Rasterman) wrote:
 On Tue, 12 Jul 2011 12:11:00 -0300 Gustavo Sverzut Barbieri
 barbi...@profusion.mobi  said:

 It would be useful to have some references in SVN and also a tool to
 diff these files, producing a hint on what's different other than pure
 cmp a b: files differ at byte X. This is done by WebKit, for
 instance

 tom's elm_testsuite.sh is a REALLY simple hack :) its not going to do these
 things. i have a more complete test suit. i actually have the start of a 
 pretty
 good test suite that has commands to create objects and manipulate them ans
 then save the current screenshot for later comparison. it actually lets u
 scripot what objects u will use, where to put them, how to modify them etc so
 we can test all sorts of combinations of objects and transitions of state. 
 it's
 lurking locally. i sent tasn a copy a bit back.

I'm already using this elm_testsuite.sh full time with textblock. I have 
this tool I created (it's in svn) called tb_test that I just run through 
tsuite with various sample format .txts and to easily verify I don't 
break anything. This hack already saved me a lot of spanking and 
possibly hours of testing in the past couple of weeks.

As for a more complex test suite, Aharon is working on it, it's getting 
really cool. It allows you to modify everything through code, take 
screenshots whenever you want, record mouse/keyboard input and etc.
It will be really awesome. :) The advantage of this test suite in 
comparison to your awesome (really awesome) test suite, is that it's 
intended for higher level testing, i.e widgets and stuff that rely on 
ecore's main loop (yeah, it works nicely with a mainloop) and not just 
for canvas objects.

--
Tom.

--
Got Input?   Slashdot Needs You.
Take our quick survey online.  Come on, we don't ask for help often.
Plus, you'll get a chance to win $100 to spend on ThinkGeek.
http://p.sf.net/sfu/slashdot-survey
___
enlightenment-devel mailing list
enlightenment-devel@lists.sourceforge.net
https://lists.sourceforge.net/lists/listinfo/enlightenment-devel


Re: [E-devel] Introducing: elm_tsuite - a pixel comparison test suite for EFL.

2011-07-27 Thread The Rasterman
On Wed, 27 Jul 2011 14:20:42 +0300 Tom Hacohen
tom.haco...@partner.samsung.com said:

 On 27/07/11 14:02, Carsten Haitzler (The Rasterman) wrote:
  On Tue, 12 Jul 2011 12:11:00 -0300 Gustavo Sverzut Barbieri
  barbi...@profusion.mobi  said:
 
  It would be useful to have some references in SVN and also a tool to
  diff these files, producing a hint on what's different other than pure
  cmp a b: files differ at byte X. This is done by WebKit, for
  instance
 
  tom's elm_testsuite.sh is a REALLY simple hack :) its not going to do these
  things. i have a more complete test suit. i actually have the start of a
  pretty good test suite that has commands to create objects and manipulate
  them ans then save the current screenshot for later comparison. it
  actually lets u scripot what objects u will use, where to put them, how to
  modify them etc so we can test all sorts of combinations of objects and
  transitions of state. it's lurking locally. i sent tasn a copy a bit back.
 
 I'm already using this elm_testsuite.sh full time with textblock. I have 
 this tool I created (it's in svn) called tb_test that I just run through 
 tsuite with various sample format .txts and to easily verify I don't 
 break anything. This hack already saved me a lot of spanking and 
 possibly hours of testing in the past couple of weeks.
 
 As for a more complex test suite, Aharon is working on it, it's getting 
 really cool. It allows you to modify everything through code, take 
 screenshots whenever you want, record mouse/keyboard input and etc.
 It will be really awesome. :) The advantage of this test suite in 
 comparison to your awesome (really awesome) test suite, is that it's 
 intended for higher level testing, i.e widgets and stuff that rely on 
 ecore's main loop (yeah, it works nicely with a mainloop) and not just 
 for canvas objects.

hmm sounds like he's doing something definitely higher up the stack. what i
started with was something to test evas itself and test that all states of
objects work and render right and that changing states - eg a move, a resize,
enable.disable map, clip and unclip, and all sorts of permutations and
combinations of these keep working correctly. i guess you could do a similar
thing for elm and bind it into the mainloop - eg instead of an explicit save
shot/image now you have a shot in 0.2 sec or wait until loop idle for 0.5
sec, then shot so u can wait until anim is done etc. you'd still need lots of
command to create widgets and manipulate them of course.

-- 
- Codito, ergo sum - I code, therefore I am --
The Rasterman (Carsten Haitzler)ras...@rasterman.com


--
Got Input?   Slashdot Needs You.
Take our quick survey online.  Come on, we don't ask for help often.
Plus, you'll get a chance to win $100 to spend on ThinkGeek.
http://p.sf.net/sfu/slashdot-survey
___
enlightenment-devel mailing list
enlightenment-devel@lists.sourceforge.net
https://lists.sourceforge.net/lists/listinfo/enlightenment-devel


Re: [E-devel] Introducing: elm_tsuite - a pixel comparison test suite for EFL.

2011-07-13 Thread Cedric BAIL
On Tue, Jul 12, 2011 at 5:23 PM, Tom Hacohen
tom.haco...@partner.samsung.com wrote:
 On 12/07/11 18:11, Gustavo Sverzut Barbieri wrote:
 It would be useful to have some references in SVN and also a tool to
 diff these files, producing a hint on what's different other than pure
 cmp a b: files differ at byte X. This is done by WebKit, for
 instance
 Well, it uses image magick to mark the different areas and also provides
 old and new images. I think it's more than enough to find out
 exactly what's wrong.

 Another useful idea from WebKit is the virtual dump of the render
 tree. This walks the canvas, layers and objects (recursively) to
 produce a textual output, that is then able to be diff'ed. This is
 simpler to check and diff, also faster to run. Something like:

 Evas #123 100x100:
     Layer #456:
        Rectangle #1 @0,0 100x100 # clipper=NONE
 pass_events=TRUE repeat_events=FALSE visible=TRUE
        Image #2 @10,10 50x50 # clipper=#1 pass_events=TRUE
 repeat_events=FALSE visible=TRUE visible=TRUE border=2,2,2,2
 scale_hint=STATIC file=bla key=NONE
        Smart #3 @0,0 100x100 # clipper=NONE ...
           Rectangle #4 ...
           Text #5...

 The idea is to create virtual pointers (#1, #2...) so we can reference
 them and diff the results in different runs.


 That's a good idea, but a lot more annoying to implement than a simple
 script, and I'm not sure there's enough benefit to even bother.

 The test suite is aimed for developers that want to make sure they don't
 break something when they change it, and also for a testing server, and
 is intended to run along with all the other unit/regression/etc. tests.
 For the developers, saying there's a difference and marking it on a
 picture is usually more than enough, and for the testing server, this
 will just help us find issues and spot the problematic commits.

 But we will consider adding those features to the other testing tool,
 which will not be just a script launching binaries and will give us a
 lot more control.

Maybe using ensure to do this could help to ease the amount of work
required by this idea (that I really like to).
-- 
Cedric BAIL

--
AppSumo Presents a FREE Video for the SourceForge Community by Eric 
Ries, the creator of the Lean Startup Methodology on Lean Startup 
Secrets Revealed. This video shows you how to validate your ideas, 
optimize your ideas and identify your business strategy.
http://p.sf.net/sfu/appsumosfdev2dev
___
enlightenment-devel mailing list
enlightenment-devel@lists.sourceforge.net
https://lists.sourceforge.net/lists/listinfo/enlightenment-devel


Re: [E-devel] Introducing: elm_tsuite - a pixel comparison test suite for EFL.

2011-07-13 Thread Tom Hacohen
On 13/07/11 11:25, Cedric BAIL wrote:
 Maybe using ensure to do this could help to ease the amount of work
 required by this idea (that I really like to).

Yeah, that's for sure.

--
Tom.


--
AppSumo Presents a FREE Video for the SourceForge Community by Eric 
Ries, the creator of the Lean Startup Methodology on Lean Startup 
Secrets Revealed. This video shows you how to validate your ideas, 
optimize your ideas and identify your business strategy.
http://p.sf.net/sfu/appsumosfdev2dev
___
enlightenment-devel mailing list
enlightenment-devel@lists.sourceforge.net
https://lists.sourceforge.net/lists/listinfo/enlightenment-devel


Re: [E-devel] Introducing: elm_tsuite - a pixel comparison test suite for EFL.

2011-07-13 Thread ChunEon Park
Compared 111 with 2 diffs.
Is it normal?
Even there are no diff files. :p

Let's run together for the best moment!
 -Regards, Hermet-
 
-Original Message-
From: Tom Hacohenlt;tom.haco...@partner.samsung.comgt; 
To: Cedric BAILlt;cedric.b...@free.frgt;
Cc: Aharon Hillellt;a.hil...@partner.samsung.comgt;; 
enlightenment-devellt;enlightenment-devel@lists.sourceforge.netgt;
Sent: 11-07-13(수) 17:41:07
Subject: Re: [E-devel] Introducing: elm_tsuite - a pixel comparison test suite 
for EFL.On 13/07/11 11:25, Cedric BAIL wrote:
 Maybe using ensure to do this could help to ease the amount of work
 required by this idea (that I really like to).
Yeah, that's for sure.
--
Tom.
--
AppSumo Presents a FREE Video for the SourceForge Community by Eric 
Ries, the creator of the Lean Startup Methodology on Lean Startup 
Secrets Revealed. This video shows you how to validate your ideas, 
optimize your ideas and identify your business strategy.
http://p.sf.net/sfu/appsumosfdev2dev
___
enlightenment-devel mailing list
enlightenment-devel@lists.sourceforge.net
https://lists.sourceforge.net/lists/listinfo/enlightenment-devel
--
AppSumo Presents a FREE Video for the SourceForge Community by Eric 
Ries, the creator of the Lean Startup Methodology on Lean Startup 
Secrets Revealed. This video shows you how to validate your ideas, 
optimize your ideas and identify your business strategy.
http://p.sf.net/sfu/appsumosfdev2dev
___
enlightenment-devel mailing list
enlightenment-devel@lists.sourceforge.net
https://lists.sourceforge.net/lists/listinfo/enlightenment-devel


Re: [E-devel] Introducing: elm_tsuite - a pixel comparison test suite for EFL.

2011-07-13 Thread Tom Hacohen
On 13/07/11 12:10, ChunEon Park wrote:
 Compared 111 with 2 diffs.


 Is it normal?

 Even there are no diff files. :p

It printed out the diffs while testing.

ls *.comp* to see the files that were different.

I guess you get Calendar_2 and Focus_2 different.

Focus_2 is an animation, I already removed it from the tests, that's why 
it doesn't work.

As for Calendar_2, it's date based, so you have to run them pretty close 
to each other in order to make sure you get the same images.

Maybe we should add a non-date related calendar test...

--
Tom.

--
AppSumo Presents a FREE Video for the SourceForge Community by Eric 
Ries, the creator of the Lean Startup Methodology on Lean Startup 
Secrets Revealed. This video shows you how to validate your ideas, 
optimize your ideas and identify your business strategy.
http://p.sf.net/sfu/appsumosfdev2dev
___
enlightenment-devel mailing list
enlightenment-devel@lists.sourceforge.net
https://lists.sourceforge.net/lists/listinfo/enlightenment-devel


Re: [E-devel] Introducing: elm_tsuite - a pixel comparison test suite for EFL.

2011-07-12 Thread Gustavo Sverzut Barbieri
On Tue, Jul 12, 2011 at 10:11 AM, Tom Hacohen
tom.haco...@partner.samsung.com wrote:
 Dear all,

 I'm glad to introduce elm_tsuite, a pixel comparison test suite for EFL.
 Thanks should go to Aharon Hillel for writing this one.
 Commit: http://trac.enlightenment.org/e/changeset/61289

 All the info below is also available in the README.

 This test suite should be used to make sure we don't break rendering of
 widgets when we do changes. For example, min size calculation breakage
 of elm_button will likely cause a resize in the test and thus an error
 will be detected by the test suite.

 This test suite is by no means limited to Elementary widgets only, more
 info in the commit log + README.

 There's currently a sample elementary.txt file there that includes a lot
 of tests for elementary widgets, just run:
 ./elm_tsuite.sh elementary.txt o
 Do changes, and then:
 ./elm_tsuite.sh elementary.txt
 to do the comparison.

 I hope to add some textblock tests soon as well.

 This is a fast and easy test suite that I hope everyone will start using.

 Aharon will soon start working on another test suite, that will let you
 write tests that directly call the screenshot calls and will possibly
 let you record user interaction. This will hopefully let us do even
 better testing to elementary widgets (or any other EFL application).

 If you have any questions, please let Aharon/me know.

It would be useful to have some references in SVN and also a tool to
diff these files, producing a hint on what's different other than pure
cmp a b: files differ at byte X. This is done by WebKit, for
instance

Another useful idea from WebKit is the virtual dump of the render
tree. This walks the canvas, layers and objects (recursively) to
produce a textual output, that is then able to be diff'ed. This is
simpler to check and diff, also faster to run. Something like:

Evas #123 100x100:
   Layer #456:
  Rectangle #1 @0,0 100x100 # clipper=NONE
pass_events=TRUE repeat_events=FALSE visible=TRUE
  Image #2 @10,10 50x50 # clipper=#1 pass_events=TRUE
repeat_events=FALSE visible=TRUE visible=TRUE border=2,2,2,2
scale_hint=STATIC file=bla key=NONE
  Smart #3 @0,0 100x100 # clipper=NONE ...
 Rectangle #4 ...
 Text #5...

The idea is to create virtual pointers (#1, #2...) so we can reference
them and diff the results in different runs.


-- 
Gustavo Sverzut Barbieri
http://profusion.mobi embedded systems
--
MSN: barbi...@gmail.com
Skype: gsbarbieri
Mobile: +55 (19) 9225-2202

--
All of the data generated in your IT infrastructure is seriously valuable.
Why? It contains a definitive record of application performance, security 
threats, fraudulent activity, and more. Splunk takes this data and makes 
sense of it. IT sense. And common sense.
http://p.sf.net/sfu/splunk-d2d-c2
___
enlightenment-devel mailing list
enlightenment-devel@lists.sourceforge.net
https://lists.sourceforge.net/lists/listinfo/enlightenment-devel


Re: [E-devel] Introducing: elm_tsuite - a pixel comparison test suite for EFL.

2011-07-12 Thread Tom Hacohen
On 12/07/11 18:11, Gustavo Sverzut Barbieri wrote:
 It would be useful to have some references in SVN and also a tool to
 diff these files, producing a hint on what's different other than pure
 cmp a b: files differ at byte X. This is done by WebKit, for
 instance
Well, it uses image magick to mark the different areas and also provides 
old and new images. I think it's more than enough to find out 
exactly what's wrong.

 Another useful idea from WebKit is the virtual dump of the render
 tree. This walks the canvas, layers and objects (recursively) to
 produce a textual output, that is then able to be diff'ed. This is
 simpler to check and diff, also faster to run. Something like:

 Evas #123 100x100:
 Layer #456:
Rectangle #1 @0,0 100x100 # clipper=NONE
 pass_events=TRUE repeat_events=FALSE visible=TRUE
Image #2 @10,10 50x50 # clipper=#1 pass_events=TRUE
 repeat_events=FALSE visible=TRUE visible=TRUE border=2,2,2,2
 scale_hint=STATIC file=bla key=NONE
Smart #3 @0,0 100x100 # clipper=NONE ...
   Rectangle #4 ...
   Text #5...

 The idea is to create virtual pointers (#1, #2...) so we can reference
 them and diff the results in different runs.


That's a good idea, but a lot more annoying to implement than a simple 
script, and I'm not sure there's enough benefit to even bother.

The test suite is aimed for developers that want to make sure they don't 
break something when they change it, and also for a testing server, and 
is intended to run along with all the other unit/regression/etc. tests. 
For the developers, saying there's a difference and marking it on a 
picture is usually more than enough, and for the testing server, this 
will just help us find issues and spot the problematic commits.

But we will consider adding those features to the other testing tool, 
which will not be just a script launching binaries and will give us a 
lot more control.

--
Tom.

--
All of the data generated in your IT infrastructure is seriously valuable.
Why? It contains a definitive record of application performance, security 
threats, fraudulent activity, and more. Splunk takes this data and makes 
sense of it. IT sense. And common sense.
http://p.sf.net/sfu/splunk-d2d-c2
___
enlightenment-devel mailing list
enlightenment-devel@lists.sourceforge.net
https://lists.sourceforge.net/lists/listinfo/enlightenment-devel