date:20130416

all fixed


On Tue, Apr 16, 2013 at 1:27 AM, Carsten Haitzler ras...@rasterman.comwrote:

 bug #1:

 1. start terminal (xterm/terminology - any window actually)
 2. type cmd: sleep 5; exit
 3. alt+left click and drag to move window. keep moving it until window
 closes/exits
 4. release mouse.
 5. segfault happiness! :)

 (a quick look says to me action_border is still pointing to a border that
 has
 now been deleted - crashdump attached).

 bug #2:

 something has happened to the shadow in comp... it's now wrong. it has
 moved
 down from the title:


 http://www.enlightenment.org/ss/display.php?image=e-516c92a8841755.22448970.png

 vs old:


 http://www.enlightenment.org/ss/display.php?image=e-515c83bebe2212.05354344.jpg

 look carefully at shadow on left/right size of the window near the titlebar
 top. now the shadow begins much lower.. just below client start.

 bug #3.

 double-click a window to shade it. it USED to just become a nice simple
 titlebar... no longer. :)

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



 --
 Precog is a next-generation analytics platform capable of advanced
 analytics on semi-structured data. The platform includes APIs for building
 apps and a phenomenal toolset for data science. Developers can use
 our toolset for easy data analysis  visualization. Get a free account!
 http://www2.precog.com/precogplatform/slashdotnewsletter
 ___
 enlightenment-devel mailing list
 enlightenment-devel@lists.sourceforge.net
 https://lists.sourceforge.net/lists/listinfo/enlightenment-devel


--
Precog is a next-generation analytics platform capable of advanced
analytics on semi-structured data. The platform includes APIs for building
apps and a phenomenal toolset for data science. Developers can use
our toolset for easy data analysis  visualization. Get a free account!
http://www2.precog.com/precogplatform/slashdotnewsletter
___
enlightenment-devel mailing list
enlightenment-devel@lists.sourceforge.net
https://lists.sourceforge.net/lists/listinfo/enlightenment-devel

Re: [E-devel] new bugs. in e (border releated)

On Tue, Apr 16, 2013 at 1:37 AM, Carsten Haitzler ras...@rasterman.comwrote:

 On Tue, 16 Apr 2013 09:27:21 +0900 Carsten Haitzler (The Rasterman)
 ras...@rasterman.com said:

 bug #4:

 moving windows around no matter how small or large they are is now really
 jerky... or it is almost all of the time... terminology seems to be smooth
 all
 the time, but xterm seems to be 50/50 and anything else (sylpheed,
 chromium,
 gedit, kcachegrind...) is jerky as all hell. it seems to be a timing issue
 of
 some sort...


I figured this was just a client redraw issue. the next compositor rewrite
will solve it altogether, but I don't know that there's anything to be done
for now



 bug #5:

 resizing of a window produces swimming of content as it is stretched to
 fill
 the new border UNTIL the client can render an updated set of content (and
 then
 fix it up). the effect basically is that content squashes and expands back
 and
 forth during resizing as opposed to remaining unchanged in scaling. comp
 used
 to address this nicely by enforcing e_comp size == size of pixmap
 dimensions. :)


I asked you about this a few weeks ago and your response was that it was
just a side effect of client redraw delay. I guess one of us misunderstood
the other, and I have no idea what your explanation means.



  bug #1:
 
  1. start terminal (xterm/terminology - any window actually)
  2. type cmd: sleep 5; exit
  3. alt+left click and drag to move window. keep moving it until window
  closes/exits
  4. release mouse.
  5. segfault happiness! :)
 
  (a quick look says to me action_border is still pointing to a border
 that has
  now been deleted - crashdump attached).
 
  bug #2:
 
  something has happened to the shadow in comp... it's now wrong. it has
 moved
  down from the title:
 
 
 http://www.enlightenment.org/ss/display.php?image=e-516c92a8841755.22448970.png
 
  vs old:
 
 
 http://www.enlightenment.org/ss/display.php?image=e-515c83bebe2212.05354344.jpg
 
  look carefully at shadow on left/right size of the window near the
 titlebar
  top. now the shadow begins much lower.. just below client start.
 
  bug #3.
 
  double-click a window to shade it. it USED to just become a nice simple
  titlebar... no longer. :)
 
  --
  - Codito, ergo sum - I code, therefore I am --
  The Rasterman (Carsten Haitzler)ras...@rasterman.com
 


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



 --
 Precog is a next-generation analytics platform capable of advanced
 analytics on semi-structured data. The platform includes APIs for building
 apps and a phenomenal toolset for data science. Developers can use
 our toolset for easy data analysis  visualization. Get a free account!
 http://www2.precog.com/precogplatform/slashdotnewsletter
 ___
 enlightenment-devel mailing list
 enlightenment-devel@lists.sourceforge.net
 https://lists.sourceforge.net/lists/listinfo/enlightenment-devel

--
Precog is a next-generation analytics platform capable of advanced
analytics on semi-structured data. The platform includes APIs for building
apps and a phenomenal toolset for data science. Developers can use
our toolset for easy data analysis  visualization. Get a free account!
http://www2.precog.com/precogplatform/slashdotnewsletter
___
enlightenment-devel mailing list
enlightenment-devel@lists.sourceforge.net
https://lists.sourceforge.net/lists/listinfo/enlightenment-devel

Re: [E-devel] new bugs. in e (border releated)

This sounds like a localized config bug. Try changing your desk flip
transition.


On Tue, Apr 16, 2013 at 3:51 AM, Carsten Haitzler ras...@rasterman.comwrote:

 On Tue, 16 Apr 2013 11:46:04 +0900 Carsten Haitzler (The Rasterman)
 ras...@rasterman.com said:

 bug #7:

 very often mouse events just stop working (going to clients). mouse
 press/move/release/wheel etc. after flipping virtual desktops a few times
 or alt
 +left dragging a window about.

  On Tue, 16 Apr 2013 09:37:54 +0900 Carsten Haitzler (The Rasterman)
  ras...@rasterman.com said:
 
  bug #6:
 
  xdaliclock -transparent is nicely broken (shape mask is applied and
 offset by
  title height). this used to work. it's one of my comp shape test cases.
 :)
 
   On Tue, 16 Apr 2013 09:27:21 +0900 Carsten Haitzler (The Rasterman)
   ras...@rasterman.com said:
  
   bug #4:
  
   moving windows around no matter how small or large they are is now
 really
   jerky... or it is almost all of the time... terminology seems to be
 smooth
   all the time, but xterm seems to be 50/50 and anything else (sylpheed,
   chromium, gedit, kcachegrind...) is jerky as all hell. it seems to be a
   timing issue of some sort...
  
   bug #5:
  
   resizing of a window produces swimming of content as it is stretched
 to
   fill the new border UNTIL the client can render an updated set of
 content
   (and then fix it up). the effect basically is that content squashes and
   expands back and forth during resizing as opposed to remaining
 unchanged
   in scaling. comp used to address this nicely by enforcing e_comp size
 ==
   size of pixmap dimensions. :)
  
bug #1:
   
1. start terminal (xterm/terminology - any window actually)
2. type cmd: sleep 5; exit
3. alt+left click and drag to move window. keep moving it until
 window
closes/exits
4. release mouse.
5. segfault happiness! :)
   
(a quick look says to me action_border is still pointing to a border
 that
has now been deleted - crashdump attached).
   
bug #2:
   
something has happened to the shadow in comp... it's now wrong. it
 has
moved down from the title:
   
   
 http://www.enlightenment.org/ss/display.php?image=e-516c92a8841755.22448970.png
   
vs old:
   
   
 http://www.enlightenment.org/ss/display.php?image=e-515c83bebe2212.05354344.jpg
   
look carefully at shadow on left/right size of the window near the
titlebar top. now the shadow begins much lower.. just below client
 start.
   
bug #3.
   
double-click a window to shade it. it USED to just become a nice
 simple
titlebar... no longer. :)
   
--
- Codito, ergo sum - I code, therefore I am
 --
The Rasterman (Carsten Haitzler)ras...@rasterman.com
   
  
  
   --
   - Codito, ergo sum - I code, therefore I am
 --
   The Rasterman (Carsten Haitzler)ras...@rasterman.com
  
  
  
 --
   Precog is a next-generation analytics platform capable of advanced
   analytics on semi-structured data. The platform includes APIs for
 building
   apps and a phenomenal toolset for data science. Developers can use
   our toolset for easy data analysis  visualization. Get a free account!
   http://www2.precog.com/precogplatform/slashdotnewsletter
   ___
   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
 
 
 
 --
  Precog is a next-generation analytics platform capable of advanced
  analytics on semi-structured data. The platform includes APIs for
 building
  apps and a phenomenal toolset for data science. Developers can use
  our toolset for easy data analysis  visualization. Get a free account!
  http://www2.precog.com/precogplatform/slashdotnewsletter
  ___
  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



 --
 Precog is a next-generation analytics platform capable of advanced
 analytics on semi-structured data. The platform includes APIs for building
 apps and a phenomenal toolset for data science. Developers can use
 our toolset for easy data analysis  visualization. Get a free account!
 http://www2.precog.com/precogplatform/slashdotnewsletter
 ___
 enlightenment-devel mailing list

Re: [E-devel] [EGIT] [apps/terminology] master 03/05: use cwd of term when creating new ones with splits and tabs.

AWESOME!
Was that this easy? I blame myself not to have a look at this code but just
complaining...
Anyhow thanks muchly!

Daniel Juyung Seo (SeoZ)




On Mon, Apr 15, 2013 at 11:42 PM, Carsten Haitzler - Enlightenment Git 
no-re...@enlightenment.org wrote:

 raster pushed a commit to branch master.

 commit fc60732d6498d19da983a98dfc181ab8b260edba
 Author: Carsten Haitzler (Rasterman) ras...@rasterman.com
 Date:   Mon Apr 15 13:32:33 2013 +0900

 use cwd of term when creating new ones with splits and tabs.
 ---
  src/bin/main.c   | 8 ++--
  src/bin/termio.c | 5 +++--
  2 files changed, 9 insertions(+), 4 deletions(-)

 diff --git a/src/bin/main.c b/src/bin/main.c
 index 7006a15..89d9c8d 100644
 --- a/src/bin/main.c
 +++ b/src/bin/main.c
 @@ -214,6 +214,7 @@ _split_split(Split *sp, Eina_Bool horizontal)
 Split *sp2;
 Evas_Object *o;
 Config *config;
 +   char buf[PATH_MAX], *wdir = NULL;

 if (!sp-term) return;

 @@ -242,8 +243,9 @@ _split_split(Split *sp, Eina_Bool horizontal)
 sp2-parent = sp;
 sp2-wn = sp-wn;
 config = config_fork(sp-term-config);
 +   if (termio_cwd_get(sp-term-term, buf, sizeof(buf))) wdir = buf;
 sp2-term = main_term_new(sp-wn, config,
 - NULL, EINA_FALSE, NULL,
 + NULL, EINA_FALSE, wdir,
   80, 24, EINA_FALSE);
 sp2-terms = eina_list_append(sp2-terms, sp2-term);
 _term_resize_track_start(sp2);
 @@ -310,14 +312,16 @@ main_new(Evas_Object *win, Evas_Object *term)
 Split *sp = _split_find(win, term);
 Config *config;
 int w, h;
 +   char buf[PATH_MAX], *wdir = NULL;

 if (!sp) return;
 _term_resize_track_stop(sp);
 evas_object_hide(sp-term-bg);
 config = config_fork(sp-term-config);
 termio_size_get(sp-term-term, w, h);
 +   if (termio_cwd_get(sp-term-term, buf, sizeof(buf))) wdir = buf;
 sp-term = main_term_new(sp-wn, config,
 -NULL, EINA_FALSE, NULL,
 +NULL, EINA_FALSE, wdir,
  w, h, EINA_FALSE);
 sp-terms = eina_list_append(sp-terms, sp-term);
 _term_resize_track_start(sp);
 diff --git a/src/bin/termio.c b/src/bin/termio.c
 index 121be37..4df0dc9 100644
 --- a/src/bin/termio.c
 +++ b/src/bin/termio.c
 @@ -3555,18 +3555,19 @@ termio_cwd_get(const Evas_Object *obj, char *buf,
 size_t size)
 char procpath[PATH_MAX];
 Termio *sd = evas_object_smart_data_get(obj);
 pid_t pid;
 +   ssize_t siz;

 if (!sd) return EINA_FALSE;

 pid = termpty_pid_get(sd-pty);
 snprintf(procpath, sizeof(procpath), /proc/%d/cwd, pid);
 -   if (readlink(procpath, buf, size)  1)
 +   if ((siz = readlink(procpath, buf, size))  1)
   {
  ERR(Could not load working directory %s: %s,
  procpath, strerror(errno));
  return EINA_FALSE;
   }
 -
 +   buf[siz] = 0;
 return EINA_TRUE;
  }


 --


 --
 Precog is a next-generation analytics platform capable of advanced
 analytics on semi-structured data. The platform includes APIs for building
 apps and a phenomenal toolset for data science. Developers can use
 our toolset for easy data analysis  visualization. Get a free account!
 http://www2.precog.com/precogplatform/slashdotnewsletter

--
Precog is a next-generation analytics platform capable of advanced
analytics on semi-structured data. The platform includes APIs for building
apps and a phenomenal toolset for data science. Developers can use
our toolset for easy data analysis  visualization. Get a free account!
http://www2.precog.com/precogplatform/slashdotnewsletter
___
enlightenment-devel mailing list
enlightenment-devel@lists.sourceforge.net
https://lists.sourceforge.net/lists/listinfo/enlightenment-devel

Re: [E-devel] [EGIT] [apps/terminology] master 03/05: use cwd of term when creating new ones with splits and tabs.

On Tue, 16 Apr 2013 16:36:32 +0900 Daniel Juyung Seo seojuyu...@gmail.com
said:

yes. it was that easy. dthats why i was perplexed tat the biggest complainers
didnt just do it... it didnt bother me in the slightest thus it was low on my
priority list :)

 AWESOME!
 Was that this easy? I blame myself not to have a look at this code but just
 complaining...
 Anyhow thanks muchly!
 
 Daniel Juyung Seo (SeoZ)
 
 
 
 
 On Mon, Apr 15, 2013 at 11:42 PM, Carsten Haitzler - Enlightenment Git 
 no-re...@enlightenment.org wrote:
 
  raster pushed a commit to branch master.
 
  commit fc60732d6498d19da983a98dfc181ab8b260edba
  Author: Carsten Haitzler (Rasterman) ras...@rasterman.com
  Date:   Mon Apr 15 13:32:33 2013 +0900
 
  use cwd of term when creating new ones with splits and tabs.
  ---
   src/bin/main.c   | 8 ++--
   src/bin/termio.c | 5 +++--
   2 files changed, 9 insertions(+), 4 deletions(-)
 
  diff --git a/src/bin/main.c b/src/bin/main.c
  index 7006a15..89d9c8d 100644
  --- a/src/bin/main.c
  +++ b/src/bin/main.c
  @@ -214,6 +214,7 @@ _split_split(Split *sp, Eina_Bool horizontal)
  Split *sp2;
  Evas_Object *o;
  Config *config;
  +   char buf[PATH_MAX], *wdir = NULL;
 
  if (!sp-term) return;
 
  @@ -242,8 +243,9 @@ _split_split(Split *sp, Eina_Bool horizontal)
  sp2-parent = sp;
  sp2-wn = sp-wn;
  config = config_fork(sp-term-config);
  +   if (termio_cwd_get(sp-term-term, buf, sizeof(buf))) wdir = buf;
  sp2-term = main_term_new(sp-wn, config,
  - NULL, EINA_FALSE, NULL,
  + NULL, EINA_FALSE, wdir,
80, 24, EINA_FALSE);
  sp2-terms = eina_list_append(sp2-terms, sp2-term);
  _term_resize_track_start(sp2);
  @@ -310,14 +312,16 @@ main_new(Evas_Object *win, Evas_Object *term)
  Split *sp = _split_find(win, term);
  Config *config;
  int w, h;
  +   char buf[PATH_MAX], *wdir = NULL;
 
  if (!sp) return;
  _term_resize_track_stop(sp);
  evas_object_hide(sp-term-bg);
  config = config_fork(sp-term-config);
  termio_size_get(sp-term-term, w, h);
  +   if (termio_cwd_get(sp-term-term, buf, sizeof(buf))) wdir = buf;
  sp-term = main_term_new(sp-wn, config,
  -NULL, EINA_FALSE, NULL,
  +NULL, EINA_FALSE, wdir,
   w, h, EINA_FALSE);
  sp-terms = eina_list_append(sp-terms, sp-term);
  _term_resize_track_start(sp);
  diff --git a/src/bin/termio.c b/src/bin/termio.c
  index 121be37..4df0dc9 100644
  --- a/src/bin/termio.c
  +++ b/src/bin/termio.c
  @@ -3555,18 +3555,19 @@ termio_cwd_get(const Evas_Object *obj, char *buf,
  size_t size)
  char procpath[PATH_MAX];
  Termio *sd = evas_object_smart_data_get(obj);
  pid_t pid;
  +   ssize_t siz;
 
  if (!sd) return EINA_FALSE;
 
  pid = termpty_pid_get(sd-pty);
  snprintf(procpath, sizeof(procpath), /proc/%d/cwd, pid);
  -   if (readlink(procpath, buf, size)  1)
  +   if ((siz = readlink(procpath, buf, size))  1)
{
   ERR(Could not load working directory %s: %s,
   procpath, strerror(errno));
   return EINA_FALSE;
}
  -
  +   buf[siz] = 0;
  return EINA_TRUE;
   }
 
 
  --
 
 
  --
  Precog is a next-generation analytics platform capable of advanced
  analytics on semi-structured data. The platform includes APIs for building
  apps and a phenomenal toolset for data science. Developers can use
  our toolset for easy data analysis  visualization. Get a free account!
  http://www2.precog.com/precogplatform/slashdotnewsletter
 
 --
 Precog is a next-generation analytics platform capable of advanced
 analytics on semi-structured data. The platform includes APIs for building
 apps and a phenomenal toolset for data science. Developers can use
 our toolset for easy data analysis  visualization. Get a free account!
 http://www2.precog.com/precogplatform/slashdotnewsletter
 ___
 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


--
Precog is a next-generation analytics platform capable of advanced
analytics on semi-structured data. The platform includes APIs for building
apps and a phenomenal toolset for data science. Developers can use
our toolset for easy data analysis  visualization. Get a free account!
http://www2.precog.com/precogplatform/slashdotnewsletter
___
enlightenment-devel mailing list

Re: [E-devel] new bugs. in e (border releated)

On Tue, 16 Apr 2013 07:53:11 +0100 Michael Blumenkrantz
michael.blumenkra...@gmail.com said:

 This sounds like a localized config bug. Try changing your desk flip
 transition.

all transitions except nonebreak here. i tried all of them.

 
 
 On Tue, Apr 16, 2013 at 3:51 AM, Carsten Haitzler ras...@rasterman.comwrote:
 
  On Tue, 16 Apr 2013 11:46:04 +0900 Carsten Haitzler (The Rasterman)
  ras...@rasterman.com said:
 
  bug #7:
 
  very often mouse events just stop working (going to clients). mouse
  press/move/release/wheel etc. after flipping virtual desktops a few times
  or alt
  +left dragging a window about.
 
   On Tue, 16 Apr 2013 09:37:54 +0900 Carsten Haitzler (The Rasterman)
   ras...@rasterman.com said:
  
   bug #6:
  
   xdaliclock -transparent is nicely broken (shape mask is applied and
  offset by
   title height). this used to work. it's one of my comp shape test cases.
  :)
  
On Tue, 16 Apr 2013 09:27:21 +0900 Carsten Haitzler (The Rasterman)
ras...@rasterman.com said:
   
bug #4:
   
moving windows around no matter how small or large they are is now
  really
jerky... or it is almost all of the time... terminology seems to be
  smooth
all the time, but xterm seems to be 50/50 and anything else (sylpheed,
chromium, gedit, kcachegrind...) is jerky as all hell. it seems to be a
timing issue of some sort...
   
bug #5:
   
resizing of a window produces swimming of content as it is stretched
  to
fill the new border UNTIL the client can render an updated set of
  content
(and then fix it up). the effect basically is that content squashes and
expands back and forth during resizing as opposed to remaining
  unchanged
in scaling. comp used to address this nicely by enforcing e_comp size
  ==
size of pixmap dimensions. :)
   
 bug #1:

 1. start terminal (xterm/terminology - any window actually)
 2. type cmd: sleep 5; exit
 3. alt+left click and drag to move window. keep moving it until
  window
 closes/exits
 4. release mouse.
 5. segfault happiness! :)

 (a quick look says to me action_border is still pointing to a border
  that
 has now been deleted - crashdump attached).

 bug #2:

 something has happened to the shadow in comp... it's now wrong. it
  has
 moved down from the title:


  http://www.enlightenment.org/ss/display.php?image=e-516c92a8841755.22448970.png

 vs old:


  http://www.enlightenment.org/ss/display.php?image=e-515c83bebe2212.05354344.jpg

 look carefully at shadow on left/right size of the window near the
 titlebar top. now the shadow begins much lower.. just below client
  start.

 bug #3.

 double-click a window to shade it. it USED to just become a nice
  simple
 titlebar... no longer. :)

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

   
   
--
- Codito, ergo sum - I code, therefore I am
  --
The Rasterman (Carsten Haitzler)ras...@rasterman.com
   
   
   
  --
Precog is a next-generation analytics platform capable of advanced
analytics on semi-structured data. The platform includes APIs for
  building
apps and a phenomenal toolset for data science. Developers can use
our toolset for easy data analysis  visualization. Get a free account!
http://www2.precog.com/precogplatform/slashdotnewsletter
___
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
  
  
  
  --
   Precog is a next-generation analytics platform capable of advanced
   analytics on semi-structured data. The platform includes APIs for
  building
   apps and a phenomenal toolset for data science. Developers can use
   our toolset for easy data analysis  visualization. Get a free account!
   http://www2.precog.com/precogplatform/slashdotnewsletter
   ___
   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
 
 
 
  --
  Precog is a next-generation analytics platform capable of advanced
  analytics on semi-structured data. The platform includes APIs for building
  apps

Re: [E-devel] [Question] About depcreated elm_object_item_object_get API

Hello, I don't want to export the internal object any more.
Anyway there is room to discuss as always.

Well.. history repeats itself.
This was already discussed before. Please refer the following thread.
http://www.mail-archive.com/enlightenment-devel@lists.sourceforge.net/msg48240.html


On Tue, Apr 16, 2013 at 3:29 PM, ChunEon Park her...@naver.com wrote:

 Does somebody agree on it?

 
 -Regards, Hermet-

 -Original Message-
 From: Bluezeryohpo...@gmail.com
 To: Enlightenment developer list
 enlightenment-devel@lists.sourceforge.net;
 Cc:
 Sent: 2013-04-12 (금) 20:16:20
 Subject: Re: [E-devel] [Question] About depcreated
 elm_object_item_object_get API

 Yes, theme dependency can be the problem.
 But there eixst pros and cons.


Yes there exists pros and cons.
But *MY* policy is not to export the internal objects as much as possible.
Managing one objects in the two different places deosn't look good.
Once we export the internal objects, people will abuse it with the very
high possibility.



 If application make customized winset theme, application can add programs
 or parts or something for their own purpose and do more things on theme and
 do not need to make their own customized winset. In this situation, there
 is no need to change the elementary and default theme.  Application install
 their app just with their extension theme.

 In this case, application should know the edje/evas APIs (or elementary
 should wrap it but this is almost impossible, I think). But I think that
 application should know those and use those for making perferctly there own
 purpose.

 Otherwise, application should make their own edje and add code for
 manupulate it (make own winset).  In this case, application also know
 edje/evas APIs.


Yes this is what I recommend as of now.
If you want to handle/manage something, you create it. Do not take other's
possessions. There must be a conflict.
Genlist supports full item style for this kind of use cases.



 Theme is already exported for applications. So application can make their
 own theme.
 And I requested to export theme (edje), currently it is hidden.



 2013/4/4 ChunEon Park hermet@naver.com

  Maybe it can be... but i don't like those kinds of apis since it will
  result in the increase of the theme compatibility break.
 


Agreed.


  For your information,
  Users even can't decide the ctxpopup size.
  Conceptually, the ctxpopup itself decides the size and position by
 itself.
 
  
  -Regards, Hermet-
 
  -Original Message-
  From: Bluezeryohpowel@gmail.com
  To: Enlightenment developer list
  enlightenment-devel@lists.sourceforge.net;
  Cc:
  Sent: 2013-04-04 (목) 16:31:02
  Subject: [E-devel] [Question] About depcreated elm_object_item_object_get
  API
 
  Dear EFL developers,
 
  I have question about getting geometry of edje part on elemenatry object
 or
  elm object item.
  If ctx popup or date time widget want to be placed on the text part, I
  cannot find the way.
 
  Actually application should depend on edje theme part on some cases.
  Applications should know the name of the swallowed part on the edje
 theme.
 
  So, How about elm_object_item_part_geomtry_get API? Or I think that


NO!! We don't wrap all the evas/edje apis from elm side.


  elm_object_item_object_get or elm_object_object_get seems to be needed.
  If applications can get the edje object from elementary widget or
  elementary object item,  they can use edje APIs directly.
  But elm_object_item_object_get API was depecated and removed. The reason
 to
  remove this API seems to prevent abuse of application developers.
  But, for the flexibility, this API can be needed.
  Previously, this API returns Evas_Object *.How about retunring
  *const*Evas_Object *. ?
  I can prevent chaning Object property.
 
 
  BRs,
  Kim.
 
  --
  BRs,
  Kim.


Thanks.

Daniel Juyung Seo (SeoZ)


 
 
 --
  Minimize network downtime and maximize team effectiveness.
  Reduce network management and security costs.Learn how to hire
  the most talented Cisco Certified professionals. Visit the
  Employer Resources Portal
  http://www.cisco.com/web/learning/employer_resources/index.html
  ___
  enlightenment-devel mailing list
  enlightenment-devel@lists.sourceforge.net
  https://lists.sourceforge.net/lists/listinfo/enlightenment-devel
 
 
 
 
 --
  Minimize network downtime and maximize team effectiveness.
  Reduce network management and security costs.Learn how to hire
  the most talented Cisco Certified professionals. Visit the
  Employer Resources Portal
  http://www.cisco.com/web/learning/employer_resources/index.html
  ___
  enlightenment-devel mailing list

Re: [E-devel] new bugs. in e (border releated)

Okay, fixed now


On Tue, Apr 16, 2013 at 9:33 AM, Carsten Haitzler ras...@rasterman.comwrote:

 On Tue, 16 Apr 2013 07:53:11 +0100 Michael Blumenkrantz
 michael.blumenkra...@gmail.com said:

  This sounds like a localized config bug. Try changing your desk flip
  transition.

 all transitions except nonebreak here. i tried all of them.

 
 
  On Tue, Apr 16, 2013 at 3:51 AM, Carsten Haitzler ras...@rasterman.com
 wrote:
 
   On Tue, 16 Apr 2013 11:46:04 +0900 Carsten Haitzler (The Rasterman)
   ras...@rasterman.com said:
  
   bug #7:
  
   very often mouse events just stop working (going to clients). mouse
   press/move/release/wheel etc. after flipping virtual desktops a few
 times
   or alt
   +left dragging a window about.
  
On Tue, 16 Apr 2013 09:37:54 +0900 Carsten Haitzler (The Rasterman)
ras...@rasterman.com said:
   
bug #6:
   
xdaliclock -transparent is nicely broken (shape mask is applied and
   offset by
title height). this used to work. it's one of my comp shape test
 cases.
   :)
   
 On Tue, 16 Apr 2013 09:27:21 +0900 Carsten Haitzler (The Rasterman)
 ras...@rasterman.com said:

 bug #4:

 moving windows around no matter how small or large they are is now
   really
 jerky... or it is almost all of the time... terminology seems to be
   smooth
 all the time, but xterm seems to be 50/50 and anything else
 (sylpheed,
 chromium, gedit, kcachegrind...) is jerky as all hell. it seems to
 be a
 timing issue of some sort...

 bug #5:

 resizing of a window produces swimming of content as it is
 stretched
   to
 fill the new border UNTIL the client can render an updated set of
   content
 (and then fix it up). the effect basically is that content
 squashes and
 expands back and forth during resizing as opposed to remaining
   unchanged
 in scaling. comp used to address this nicely by enforcing e_comp
 size
   ==
 size of pixmap dimensions. :)

  bug #1:
 
  1. start terminal (xterm/terminology - any window actually)
  2. type cmd: sleep 5; exit
  3. alt+left click and drag to move window. keep moving it until
   window
  closes/exits
  4. release mouse.
  5. segfault happiness! :)
 
  (a quick look says to me action_border is still pointing to a
 border
   that
  has now been deleted - crashdump attached).
 
  bug #2:
 
  something has happened to the shadow in comp... it's now wrong.
 it
   has
  moved down from the title:
 
 
  
 http://www.enlightenment.org/ss/display.php?image=e-516c92a8841755.22448970.png
 
  vs old:
 
 
  
 http://www.enlightenment.org/ss/display.php?image=e-515c83bebe2212.05354344.jpg
 
  look carefully at shadow on left/right size of the window near
 the
  titlebar top. now the shadow begins much lower.. just below
 client
   start.
 
  bug #3.
 
  double-click a window to shade it. it USED to just become a nice
   simple
  titlebar... no longer. :)
 
  --
  - Codito, ergo sum - I code, therefore I am
   --
  The Rasterman (Carsten Haitzler)ras...@rasterman.com
 


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



  
 --
 Precog is a next-generation analytics platform capable of advanced
 analytics on semi-structured data. The platform includes APIs for
   building
 apps and a phenomenal toolset for data science. Developers can use
 our toolset for easy data analysis  visualization. Get a free
 account!
 http://www2.precog.com/precogplatform/slashdotnewsletter
 ___
 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
   
   
   
  
 --
Precog is a next-generation analytics platform capable of advanced
analytics on semi-structured data. The platform includes APIs for
   building
apps and a phenomenal toolset for data science. Developers can use
our toolset for easy data analysis  visualization. Get a free
 account!
http://www2.precog.com/precogplatform/slashdotnewsletter
___
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)

Re: [E-devel] [EGIT] [core/elementary] master 01/01: elementary/elm_entry : When entry is disabled, scrolling is also disabled.

Why? I'm not sure that's wanted.

--
Tom.

On 16/04/13 09:59, WooHyun Jung - Enlightenment Git wrote:
 woohyun pushed a commit to branch master.

 commit 1304f6756ef4ae82cc0bced10d8d4873a5657670
 Author: WooHyun Jung wh0705.j...@samsung.com
 Date:   Tue Apr 16 17:59:02 2013 +0900

  elementary/elm_entry : When entry is disabled, scrolling is also 
 disabled.
 ---
   ChangeLog   | 4 
   NEWS| 1 +
   src/lib/elm_entry.c | 4 
   3 files changed, 9 insertions(+)

 diff --git a/ChangeLog b/ChangeLog
 index c1c1012..d3bc3f6 100644
 --- a/ChangeLog
 +++ b/ChangeLog
 @@ -1256,3 +1256,7 @@

   * Fix the policy is not changed when the theme is changed.
   The reload signal from edc is necessary always.
 +
 +2013-04-16  WooHyun Jung
 +
 +* When entry is disabled, scrolling is also disabled.
 diff --git a/NEWS b/NEWS
 index 611964c..89e0dc0 100644
 --- a/NEWS
 +++ b/NEWS
 @@ -208,6 +208,7 @@ Fixes:
  * Fix the region_show/region_bring_in don't have a limit at a paging 
 movement.
  * Fix the calculation double type number.
  * Fix the policy is not changed when the theme is changed.
 +   * When entry is disabled, scrolling is also disabled.

   Removals:

 diff --git a/src/lib/elm_entry.c b/src/lib/elm_entry.c
 index e9852f3..f6dec53 100644
 --- a/src/lib/elm_entry.c
 +++ b/src/lib/elm_entry.c
 @@ -463,11 +463,15 @@ _elm_entry_smart_disable(Eo *obj, void *_pd, va_list 
 *list)
  if (elm_object_disabled_get(obj))
{
   edje_object_signal_emit(sd-entry_edje, elm,state,disabled, 
 elm);
 +if (sd-scroll)
 +  eo_do(obj, elm_scrollable_interface_freeze_set(EINA_TRUE));
   sd-disabled = EINA_TRUE;
}
  else
{
   edje_object_signal_emit(sd-entry_edje, elm,state,enabled, elm);
 +if (sd-scroll)
 +  eo_do(obj, elm_scrollable_interface_freeze_set(EINA_FALSE));
   sd-disabled = EINA_FALSE;
}




--
Precog is a next-generation analytics platform capable of advanced
analytics on semi-structured data. The platform includes APIs for building
apps and a phenomenal toolset for data science. Developers can use
our toolset for easy data analysis  visualization. Get a free account!
http://www2.precog.com/precogplatform/slashdotnewsletter
___
enlightenment-devel mailing list
enlightenment-devel@lists.sourceforge.net
https://lists.sourceforge.net/lists/listinfo/enlightenment-devel

Re: [E-devel] [EGIT] [core/efl] master 01/03: Install CMake configs to the proper path

AWESOME!! This works perfectly. Now I can get rid of some CMAKE_PREFIX_PATH
definition from my build script!
Today is awesome day :)

Daniel Juyung Seo (SeoZ)


On Tue, Apr 16, 2013 at 5:08 PM, Igor Murzov - Enlightenment Git 
no-re...@enlightenment.org wrote:

 garik pushed a commit to branch master.

 commit 380df67f84f5be3ca9f93846a1d9dd76a608a68b
 Author: Igor Murzov e-m...@date.by
 Date:   Sat Apr 13 23:53:49 2013 +0400

 Install CMake configs to the proper path
 ---
  Makefile.am | 49 ++---
  1 file changed, 34 insertions(+), 15 deletions(-)

 diff --git a/Makefile.am b/Makefile.am
 index 9b2072c..d8d1131 100644
 --- a/Makefile.am
 +++ b/Makefile.am
 @@ -234,21 +234,40 @@ pkgconfig_DATA += pc/ecore-audio.pc
  endif

  # Cmake configs:
 -cmakeconfigdir = ${prefix}/share/cmake/Modules/
 -cmakeconfig_DATA = cmakeconfig/EinaConfig.cmake \
 -  cmakeconfig/EinaConfigVersion.cmake \
 -  cmakeconfig/EoConfig.cmake \
 -  cmakeconfig/EoConfigVersion.cmake \
 -  cmakeconfig/EetConfig.cmake \
 -  cmakeconfig/EetConfigVersion.cmake \
 -  cmakeconfig/EvasConfig.cmake \
 -  cmakeconfig/EvasConfigVersion.cmake \
 -  cmakeconfig/EcoreConfig.cmake \
 -  cmakeconfig/EcoreConfigVersion.cmake \
 -  cmakeconfig/EdjeConfig.cmake \
 -  cmakeconfig/EdjeConfigVersion.cmake \
 -  cmakeconfig/EfreetConfig.cmake \
 -  cmakeconfig/EfreetConfigVersion.cmake
 +eina_cmakeconfigdir = $(libdir)/cmake/Eina/
 +eina_cmakeconfig_DATA = \
 +cmakeconfig/EinaConfig.cmake \
 +cmakeconfig/EinaConfigVersion.cmake
 +
 +eo_cmakeconfigdir = $(libdir)/cmake/Eo/
 +eo_cmakeconfig_DATA = \
 +cmakeconfig/EoConfig.cmake \
 +cmakeconfig/EoConfigVersion.cmake
 +
 +eet_cmakeconfigdir = $(libdir)/cmake/Eet/
 +eet_cmakeconfig_DATA = \
 +cmakeconfig/EetConfig.cmake \
 +cmakeconfig/EetConfigVersion.cmake
 +
 +evas_cmakeconfigdir =  $(libdir)/cmake/Evas/
 +evas_cmakeconfig_DATA = \
 +cmakeconfig/EvasConfig.cmake \
 +cmakeconfig/EvasConfigVersion.cmake
 +
 +ecore_cmakeconfigdir =  $(libdir)/cmake/Ecore/
 +ecore_cmakeconfig_DATA = \
 +cmakeconfig/EcoreConfig.cmake \
 +cmakeconfig/EcoreConfigVersion.cmake
 +
 +edje_cmakeconfigdir =  $(libdir)/cmake/Edje/
 +edje_cmakeconfig_DATA = \
 +cmakeconfig/EdjeConfig.cmake \
 +cmakeconfig/EdjeConfigVersion.cmake
 +
 +efreet_cmakeconfigdir =  $(libdir)/cmake/Efreet/
 +efreet_cmakeconfig_DATA = \
 +cmakeconfig/EfreetConfig.cmake \
 +cmakeconfig/EfreetConfigVersion.cmake

  # D-Bus services:


 --


 --
 Precog is a next-generation analytics platform capable of advanced
 analytics on semi-structured data. The platform includes APIs for building
 apps and a phenomenal toolset for data science. Developers can use
 our toolset for easy data analysis  visualization. Get a free account!
 http://www2.precog.com/precogplatform/slashdotnewsletter

--
Precog is a next-generation analytics platform capable of advanced
analytics on semi-structured data. The platform includes APIs for building
apps and a phenomenal toolset for data science. Developers can use
our toolset for easy data analysis  visualization. Get a free account!
http://www2.precog.com/precogplatform/slashdotnewsletter
___
enlightenment-devel mailing list
enlightenment-devel@lists.sourceforge.net
https://lists.sourceforge.net/lists/listinfo/enlightenment-devel

[E-devel] Added a new gdb command: eo_backtrace + one related improvement

Hey there,

I added a new gdb command eo_backtrace. It essentially prints the 
backtrace but filters out all the lines that are from inside eo.

Also, if you use a recent enough gdb (I don't know how recent it should 
be, I know it doesn't work on Ubuntu 12.04 but does on recent arch), you 
get automatic module loading. This means you don't have to load the 
python module manually, it'll be automatically loaded when debugging 
objects that link against eo. You can also set older gdb version to 
always load the module (only adds commands so it's fine).

If you install the efl to a different prefix than the one gdb was 
installed to, you have to either set gdb's data dir/trusted module path 
to the correct one, or just symlink the gdb python file to the proper 
place. For example, assuming you installed eo to /usr/local and gdb to 
/usr, link /usr/local/share/gdb/auto-load/libeo* to 
/usr/share/gdb/auto-load/libeo*.

Have fun, and let me know if you face any issues.

Cheers,
Tom.

--
Precog is a next-generation analytics platform capable of advanced
analytics on semi-structured data. The platform includes APIs for building
apps and a phenomenal toolset for data science. Developers can use
our toolset for easy data analysis  visualization. Get a free account!
http://www2.precog.com/precogplatform/slashdotnewsletter
___
enlightenment-devel mailing list
enlightenment-devel@lists.sourceforge.net
https://lists.sourceforge.net/lists/listinfo/enlightenment-devel

Re: [E-devel] Added a new gdb command: eo_backtrace + one related improvement

please put this info on the site somewhere


On Tue, Apr 16, 2013 at 11:54 AM, Tom Hacohen tom.haco...@samsung.comwrote:

 Hey there,

 I added a new gdb command eo_backtrace. It essentially prints the
 backtrace but filters out all the lines that are from inside eo.

 Also, if you use a recent enough gdb (I don't know how recent it should
 be, I know it doesn't work on Ubuntu 12.04 but does on recent arch), you
 get automatic module loading. This means you don't have to load the
 python module manually, it'll be automatically loaded when debugging
 objects that link against eo. You can also set older gdb version to
 always load the module (only adds commands so it's fine).

 If you install the efl to a different prefix than the one gdb was
 installed to, you have to either set gdb's data dir/trusted module path
 to the correct one, or just symlink the gdb python file to the proper
 place. For example, assuming you installed eo to /usr/local and gdb to
 /usr, link /usr/local/share/gdb/auto-load/libeo* to
 /usr/share/gdb/auto-load/libeo*.

 Have fun, and let me know if you face any issues.

 Cheers,
 Tom.


 --
 Precog is a next-generation analytics platform capable of advanced
 analytics on semi-structured data. The platform includes APIs for building
 apps and a phenomenal toolset for data science. Developers can use
 our toolset for easy data analysis  visualization. Get a free account!
 http://www2.precog.com/precogplatform/slashdotnewsletter
 ___
 enlightenment-devel mailing list
 enlightenment-devel@lists.sourceforge.net
 https://lists.sourceforge.net/lists/listinfo/enlightenment-devel

--
Precog is a next-generation analytics platform capable of advanced
analytics on semi-structured data. The platform includes APIs for building
apps and a phenomenal toolset for data science. Developers can use
our toolset for easy data analysis  visualization. Get a free account!
http://www2.precog.com/precogplatform/slashdotnewsletter
___
enlightenment-devel mailing list
enlightenment-devel@lists.sourceforge.net
https://lists.sourceforge.net/lists/listinfo/enlightenment-devel

Re: [E-devel] [Question] About depcreated elm_object_item_object_get API

2013-04-16 Thread Bluezery

2013/4/16 Daniel Juyung Seo seojuyu...@gmail.com

 Hello, I don't want to export the internal object any more.


That depends on point of views.
Think about 3 mans
Designer / Frame developer / Application Developer

Edje is designed for designer. Designer makes edje. It can be compiled and
distributed itself.
Frame developer makes elementary widget. It can be also distributed itself.
Application developer uses elementary and edje.

Designer
  /   \
Frame Developer -  Application Developer

They depends on each others like above.
If application cannot access the edje. Above can be changed like below.

Framework Designer (widget theme) - Frame Developer - Application
Developer - Applicaiton Designer

So the problem is that edje is designed from Framework Designer can not be
reused for application developer.
Application designer overwrites (cannot override) it.



 Anyway there is room to discuss as always.

 Well.. history repeats itself.
 This was already discussed before. Please refer the following thread.

 http://www.mail-archive.com/enlightenment-devel@lists.sourceforge.net/msg48240.html


 On Tue, Apr 16, 2013 at 3:29 PM, ChunEon Park her...@naver.com wrote:

  Does somebody agree on it?
 
  
  -Regards, Hermet-
 
  -Original Message-
  From: Bluezeryohpo...@gmail.com
  To: Enlightenment developer list
  enlightenment-devel@lists.sourceforge.net;
  Cc:
  Sent: 2013-04-12 (금) 20:16:20
  Subject: Re: [E-devel] [Question] About depcreated
  elm_object_item_object_get API
 
  Yes, theme dependency can be the problem.
  But there eixst pros and cons.
 

 Yes there exists pros and cons.
 But *MY* policy is not to export the internal objects as much as possible.
 Managing one objects in the two different places deosn't look good.
 Once we export the internal objects, people will abuse it with the very
 high possibility.


 
  If application make customized winset theme, application can add programs
  or parts or something for their own purpose and do more things on theme
 and
  do not need to make their own customized winset. In this situation, there
  is no need to change the elementary and default theme.  Application
 install
  their app just with their extension theme.
 
  In this case, application should know the edje/evas APIs (or elementary
  should wrap it but this is almost impossible, I think). But I think that
  application should know those and use those for making perferctly there
 own
  purpose.
 
  Otherwise, application should make their own edje and add code for
  manupulate it (make own winset).  In this case, application also know
  edje/evas APIs.
 

 Yes this is what I recommend as of now.
 If you want to handle/manage something, you create it. Do not take other's
 possessions. There must be a conflict.
 Genlist supports full item style for this kind of use cases.


 
  Theme is already exported for applications. So application can make their
  own theme.
  And I requested to export theme (edje), currently it is hidden.
 
 
 
  2013/4/4 ChunEon Park hermet@naver.com
 
   Maybe it can be... but i don't like those kinds of apis since it will
   result in the increase of the theme compatibility break.
  
 

 Agreed.


   For your information,
   Users even can't decide the ctxpopup size.
   Conceptually, the ctxpopup itself decides the size and position by
  itself.
  
   
   -Regards, Hermet-
  
   -Original Message-
   From: Bluezeryohpowel@gmail.com
   To: Enlightenment developer list
   enlightenment-devel@lists.sourceforge.net;
   Cc:
   Sent: 2013-04-04 (목) 16:31:02
   Subject: [E-devel] [Question] About depcreated
 elm_object_item_object_get
   API
  
   Dear EFL developers,
  
   I have question about getting geometry of edje part on elemenatry
 object
  or
   elm object item.
   If ctx popup or date time widget want to be placed on the text part, I
   cannot find the way.
  
   Actually application should depend on edje theme part on some cases.
   Applications should know the name of the swallowed part on the edje
  theme.
  
   So, How about elm_object_item_part_geomtry_get API? Or I think that
 

 NO!! We don't wrap all the evas/edje apis from elm side.


   elm_object_item_object_get or elm_object_object_get seems to be needed.
   If applications can get the edje object from elementary widget or
   elementary object item,  they can use edje APIs directly.
   But elm_object_item_object_get API was depecated and removed. The
 reason
  to
   remove this API seems to prevent abuse of application developers.
   But, for the flexibility, this API can be needed.
   Previously, this API returns Evas_Object *.How about retunring
   *const*Evas_Object *. ?
   I can prevent chaning Object property.
  
  
   BRs,
   Kim.
  
   --
   BRs,
   Kim.
 

 Thanks.

 Daniel Juyung Seo (SeoZ)

Re: [E-devel] new bugs. in e (border releated)

fixed


On Tue, Apr 16, 2013 at 3:46 AM, Carsten Haitzler ras...@rasterman.comwrote:

 On Tue, 16 Apr 2013 09:37:54 +0900 Carsten Haitzler (The Rasterman)
 ras...@rasterman.com said:

 bug #6:

 xdaliclock -transparent is nicely broken (shape mask is applied and offset
 by
 title height). this used to work. it's one of my comp shape test cases. :)

  On Tue, 16 Apr 2013 09:27:21 +0900 Carsten Haitzler (The Rasterman)
  ras...@rasterman.com said:
 
  bug #4:
 
  moving windows around no matter how small or large they are is now really
  jerky... or it is almost all of the time... terminology seems to be
 smooth all
  the time, but xterm seems to be 50/50 and anything else (sylpheed,
 chromium,
  gedit, kcachegrind...) is jerky as all hell. it seems to be a timing
 issue of
  some sort...
 
  bug #5:
 
  resizing of a window produces swimming of content as it is stretched
 to fill
  the new border UNTIL the client can render an updated set of content
 (and then
  fix it up). the effect basically is that content squashes and expands
 back and
  forth during resizing as opposed to remaining unchanged in scaling.
 comp
  used to address this nicely by enforcing e_comp size == size of pixmap
  dimensions. :)
 
   bug #1:
  
   1. start terminal (xterm/terminology - any window actually)
   2. type cmd: sleep 5; exit
   3. alt+left click and drag to move window. keep moving it until window
   closes/exits
   4. release mouse.
   5. segfault happiness! :)
  
   (a quick look says to me action_border is still pointing to a border
 that
   has now been deleted - crashdump attached).
  
   bug #2:
  
   something has happened to the shadow in comp... it's now wrong. it has
 moved
   down from the title:
  
  
 http://www.enlightenment.org/ss/display.php?image=e-516c92a8841755.22448970.png
  
   vs old:
  
  
 http://www.enlightenment.org/ss/display.php?image=e-515c83bebe2212.05354344.jpg
  
   look carefully at shadow on left/right size of the window near the
 titlebar
   top. now the shadow begins much lower.. just below client start.
  
   bug #3.
  
   double-click a window to shade it. it USED to just become a nice simple
   titlebar... no longer. :)
  
   --
   - Codito, ergo sum - I code, therefore I am
 --
   The Rasterman (Carsten Haitzler)ras...@rasterman.com
  
 
 
  --
  - Codito, ergo sum - I code, therefore I am --
  The Rasterman (Carsten Haitzler)ras...@rasterman.com
 
 
 
 --
  Precog is a next-generation analytics platform capable of advanced
  analytics on semi-structured data. The platform includes APIs for
 building
  apps and a phenomenal toolset for data science. Developers can use
  our toolset for easy data analysis  visualization. Get a free account!
  http://www2.precog.com/precogplatform/slashdotnewsletter
  ___
  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



 --
 Precog is a next-generation analytics platform capable of advanced
 analytics on semi-structured data. The platform includes APIs for building
 apps and a phenomenal toolset for data science. Developers can use
 our toolset for easy data analysis  visualization. Get a free account!
 http://www2.precog.com/precogplatform/slashdotnewsletter
 ___
 enlightenment-devel mailing list
 enlightenment-devel@lists.sourceforge.net
 https://lists.sourceforge.net/lists/listinfo/enlightenment-devel

--
Precog is a next-generation analytics platform capable of advanced
analytics on semi-structured data. The platform includes APIs for building
apps and a phenomenal toolset for data science. Developers can use
our toolset for easy data analysis  visualization. Get a free account!
http://www2.precog.com/precogplatform/slashdotnewsletter
___
enlightenment-devel mailing list
enlightenment-devel@lists.sourceforge.net
https://lists.sourceforge.net/lists/listinfo/enlightenment-devel

Re: [E-devel] Added a new gdb command: eo_backtrace + one related improvement

Any specific locations you have in mind?

On 16/04/13 11:57, Michael Blumenkrantz wrote:
 please put this info on the site somewhere


 On Tue, Apr 16, 2013 at 11:54 AM, Tom Hacohen tom.haco...@samsung.comwrote:

 Hey there,

 I added a new gdb command eo_backtrace. It essentially prints the
 backtrace but filters out all the lines that are from inside eo.

 Also, if you use a recent enough gdb (I don't know how recent it should
 be, I know it doesn't work on Ubuntu 12.04 but does on recent arch), you
 get automatic module loading. This means you don't have to load the
 python module manually, it'll be automatically loaded when debugging
 objects that link against eo. You can also set older gdb version to
 always load the module (only adds commands so it's fine).

 If you install the efl to a different prefix than the one gdb was
 installed to, you have to either set gdb's data dir/trusted module path
 to the correct one, or just symlink the gdb python file to the proper
 place. For example, assuming you installed eo to /usr/local and gdb to
 /usr, link /usr/local/share/gdb/auto-load/libeo* to
 /usr/share/gdb/auto-load/libeo*.

 Have fun, and let me know if you face any issues.

 Cheers,
 Tom.


 --
 Precog is a next-generation analytics platform capable of advanced
 analytics on semi-structured data. The platform includes APIs for building
 apps and a phenomenal toolset for data science. Developers can use
 our toolset for easy data analysis  visualization. Get a free account!
 http://www2.precog.com/precogplatform/slashdotnewsletter
 ___
 enlightenment-devel mailing list
 enlightenment-devel@lists.sourceforge.net
 https://lists.sourceforge.net/lists/listinfo/enlightenment-devel

 --
 Precog is a next-generation analytics platform capable of advanced
 analytics on semi-structured data. The platform includes APIs for building
 apps and a phenomenal toolset for data science. Developers can use
 our toolset for easy data analysis  visualization. Get a free account!
 http://www2.precog.com/precogplatform/slashdotnewsletter
 ___
 enlightenment-devel mailing list
 enlightenment-devel@lists.sourceforge.net
 https://lists.sourceforge.net/lists/listinfo/enlightenment-devel



--
Precog is a next-generation analytics platform capable of advanced
analytics on semi-structured data. The platform includes APIs for building
apps and a phenomenal toolset for data science. Developers can use
our toolset for easy data analysis  visualization. Get a free account!
http://www2.precog.com/precogplatform/slashdotnewsletter
___
enlightenment-devel mailing list
enlightenment-devel@lists.sourceforge.net
https://lists.sourceforge.net/lists/listinfo/enlightenment-devel

Re: [E-devel] Added a new gdb command: eo_backtrace + one related improvement

AWESOMENESS! x 100.
Today is the best AWESOME prize day.

This works perfectly here in Ubuntu 12.04 with .gdbinit configuration.
Coolz

Daniel Juyung Seo (SeoZ)



On Tue, Apr 16, 2013 at 7:54 PM, Tom Hacohen tom.haco...@samsung.comwrote:

 Hey there,

 I added a new gdb command eo_backtrace. It essentially prints the
 backtrace but filters out all the lines that are from inside eo.

 Also, if you use a recent enough gdb (I don't know how recent it should
 be, I know it doesn't work on Ubuntu 12.04 but does on recent arch), you
 get automatic module loading. This means you don't have to load the
 python module manually, it'll be automatically loaded when debugging
 objects that link against eo. You can also set older gdb version to
 always load the module (only adds commands so it's fine).

 If you install the efl to a different prefix than the one gdb was
 installed to, you have to either set gdb's data dir/trusted module path
 to the correct one, or just symlink the gdb python file to the proper
 place. For example, assuming you installed eo to /usr/local and gdb to
 /usr, link /usr/local/share/gdb/auto-load/libeo* to
 /usr/share/gdb/auto-load/libeo*.

 Have fun, and let me know if you face any issues.

 Cheers,
 Tom.


 --
 Precog is a next-generation analytics platform capable of advanced
 analytics on semi-structured data. The platform includes APIs for building
 apps and a phenomenal toolset for data science. Developers can use
 our toolset for easy data analysis  visualization. Get a free account!
 http://www2.precog.com/precogplatform/slashdotnewsletter
 ___
 enlightenment-devel mailing list
 enlightenment-devel@lists.sourceforge.net
 https://lists.sourceforge.net/lists/listinfo/enlightenment-devel

--
Precog is a next-generation analytics platform capable of advanced
analytics on semi-structured data. The platform includes APIs for building
apps and a phenomenal toolset for data science. Developers can use
our toolset for easy data analysis  visualization. Get a free account!
http://www2.precog.com/precogplatform/slashdotnewsletter
___
enlightenment-devel mailing list
enlightenment-devel@lists.sourceforge.net
https://lists.sourceforge.net/lists/listinfo/enlightenment-devel

Re: [E-devel] Added a new gdb command: eo_backtrace + one related improvement

Glad you like it. Please consider pasting your gdbinit here so other 
people may benefit from it.

--
Tom.

On 16/04/13 12:27, Daniel Juyung Seo wrote:
 AWESOMENESS! x 100.
 Today is the best AWESOME prize day.

 This works perfectly here in Ubuntu 12.04 with .gdbinit configuration.
 Coolz

 Daniel Juyung Seo (SeoZ)



 On Tue, Apr 16, 2013 at 7:54 PM, Tom Hacohen tom.haco...@samsung.comwrote:

 Hey there,

 I added a new gdb command eo_backtrace. It essentially prints the
 backtrace but filters out all the lines that are from inside eo.

 Also, if you use a recent enough gdb (I don't know how recent it should
 be, I know it doesn't work on Ubuntu 12.04 but does on recent arch), you
 get automatic module loading. This means you don't have to load the
 python module manually, it'll be automatically loaded when debugging
 objects that link against eo. You can also set older gdb version to
 always load the module (only adds commands so it's fine).

 If you install the efl to a different prefix than the one gdb was
 installed to, you have to either set gdb's data dir/trusted module path
 to the correct one, or just symlink the gdb python file to the proper
 place. For example, assuming you installed eo to /usr/local and gdb to
 /usr, link /usr/local/share/gdb/auto-load/libeo* to
 /usr/share/gdb/auto-load/libeo*.

 Have fun, and let me know if you face any issues.

 Cheers,
 Tom.


 --
 Precog is a next-generation analytics platform capable of advanced
 analytics on semi-structured data. The platform includes APIs for building
 apps and a phenomenal toolset for data science. Developers can use
 our toolset for easy data analysis  visualization. Get a free account!
 http://www2.precog.com/precogplatform/slashdotnewsletter
 ___
 enlightenment-devel mailing list
 enlightenment-devel@lists.sourceforge.net
 https://lists.sourceforge.net/lists/listinfo/enlightenment-devel

 --
 Precog is a next-generation analytics platform capable of advanced
 analytics on semi-structured data. The platform includes APIs for building
 apps and a phenomenal toolset for data science. Developers can use
 our toolset for easy data analysis  visualization. Get a free account!
 http://www2.precog.com/precogplatform/slashdotnewsletter
 ___
 enlightenment-devel mailing list
 enlightenment-devel@lists.sourceforge.net
 https://lists.sourceforge.net/lists/listinfo/enlightenment-devel



--
Precog is a next-generation analytics platform capable of advanced
analytics on semi-structured data. The platform includes APIs for building
apps and a phenomenal toolset for data science. Developers can use
our toolset for easy data analysis  visualization. Get a free account!
http://www2.precog.com/precogplatform/slashdotnewsletter
___
enlightenment-devel mailing list
enlightenment-devel@lists.sourceforge.net
https://lists.sourceforge.net/lists/listinfo/enlightenment-devel

Re: [E-devel] Added a new gdb command: eo_backtrace + one related improvement

Sure.

~/.gdbinit
source /usr/local/share/eo/gdb/eo_gdb.py

It's that simple. This works on Ubuntu 12.04 without additional
configuration.
And here is the result of bt vs eo_backtrace.
https://phab.enlightenment.org/P21

Enjoy it :)

Daniel Juyung Seo (SeoZ)

On Tue, Apr 16, 2013 at 8:33 PM, Tom Hacohen tom.haco...@samsung.comwrote:

Glad you like it. Please consider pasting your gdbinit here so other
people may benefit from it.

--
Tom.

On 16/04/13 12:27, Daniel Juyung Seo wrote:

AWESOMENESS! x 100.
Today is the best AWESOME prize day.

This works perfectly here in Ubuntu 12.04 with .gdbinit configuration.
Coolz

Daniel Juyung Seo (SeoZ)

On Tue, Apr 16, 2013 at 7:54 PM, Tom Hacohen tom.haco...@samsung.com**
wrote:

Hey there,

I added a new gdb command eo_backtrace. It essentially prints the
backtrace but filters out all the lines that are from inside eo.

Also, if you use a recent enough gdb (I don't know how recent it should
be, I know it doesn't work on Ubuntu 12.04 but does on recent arch), you
get automatic module loading. This means you don't have to load the
python module manually, it'll be automatically loaded when debugging
objects that link against eo. You can also set older gdb version to
always load the module (only adds commands so it's fine).

If you install the efl to a different prefix than the one gdb was
installed to, you have to either set gdb's data dir/trusted module path
to the correct one, or just symlink the gdb python file to the proper
place. For example, assuming you installed eo to /usr/local and gdb to
/usr, link /usr/local/share/gdb/auto-**load/libeo* to
/usr/share/gdb/auto-load/**libeo*.

Have fun, and let me know if you face any issues.

Cheers,
Tom.

--**--**
--
Precog is a next-generation analytics platform capable of advanced
analytics on semi-structured data. The platform includes APIs for
building
apps and a phenomenal toolset for data science. Developers can use
our toolset for easy data analysis visualization. Get a free account!
http://www2.precog.com/**precogplatform/**slashdotnewsletterhttp://www2.precog.com/precogplatform/slashdotnewsletter
__**_
enlightenment-devel mailing list
enlightenment-devel@lists.**sourceforge.netenlightenment-devel@lists.sourceforge.net
https://lists.sourceforge.net/**lists/listinfo/enlightenment-**develhttps://lists.sourceforge.net/lists/listinfo/enlightenment-devel

--**--**
--
Precog is a next-generation analytics platform capable of advanced
analytics on semi-structured data. The platform includes APIs for building
apps and a phenomenal toolset for data science. Developers can use
our toolset for easy data analysis visualization. Get a free account!
http://www2.precog.com/**precogplatform/**slashdotnewsletterhttp://www2.precog.com/precogplatform/slashdotnewsletter
__**_
enlightenment-devel mailing list
enlightenment-devel@lists.**sourceforge.netenlightenment-devel@lists.sourceforge.net
https://lists.sourceforge.net/**lists/listinfo/enlightenment-**develhttps://lists.sourceforge.net/lists/listinfo/enlightenment-devel

--
Precog is a next-generation analytics platform capable of advanced
analytics on semi-structured data. The platform includes APIs for building
apps and a phenomenal toolset for data science. Developers can use
our toolset for easy data analysis visualization. Get a free account!
http://www2.precog.com/precogplatform/slashdotnewsletter
___
enlightenment-devel mailing list
enlightenment-devel@lists.sourceforge.net
https://lists.sourceforge.net/lists/listinfo/enlightenment-devel

Re: [E-devel] Added a new gdb command: eo_backtrace + one related improvement

On 16/04/13 12:40, Daniel Juyung Seo wrote:
 Sure.

 ~/.gdbinit
 source /usr/local/share/eo/gdb/eo_gdb.py

 It's that simple. This works on Ubuntu 12.04 without additional
 configuration.
 And here is the result of bt vs eo_backtrace.
 https://phab.enlightenment.org/P21

Oh wow, it's really nice with all the constructors chained like that. 
Thanks for sharing your gdbinit, I hope people find good use for it.

--
Tom.


--
Precog is a next-generation analytics platform capable of advanced
analytics on semi-structured data. The platform includes APIs for building
apps and a phenomenal toolset for data science. Developers can use
our toolset for easy data analysis  visualization. Get a free account!
http://www2.precog.com/precogplatform/slashdotnewsletter
___
enlightenment-devel mailing list
enlightenment-devel@lists.sourceforge.net
https://lists.sourceforge.net/lists/listinfo/enlightenment-devel

Re: [E-devel] [Question] About depcreated elm_object_item_object_get API

On Tue, Apr 16, 2013 at 8:02 PM, Bluezery ohpo...@gmail.com wrote:

 2013/4/16 Daniel Juyung Seo seojuyu...@gmail.com

  Hello, I don't want to export the internal object any more.
 

 That depends on point of views.
 Think about 3 mans
 Designer / Frame developer / Application Developer

 Edje is designed for designer. Designer makes edje. It can be compiled and
 distributed itself.
 Frame developer makes elementary widget. It can be also distributed itself.
 Application developer uses elementary and edje.

 Designer
   /   \
 Frame Developer -  Application Developer

 They depends on each others like above.


True.


 If application cannot access the edje. Above can be changed like below.


Well you are talking about general edje object. THE EDJE OBJECT we were
talking about was an edje object which was created inside the widget.
Exposing this internal object is same as exposing class' private members to
public.

If you create your own edje object and you handle it, I don't care.



 Framework Designer (widget theme) - Frame Developer - Application
 Developer - Applicaiton Designer

 So the problem is that edje is designed from Framework Designer can not be
 reused for application developer.
 Application designer overwrites (cannot override) it.


Are you talking about theme by edje is designed from Framework Designer?
Of course you can use it by elm_layout_theme_set. There is no limitation
for reusing the any styles in theme.
You are just not allowed to hook the other's edje object which is
instanciated version of styles/themes whatsoever.

It looks like this discussion is headed to the wrong direction :(

Thanks.

Daniel Juyung Seo (SeoZ)





  Anyway there is room to discuss as always.
 
  Well.. history repeats itself.
  This was already discussed before. Please refer the following thread.
 
 
 http://www.mail-archive.com/enlightenment-devel@lists.sourceforge.net/msg48240.html
 
 
  On Tue, Apr 16, 2013 at 3:29 PM, ChunEon Park her...@naver.com wrote:
 
   Does somebody agree on it?
  
   
   -Regards, Hermet-
  
   -Original Message-
   From: Bluezeryohpo...@gmail.com
   To: Enlightenment developer list
   enlightenment-devel@lists.sourceforge.net;
   Cc:
   Sent: 2013-04-12 (금) 20:16:20
   Subject: Re: [E-devel] [Question] About depcreated
   elm_object_item_object_get API
  
   Yes, theme dependency can be the problem.
   But there eixst pros and cons.
  
 
  Yes there exists pros and cons.
  But *MY* policy is not to export the internal objects as much as
 possible.
  Managing one objects in the two different places deosn't look good.
  Once we export the internal objects, people will abuse it with the very
  high possibility.
 
 
  
   If application make customized winset theme, application can add
 programs
   or parts or something for their own purpose and do more things on theme
  and
   do not need to make their own customized winset. In this situation,
 there
   is no need to change the elementary and default theme.  Application
  install
   their app just with their extension theme.
  
   In this case, application should know the edje/evas APIs (or elementary
   should wrap it but this is almost impossible, I think). But I think
 that
   application should know those and use those for making perferctly there
  own
   purpose.
  
   Otherwise, application should make their own edje and add code for
   manupulate it (make own winset).  In this case, application also know
   edje/evas APIs.
  
 
  Yes this is what I recommend as of now.
  If you want to handle/manage something, you create it. Do not take
 other's
  possessions. There must be a conflict.
  Genlist supports full item style for this kind of use cases.
 
 
  
   Theme is already exported for applications. So application can make
 their
   own theme.
   And I requested to export theme (edje), currently it is hidden.
  
  
  
   2013/4/4 ChunEon Park hermet@naver.com
  
Maybe it can be... but i don't like those kinds of apis since it will
result in the increase of the theme compatibility break.
   
  
 
  Agreed.
 
 
For your information,
Users even can't decide the ctxpopup size.
Conceptually, the ctxpopup itself decides the size and position by
   itself.
   

-Regards, Hermet-
   
-Original Message-
From: Bluezeryohpowel@gmail.com
To: Enlightenment developer list
enlightenment-devel@lists.sourceforge.net;
Cc:
Sent: 2013-04-04 (목) 16:31:02
Subject: [E-devel] [Question] About depcreated
  elm_object_item_object_get
API
   
Dear EFL developers,
   
I have question about getting geometry of edje part on elemenatry
  object
   or
elm object item.
If ctx popup or date time widget want to be placed on the text part,
 I
cannot find the way.
   
Actually application should depend on edje theme part on some cases.

Re: [E-devel] [EGIT] [core/enlightenment] master 01/01: add zoomap so we get back our zooming comp client things with.. menu for today. popups need doing... anything that bypasses the comp native surf

2013-04-16 Thread Leif Middelschulte

http://idahofalls.webimpakt.com/wwwroot/userfiles/images/pr/zoo/mapforweb2.gif

Let's meet at the bar aka. drinking fountain later. 

-- 
Leif


Am Dienstag, 16. April 2013 um 13:07 schrieb Carsten Haitzler - Enlightenment 
Git:

 raster pushed a commit to branch master.
 
 commit e0e513f457866042c0099dd6927e1784738bdb91
 Author: Carsten Haitzler (Rasterman) ras...@rasterman.com 
 (mailto:ras...@rasterman.com)
 Date: Tue Apr 16 20:06:47 2013 +0900
 
 add zoomap so we get back our zooming comp client things with.. menu
 for today. popups need doing... anything that bypasses the comp native
 surface stuff needs this. :)
 ---
 src/bin/Makefile.am (http://Makefile.am) | 2 +
 src/bin/e_includes.h | 1 +
 src/bin/e_menu.c | 29 -
 src/bin/e_menu.h | 1 +
 src/bin/e_zoomap.c | 323 +++
 src/bin/e_zoomap.h | 17 +++
 6 files changed, 367 insertions(+), 6 deletions(-)
 
 diff --git a/src/bin/Makefile.am (http://Makefile.am) b/src/bin/Makefile.am 
 (http://Makefile.am)
 index 675bccf..710112b 100644
 --- a/src/bin/Makefile.am (http://Makefile.am)
 +++ b/src/bin/Makefile.am (http://Makefile.am)
 @@ -201,6 +201,7 @@ e_win.h \
 e_xinerama.h \
 e_xkb.h \
 e_xsettings.h \
 +e_zoomap.h \
 e_zone.h
 
 if HAVE_WAYLAND_CLIENTS
 @@ -371,6 +372,7 @@ e_win.c \
 e_xinerama.c \
 e_xkb.c \
 e_xsettings.c \
 +e_zoomap.c \
 e_zone.c \
 $(ENLIGHTENMENTHEADERS)
 
 diff --git a/src/bin/e_includes.h b/src/bin/e_includes.h
 index 5863b22..70b5ceb 100644
 --- a/src/bin/e_includes.h
 +++ b/src/bin/e_includes.h
 @@ -55,6 +55,7 @@
 #include e_remember.h
 #include e_win.h
 #include e_pan.h
 +#include e_zoomap.h
 #include e_dialog.h
 #include e_configure.h
 #include e_configure_option.h
 diff --git a/src/bin/e_menu.c b/src/bin/e_menu.c
 index e333eb8..0c37fad 100644
 --- a/src/bin/e_menu.c
 +++ b/src/bin/e_menu.c
 @@ -1170,6 +1170,8 @@ e_menu_idler_before(void)
 m-prev.h = m-cur.h;
 w = m-cur.w;
 h = m-cur.h;
 + evas_object_resize(m-bg_object, w, h);
 + evas_object_resize(m-bg_object_wrap, w, h);
 if (m-cw)
 e_comp_win_resize(m-cw, w, h);
 }
 @@ -1210,9 +1212,10 @@ e_menu_idler_before(void)
 m-prev.visible = m-cur.visible;
 if (!m-cw)
 {
 - evas_object_move(m-bg_object, m-cur.x, m-cur.y);
 evas_object_resize(m-bg_object, m-cur.w, m-cur.h);
 - E_LAYER_SET(m-bg_object, E_COMP_CANVAS_LAYER_MENU);
 + evas_object_move(m-bg_object_wrap, m-cur.x, m-cur.y);
 + evas_object_resize(m-bg_object_wrap, m-cur.w, m-cur.h);
 + E_LAYER_SET(m-bg_object_wrap, E_COMP_CANVAS_LAYER_MENU);
 }
 e_comp_win_show(m-cw);
 }
 @@ -1284,7 +1287,8 @@ _e_menu_free(E_Menu *m)
 if (m-parent_item)
 m-parent_item-submenu = NULL;
 /* del callback causes this to unrealize the menu */
 - if (m-bg_object) evas_object_del(m-bg_object);
 + if (m-bg_object_wrap) evas_object_del(m-bg_object_wrap);
 + m-bg_object_wrap = NULL;
 EINA_LIST_FOREACH_SAFE(m-items, l, l_next, mi)
 e_object_del(E_OBJECT(mi));
 if (m-in_active_list)
 @@ -1393,6 +1397,8 @@ _e_menu_del_cb(void *data, Evas *e EINA_UNUSED, 
 Evas_Object *obj EINA_UNUSED, vo
 E_Menu *m = data;
 
 m-bg_object = NULL;
 + evas_object_del(m-bg_object_wrap);
 + m-bg_object_wrap = NULL;
 _e_menu_unrealize(m);
 }
 
 @@ -1688,6 +1694,7 @@ _e_menu_realize(E_Menu *m)
 Evas_Object *o;
 Eina_List *l;
 E_Menu_Item *mi;
 + const char *s;
 
 if (m-realized || (!m-items)) return;
 m-realized = 1;
 @@ -1701,9 +1708,6 @@ _e_menu_realize(E_Menu *m)
 o = edje_object_add(m-evas);
 evas_object_event_callback_add(o, EVAS_CALLBACK_DEL, _e_menu_del_cb, m);
 m-bg_object = o;
 - evas_object_name_set(o, menu-bg_object);
 - evas_object_data_set(o, e_menu, m);
 - evas_object_data_set(o, eobj, m);
 e_theme_edje_object_set(o, base/theme/menus, 
 e/widgets/menu/default/background);
 if (m-header.title)
 {
 @@ -1712,6 +1716,17 @@ _e_menu_realize(E_Menu *m)
 edje_object_message_signal_process(o);
 }
 
 + o = e_zoomap_add(m-evas);
 + evas_object_name_set(o, menu-bg_object_wrap);
 + evas_object_data_set(o, e_menu, m);
 + evas_object_data_set(o, eobj, m);
 + m-bg_object_wrap = o;
 + s = edje_object_data_get(m-bg_object, argb);
 + if (!s) s = edje_object_data_get(m-bg_object, shaped);
 + if ((s)  (s[0] == '1')) e_zoomap_solid_set(o, EINA_FALSE);
 + else e_zoomap_solid_set(o, EINA_TRUE);
 + e_zoomap_child_set(o, m-bg_object);
 +
 o = e_box_add(m-evas);
 evas_object_name_set(o, menu-container_object);
 m-container_object = o;
 @@ -1981,6 +1996,8 @@ _e_menu_unrealize(E_Menu *m)
 m-header.icon = NULL;
 if (m-bg_object) evas_object_del(m-bg_object);
 m-bg_object = NULL;
 + if (m-bg_object_wrap) evas_object_del(m-bg_object_wrap);
 + m-bg_object_wrap = NULL;
 if (m-container_object) evas_object_del(m-container_object);
 m-container_object = NULL;
 m-cur.visible = 0;
 diff --git a/src/bin/e_menu.h b/src/bin/e_menu.h
 index 8c4d05c..36f22cc 100644
 --- a/src/bin/e_menu.h
 +++ b/src/bin/e_menu.h
 @@ -56,6 +56,7 @@ struct _E_Menu
 E_Container_Shape *shape;
 Ecore_Job *dangling_job;
 Evas *evas;
 + Evas_Object *bg_object_wrap;
 Evas_Object

Re: [E-devel] Added a new gdb command: eo_backtrace + one related improvement

2013-04-16 Thread Cedric BAIL

Cedric Bail
On Apr 16, 2013 4:09 AM, Tom Hacohen tom.haco...@samsung.com wrote:

 Any specific locations you have in mind?

Maybe it's a good time to add a page on phab on how to debug efl
application. Describing also eina log, Clouseau and more generic stuff like
valgring.

 On 16/04/13 11:57, Michael Blumenkrantz wrote:
  please put this info on the site somewhere
 
 
  On Tue, Apr 16, 2013 at 11:54 AM, Tom Hacohen tom.haco...@samsung.com
wrote:
 
  Hey there,
 
  I added a new gdb command eo_backtrace. It essentially prints the
  backtrace but filters out all the lines that are from inside eo.
 
  Also, if you use a recent enough gdb (I don't know how recent it should
  be, I know it doesn't work on Ubuntu 12.04 but does on recent arch),
you
  get automatic module loading. This means you don't have to load the
  python module manually, it'll be automatically loaded when debugging
  objects that link against eo. You can also set older gdb version to
  always load the module (only adds commands so it's fine).
 
  If you install the efl to a different prefix than the one gdb was
  installed to, you have to either set gdb's data dir/trusted module path
  to the correct one, or just symlink the gdb python file to the proper
  place. For example, assuming you installed eo to /usr/local and gdb
to
  /usr, link /usr/local/share/gdb/auto-load/libeo* to
  /usr/share/gdb/auto-load/libeo*.
 
  Have fun, and let me know if you face any issues.
 
  Cheers,
  Tom.
 
 
 
--
  Precog is a next-generation analytics platform capable of advanced
  analytics on semi-structured data. The platform includes APIs for
building
  apps and a phenomenal toolset for data science. Developers can use
  our toolset for easy data analysis  visualization. Get a free account!
  http://www2.precog.com/precogplatform/slashdotnewsletter
  ___
  enlightenment-devel mailing list
  enlightenment-devel@lists.sourceforge.net
  https://lists.sourceforge.net/lists/listinfo/enlightenment-devel
 
 
--
  Precog is a next-generation analytics platform capable of advanced
  analytics on semi-structured data. The platform includes APIs for
building
  apps and a phenomenal toolset for data science. Developers can use
  our toolset for easy data analysis  visualization. Get a free account!
  http://www2.precog.com/precogplatform/slashdotnewsletter
  ___
  enlightenment-devel mailing list
  enlightenment-devel@lists.sourceforge.net
  https://lists.sourceforge.net/lists/listinfo/enlightenment-devel
 



--
 Precog is a next-generation analytics platform capable of advanced
 analytics on semi-structured data. The platform includes APIs for building
 apps and a phenomenal toolset for data science. Developers can use
 our toolset for easy data analysis  visualization. Get a free account!
 http://www2.precog.com/precogplatform/slashdotnewsletter
 ___
 enlightenment-devel mailing list
 enlightenment-devel@lists.sourceforge.net
 https://lists.sourceforge.net/lists/listinfo/enlightenment-devel

--
Precog is a next-generation analytics platform capable of advanced
analytics on semi-structured data. The platform includes APIs for building
apps and a phenomenal toolset for data science. Developers can use
our toolset for easy data analysis  visualization. Get a free account!
http://www2.precog.com/precogplatform/slashdotnewsletter
___
enlightenment-devel mailing list
enlightenment-devel@lists.sourceforge.net
https://lists.sourceforge.net/lists/listinfo/enlightenment-devel

Re: [E-devel] [Question] About depcreated elm_object_item_object_get API

2013-04-16 Thread Bluezery

2013/4/16 Daniel Juyung Seo seojuyu...@gmail.com

 On Tue, Apr 16, 2013 at 8:02 PM, Bluezery ohpo...@gmail.com wrote:

  2013/4/16 Daniel Juyung Seo seojuyu...@gmail.com
 
   Hello, I don't want to export the internal object any more.
  
 
  That depends on point of views.
  Think about 3 mans
  Designer / Frame developer / Application Developer
 
  Edje is designed for designer. Designer makes edje. It can be compiled
 and
  distributed itself.
  Frame developer makes elementary widget. It can be also distributed
 itself.
  Application developer uses elementary and edje.
 
  Designer
/   \
  Frame Developer -  Application Developer
 
  They depends on each others like above.
 

 True.


  If application cannot access the edje. Above can be changed like below.
 

 Well you are talking about general edje object. THE EDJE OBJECT we were
 talking about was an edje object which was created inside the widget.
 Exposing this internal object is same as exposing class' private members to
 public.


edje object is created on widget side initilally.
But elm_object_style_set or elm_theme_extension_add can load edje fo
application.


 If you create your own edje object and you handle it, I don't care.


 
  Framework Designer (widget theme) - Frame Developer - Application
  Developer - Applicaiton Designer
 
  So the problem is that edje is designed from Framework Designer can not
 be
  reused for application developer.
  Application designer overwrites (cannot override) it.
 

 Are you talking about theme by edje is designed from Framework Designer?
 Of course you can use it by elm_layout_theme_set. There is no limitation
 for reusing the any styles in theme.
 You are just not allowed to hook the other's edje object which is
 instanciated version of styles/themes whatsoever.


Yes.
But this edje cannot interact with frameworks.
So framework functionality cannot be reused.

I don't know the policy of the elementary widget.
But I want to find better way. :)



 It looks like this discussion is headed to the wrong direction :(

 Thanks.

 Daniel Juyung Seo (SeoZ)


 
 
 
   Anyway there is room to discuss as always.
  
   Well.. history repeats itself.
   This was already discussed before. Please refer the following thread.
  
  
 
 http://www.mail-archive.com/enlightenment-devel@lists.sourceforge.net/msg48240.html
  
  
   On Tue, Apr 16, 2013 at 3:29 PM, ChunEon Park her...@naver.com
 wrote:
  
Does somebody agree on it?
   

-Regards, Hermet-
   
-Original Message-
From: Bluezeryohpo...@gmail.com
To: Enlightenment developer list
enlightenment-devel@lists.sourceforge.net;
Cc:
Sent: 2013-04-12 (금) 20:16:20
Subject: Re: [E-devel] [Question] About depcreated
elm_object_item_object_get API
   
Yes, theme dependency can be the problem.
But there eixst pros and cons.
   
  
   Yes there exists pros and cons.
   But *MY* policy is not to export the internal objects as much as
  possible.
   Managing one objects in the two different places deosn't look good.
   Once we export the internal objects, people will abuse it with the very
   high possibility.
  
  
   
If application make customized winset theme, application can add
  programs
or parts or something for their own purpose and do more things on
 theme
   and
do not need to make their own customized winset. In this situation,
  there
is no need to change the elementary and default theme.  Application
   install
their app just with their extension theme.
   
In this case, application should know the edje/evas APIs (or
 elementary
should wrap it but this is almost impossible, I think). But I think
  that
application should know those and use those for making perferctly
 there
   own
purpose.
   
Otherwise, application should make their own edje and add code for
manupulate it (make own winset).  In this case, application also know
edje/evas APIs.
   
  
   Yes this is what I recommend as of now.
   If you want to handle/manage something, you create it. Do not take
  other's
   possessions. There must be a conflict.
   Genlist supports full item style for this kind of use cases.
  
  
   
Theme is already exported for applications. So application can make
  their
own theme.
And I requested to export theme (edje), currently it is hidden.
   
   
   
2013/4/4 ChunEon Park hermet@naver.com
   
 Maybe it can be... but i don't like those kinds of apis since it
 will
 result in the increase of the theme compatibility break.

   
  
   Agreed.
  
  
 For your information,
 Users even can't decide the ctxpopup size.
 Conceptually, the ctxpopup itself decides the size and position by
itself.

 
 -Regards, Hermet-

 -Original Message-
 From: Bluezeryohpowel@gmail.com
 To:

Re: [E-devel] [EGIT] [core/efl] master 01/03: Install CMake configs to the proper path

2013-04-16 Thread ryuan Choi

Oops. It's really nice.

Thank you so much


2013/4/16 Daniel Juyung Seo seojuyu...@gmail.com

 AWESOME!! This works perfectly. Now I can get rid of some CMAKE_PREFIX_PATH
 definition from my build script!
 Today is awesome day :)

 Daniel Juyung Seo (SeoZ)


 On Tue, Apr 16, 2013 at 5:08 PM, Igor Murzov - Enlightenment Git 
 no-re...@enlightenment.org wrote:

  garik pushed a commit to branch master.
 
  commit 380df67f84f5be3ca9f93846a1d9dd76a608a68b
  Author: Igor Murzov e-m...@date.by
  Date:   Sat Apr 13 23:53:49 2013 +0400
 
  Install CMake configs to the proper path
  ---
   Makefile.am | 49 ++---
   1 file changed, 34 insertions(+), 15 deletions(-)
 
  diff --git a/Makefile.am b/Makefile.am
  index 9b2072c..d8d1131 100644
  --- a/Makefile.am
  +++ b/Makefile.am
  @@ -234,21 +234,40 @@ pkgconfig_DATA += pc/ecore-audio.pc
   endif
 
   # Cmake configs:
  -cmakeconfigdir = ${prefix}/share/cmake/Modules/
  -cmakeconfig_DATA = cmakeconfig/EinaConfig.cmake \
  -  cmakeconfig/EinaConfigVersion.cmake \
  -  cmakeconfig/EoConfig.cmake \
  -  cmakeconfig/EoConfigVersion.cmake \
  -  cmakeconfig/EetConfig.cmake \
  -  cmakeconfig/EetConfigVersion.cmake \
  -  cmakeconfig/EvasConfig.cmake \
  -  cmakeconfig/EvasConfigVersion.cmake \
  -  cmakeconfig/EcoreConfig.cmake \
  -  cmakeconfig/EcoreConfigVersion.cmake \
  -  cmakeconfig/EdjeConfig.cmake \
  -  cmakeconfig/EdjeConfigVersion.cmake \
  -  cmakeconfig/EfreetConfig.cmake \
  -  cmakeconfig/EfreetConfigVersion.cmake
  +eina_cmakeconfigdir = $(libdir)/cmake/Eina/
  +eina_cmakeconfig_DATA = \
  +cmakeconfig/EinaConfig.cmake \
  +cmakeconfig/EinaConfigVersion.cmake
  +
  +eo_cmakeconfigdir = $(libdir)/cmake/Eo/
  +eo_cmakeconfig_DATA = \
  +cmakeconfig/EoConfig.cmake \
  +cmakeconfig/EoConfigVersion.cmake
  +
  +eet_cmakeconfigdir = $(libdir)/cmake/Eet/
  +eet_cmakeconfig_DATA = \
  +cmakeconfig/EetConfig.cmake \
  +cmakeconfig/EetConfigVersion.cmake
  +
  +evas_cmakeconfigdir =  $(libdir)/cmake/Evas/
  +evas_cmakeconfig_DATA = \
  +cmakeconfig/EvasConfig.cmake \
  +cmakeconfig/EvasConfigVersion.cmake
  +
  +ecore_cmakeconfigdir =  $(libdir)/cmake/Ecore/
  +ecore_cmakeconfig_DATA = \
  +cmakeconfig/EcoreConfig.cmake \
  +cmakeconfig/EcoreConfigVersion.cmake
  +
  +edje_cmakeconfigdir =  $(libdir)/cmake/Edje/
  +edje_cmakeconfig_DATA = \
  +cmakeconfig/EdjeConfig.cmake \
  +cmakeconfig/EdjeConfigVersion.cmake
  +
  +efreet_cmakeconfigdir =  $(libdir)/cmake/Efreet/
  +efreet_cmakeconfig_DATA = \
  +cmakeconfig/EfreetConfig.cmake \
  +cmakeconfig/EfreetConfigVersion.cmake
 
   # D-Bus services:
 
 
  --
 
 
 
 --
  Precog is a next-generation analytics platform capable of advanced
  analytics on semi-structured data. The platform includes APIs for
 building
  apps and a phenomenal toolset for data science. Developers can use
  our toolset for easy data analysis  visualization. Get a free account!
  http://www2.precog.com/precogplatform/slashdotnewsletter
 

 --
 Precog is a next-generation analytics platform capable of advanced
 analytics on semi-structured data. The platform includes APIs for building
 apps and a phenomenal toolset for data science. Developers can use
 our toolset for easy data analysis  visualization. Get a free account!
 http://www2.precog.com/precogplatform/slashdotnewsletter
 ___
 enlightenment-devel mailing list
 enlightenment-devel@lists.sourceforge.net
 https://lists.sourceforge.net/lists/listinfo/enlightenment-devel

--
Precog is a next-generation analytics platform capable of advanced
analytics on semi-structured data. The platform includes APIs for building
apps and a phenomenal toolset for data science. Developers can use
our toolset for easy data analysis  visualization. Get a free account!
http://www2.precog.com/precogplatform/slashdotnewsletter
___
enlightenment-devel mailing list
enlightenment-devel@lists.sourceforge.net
https://lists.sourceforge.net/lists/listinfo/enlightenment-devel

[E-devel] edbus - Annoying inconsistency

Hey guys,

How are you? I just noticed that 'edbus' installs headers under:
prefix/include/edbus-1
while using the pkgconfig name 'edbus2'
the library name 'libedbus2'
and an header called 'EDbus.h'.

Luckily for us, it's not too late to fix it. Only question, what should 
we change it to? Everywhere with the 2 or everywhere without the 2?

--
Tom.

--
Precog is a next-generation analytics platform capable of advanced
analytics on semi-structured data. The platform includes APIs for building
apps and a phenomenal toolset for data science. Developers can use
our toolset for easy data analysis  visualization. Get a free account!
http://www2.precog.com/precogplatform/slashdotnewsletter
___
enlightenment-devel mailing list
enlightenment-devel@lists.sourceforge.net
https://lists.sourceforge.net/lists/listinfo/enlightenment-devel

[E-devel] State of E

People will likely want to avoid updating their E18 for the time being until a 
mildly crippling pixmap bug has been resolved.

To anyone considering not reading this message, don't be a hero.

--
Precog is a next-generation analytics platform capable of advanced
analytics on semi-structured data. The platform includes APIs for building
apps and a phenomenal toolset for data science. Developers can use
our toolset for easy data analysis  visualization. Get a free account!
http://www2.precog.com/precogplatform/slashdotnewsletter
___
enlightenment-devel mailing list
enlightenment-devel@lists.sourceforge.net
https://lists.sourceforge.net/lists/listinfo/enlightenment-devel

Re: [E-devel] edbus - Annoying inconsistency

On Tue, Apr 16, 2013 at 11:12 AM, Tom Hacohen tom.haco...@samsung.com wrote:
 Hey guys,

 How are you? I just noticed that 'edbus' installs headers under:
 prefix/include/edbus-1
 while using the pkgconfig name 'edbus2'
 the library name 'libedbus2'
 and an header called 'EDbus.h'.

 Luckily for us, it's not too late to fix it. Only question, what should we
 change it to? Everywhere with the 2 or everywhere without the 2?

E_DBus installed edbus.pc and libedbus.so - that's why we needed to
put a 2 in the name.

However the name is edbus and has no relations with the previous
library. It's not a new version, it's a complete different one. So
appending a 2 in headers and include path is the wrong thing to do.

If ebus was available I would vote for that. It seems we are short  in
options for names.

Lucas De Marchi

--
Precog is a next-generation analytics platform capable of advanced
analytics on semi-structured data. The platform includes APIs for building
apps and a phenomenal toolset for data science. Developers can use
our toolset for easy data analysis  visualization. Get a free account!
http://www2.precog.com/precogplatform/slashdotnewsletter
___
enlightenment-devel mailing list
enlightenment-devel@lists.sourceforge.net
https://lists.sourceforge.net/lists/listinfo/enlightenment-devel

Re: [E-devel] [EGIT] [core/enlightenment] master 01/01: add zoomap so we get back our zooming comp client things with.. menu for today. popups need doing... anything that bypasses the comp native surf

On Tue, 16 Apr 2013 14:15:56 +0200 Leif Middelschulte
leif.middelschu...@gmail.com said:

 http://idahofalls.webimpakt.com/wwwroot/userfiles/images/pr/zoo/mapforweb2.gif
 
 Let's meet at the bar aka. drinking fountain later. 

i plan on hanging out with the goats. how can you been an evenining with a
couple of goats! :)

 -- 
 Leif
 
 
 Am Dienstag, 16. April 2013 um 13:07 schrieb Carsten Haitzler - Enlightenment
 Git:
 
  raster pushed a commit to branch master.
  
  commit e0e513f457866042c0099dd6927e1784738bdb91
  Author: Carsten Haitzler (Rasterman) ras...@rasterman.com
  (mailto:ras...@rasterman.com) Date: Tue Apr 16 20:06:47 2013 +0900
  
  add zoomap so we get back our zooming comp client things with.. menu
  for today. popups need doing... anything that bypasses the comp native
  surface stuff needs this. :)
  ---
  src/bin/Makefile.am (http://Makefile.am) | 2 +
  src/bin/e_includes.h | 1 +
  src/bin/e_menu.c | 29 -
  src/bin/e_menu.h | 1 +
  src/bin/e_zoomap.c | 323 +++
  src/bin/e_zoomap.h | 17 +++
  6 files changed, 367 insertions(+), 6 deletions(-)
  
  diff --git a/src/bin/Makefile.am (http://Makefile.am) b/src/bin/Makefile.am
  (http://Makefile.am) index 675bccf..710112b 100644
  --- a/src/bin/Makefile.am (http://Makefile.am)
  +++ b/src/bin/Makefile.am (http://Makefile.am)
  @@ -201,6 +201,7 @@ e_win.h \
  e_xinerama.h \
  e_xkb.h \
  e_xsettings.h \
  +e_zoomap.h \
  e_zone.h
  
  if HAVE_WAYLAND_CLIENTS
  @@ -371,6 +372,7 @@ e_win.c \
  e_xinerama.c \
  e_xkb.c \
  e_xsettings.c \
  +e_zoomap.c \
  e_zone.c \
  $(ENLIGHTENMENTHEADERS)
  
  diff --git a/src/bin/e_includes.h b/src/bin/e_includes.h
  index 5863b22..70b5ceb 100644
  --- a/src/bin/e_includes.h
  +++ b/src/bin/e_includes.h
  @@ -55,6 +55,7 @@
  #include e_remember.h
  #include e_win.h
  #include e_pan.h
  +#include e_zoomap.h
  #include e_dialog.h
  #include e_configure.h
  #include e_configure_option.h
  diff --git a/src/bin/e_menu.c b/src/bin/e_menu.c
  index e333eb8..0c37fad 100644
  --- a/src/bin/e_menu.c
  +++ b/src/bin/e_menu.c
  @@ -1170,6 +1170,8 @@ e_menu_idler_before(void)
  m-prev.h = m-cur.h;
  w = m-cur.w;
  h = m-cur.h;
  + evas_object_resize(m-bg_object, w, h);
  + evas_object_resize(m-bg_object_wrap, w, h);
  if (m-cw)
  e_comp_win_resize(m-cw, w, h);
  }
  @@ -1210,9 +1212,10 @@ e_menu_idler_before(void)
  m-prev.visible = m-cur.visible;
  if (!m-cw)
  {
  - evas_object_move(m-bg_object, m-cur.x, m-cur.y);
  evas_object_resize(m-bg_object, m-cur.w, m-cur.h);
  - E_LAYER_SET(m-bg_object, E_COMP_CANVAS_LAYER_MENU);
  + evas_object_move(m-bg_object_wrap, m-cur.x, m-cur.y);
  + evas_object_resize(m-bg_object_wrap, m-cur.w, m-cur.h);
  + E_LAYER_SET(m-bg_object_wrap, E_COMP_CANVAS_LAYER_MENU);
  }
  e_comp_win_show(m-cw);
  }
  @@ -1284,7 +1287,8 @@ _e_menu_free(E_Menu *m)
  if (m-parent_item)
  m-parent_item-submenu = NULL;
  /* del callback causes this to unrealize the menu */
  - if (m-bg_object) evas_object_del(m-bg_object);
  + if (m-bg_object_wrap) evas_object_del(m-bg_object_wrap);
  + m-bg_object_wrap = NULL;
  EINA_LIST_FOREACH_SAFE(m-items, l, l_next, mi)
  e_object_del(E_OBJECT(mi));
  if (m-in_active_list)
  @@ -1393,6 +1397,8 @@ _e_menu_del_cb(void *data, Evas *e EINA_UNUSED,
  Evas_Object *obj EINA_UNUSED, vo E_Menu *m = data;
  
  m-bg_object = NULL;
  + evas_object_del(m-bg_object_wrap);
  + m-bg_object_wrap = NULL;
  _e_menu_unrealize(m);
  }
  
  @@ -1688,6 +1694,7 @@ _e_menu_realize(E_Menu *m)
  Evas_Object *o;
  Eina_List *l;
  E_Menu_Item *mi;
  + const char *s;
  
  if (m-realized || (!m-items)) return;
  m-realized = 1;
  @@ -1701,9 +1708,6 @@ _e_menu_realize(E_Menu *m)
  o = edje_object_add(m-evas);
  evas_object_event_callback_add(o, EVAS_CALLBACK_DEL, _e_menu_del_cb, m);
  m-bg_object = o;
  - evas_object_name_set(o, menu-bg_object);
  - evas_object_data_set(o, e_menu, m);
  - evas_object_data_set(o, eobj, m);
  e_theme_edje_object_set(o, base/theme/menus,
  e/widgets/menu/default/background); if (m-header.title)
  {
  @@ -1712,6 +1716,17 @@ _e_menu_realize(E_Menu *m)
  edje_object_message_signal_process(o);
  }
  
  + o = e_zoomap_add(m-evas);
  + evas_object_name_set(o, menu-bg_object_wrap);
  + evas_object_data_set(o, e_menu, m);
  + evas_object_data_set(o, eobj, m);
  + m-bg_object_wrap = o;
  + s = edje_object_data_get(m-bg_object, argb);
  + if (!s) s = edje_object_data_get(m-bg_object, shaped);
  + if ((s)  (s[0] == '1')) e_zoomap_solid_set(o, EINA_FALSE);
  + else e_zoomap_solid_set(o, EINA_TRUE);
  + e_zoomap_child_set(o, m-bg_object);
  +
  o = e_box_add(m-evas);
  evas_object_name_set(o, menu-container_object);
  m-container_object = o;
  @@ -1981,6 +1996,8 @@ _e_menu_unrealize(E_Menu *m)
  m-header.icon = NULL;
  if (m-bg_object) evas_object_del(m-bg_object);
  m-bg_object = NULL;
  + if (m-bg_object_wrap) evas_object_del(m-bg_object_wrap);
  + m-bg_object_wrap = NULL;
  if (m-container_object)

Re: [E-devel] State of E

2013-04-16 Thread Leif Middelschulte

For all those who have pulled latest revision and are looking for a working 
revision:

Using revision 93f8f16bfde302f13282be49a7d912bcfa19df04 should to the trick :-) 

-- 
Leif


Am Dienstag, 16. April 2013 um 16:26 schrieb Michael Blumenkrantz:

 People will likely want to avoid updating their E18 for the time being until 
 a mildly crippling pixmap bug has been resolved.
 
 To anyone considering not reading this message, don't be a hero.
 
 --
 Precog is a next-generation analytics platform capable of advanced
 analytics on semi-structured data. The platform includes APIs for building
 apps and a phenomenal toolset for data science. Developers can use
 our toolset for easy data analysis  visualization. Get a free account!
 http://www2.precog.com/precogplatform/slashdotnewsletter
 ___
 enlightenment-devel mailing list
 enlightenment-devel@lists.sourceforge.net 
 (mailto:enlightenment-devel@lists.sourceforge.net)
 https://lists.sourceforge.net/lists/listinfo/enlightenment-devel
 
 


--
Precog is a next-generation analytics platform capable of advanced
analytics on semi-structured data. The platform includes APIs for building
apps and a phenomenal toolset for data science. Developers can use
our toolset for easy data analysis  visualization. Get a free account!
http://www2.precog.com/precogplatform/slashdotnewsletter
___
enlightenment-devel mailing list
enlightenment-devel@lists.sourceforge.net
https://lists.sourceforge.net/lists/listinfo/enlightenment-devel

Re: [E-devel] edbus - Annoying inconsistency

On 16/04/13 15:26, Lucas De Marchi wrote:
 On Tue, Apr 16, 2013 at 11:12 AM, Tom Hacohen tom.haco...@samsung.com wrote:
 Hey guys,

 How are you? I just noticed that 'edbus' installs headers under:
 prefix/include/edbus-1
 while using the pkgconfig name 'edbus2'
 the library name 'libedbus2'
 and an header called 'EDbus.h'.

 Luckily for us, it's not too late to fix it. Only question, what should we
 change it to? Everywhere with the 2 or everywhere without the 2?

 E_DBus installed edbus.pc and libedbus.so - that's why we needed to
 put a 2 in the name.

 However the name is edbus and has no relations with the previous
 library. It's not a new version, it's a complete different one. So
 appending a 2 in headers and include path is the wrong thing to do.

 If ebus was available I would vote for that. It seems we are short  in
 options for names.

 Lucas De Marchi



I understand putting the 2 was needed, but if we add it once, we should 
add it everywhere. It's annoyingly inconsistent.

Why did you choose 2 then? Why not 4? Or A? You chose 2 because it's the 
best explanation, as it is, although mostly unrelated, e_dbus v2. And if 
that's good enough there, it's good enough everywhere.

In conclusion, I see no harm in using edbus2 all around the place.

--
Tom.

--
Precog is a next-generation analytics platform capable of advanced
analytics on semi-structured data. The platform includes APIs for building
apps and a phenomenal toolset for data science. Developers can use
our toolset for easy data analysis  visualization. Get a free account!
http://www2.precog.com/precogplatform/slashdotnewsletter
___
enlightenment-devel mailing list
enlightenment-devel@lists.sourceforge.net
https://lists.sourceforge.net/lists/listinfo/enlightenment-devel

Re: [E-devel] State of E

2013-04-16 Thread Leif Middelschulte

Am Dienstag, 16. April 2013 um 16:38 schrieb Leif Middelschulte:
 For all those who have pulled latest revision and are looking for a working 
 revision:
 
 Using revision 93f8f16bfde302f13282be49a7d912bcfa19df04 should to the trick 
 :-)
Sorry for the wrong info I forwarded without prior confirmation.

26f26283a4da25f08a385e7a0143434218318044 does work (including window resizing). 

--
Leif
 
 -- 
 Leif
 
 
 Am Dienstag, 16. April 2013 um 16:26 schrieb Michael Blumenkrantz:
 
  People will likely want to avoid updating their E18 for the time being 
  until a mildly crippling pixmap bug has been resolved.
  
  To anyone considering not reading this message, don't be a hero.
  
  --
  Precog is a next-generation analytics platform capable of advanced
  analytics on semi-structured data. The platform includes APIs for building
  apps and a phenomenal toolset for data science. Developers can use
  our toolset for easy data analysis  visualization. Get a free account!
  http://www2.precog.com/precogplatform/slashdotnewsletter
  ___
  enlightenment-devel mailing list
  enlightenment-devel@lists.sourceforge.net 
  (mailto:enlightenment-devel@lists.sourceforge.net)
  https://lists.sourceforge.net/lists/listinfo/enlightenment-devel
  
  
  
 
 

--
Precog is a next-generation analytics platform capable of advanced
analytics on semi-structured data. The platform includes APIs for building
apps and a phenomenal toolset for data science. Developers can use
our toolset for easy data analysis  visualization. Get a free account!
http://www2.precog.com/precogplatform/slashdotnewsletter
___
enlightenment-devel mailing list
enlightenment-devel@lists.sourceforge.net
https://lists.sourceforge.net/lists/listinfo/enlightenment-devel

Re: [E-devel] edbus - Annoying inconsistency

On Tue, Apr 16, 2013 at 11:54 AM, Tom Hacohen tom.haco...@samsung.com wrote:
 On 16/04/13 15:26, Lucas De Marchi wrote:

 On Tue, Apr 16, 2013 at 11:12 AM, Tom Hacohen tom.haco...@samsung.com
 wrote:

 Hey guys,

 How are you? I just noticed that 'edbus' installs headers under:
 prefix/include/edbus-1
 while using the pkgconfig name 'edbus2'
 the library name 'libedbus2'
 and an header called 'EDbus.h'.

 Luckily for us, it's not too late to fix it. Only question, what should
 we
 change it to? Everywhere with the 2 or everywhere without the 2?


 E_DBus installed edbus.pc and libedbus.so - that's why we needed to
 put a 2 in the name.

 However the name is edbus and has no relations with the previous
 library. It's not a new version, it's a complete different one. So
 appending a 2 in headers and include path is the wrong thing to do.

 If ebus was available I would vote for that. It seems we are short  in
 options for names.

 Lucas De Marchi



 I understand putting the 2 was needed, but if we add it once, we should add
 it everywhere. It's annoyingly inconsistent.

 Why did you choose 2 then? Why not 4? Or A? You chose 2 because it's the
 best explanation, as it is, although mostly unrelated, e_dbus v2. And if

because it was not me is the best explanation.

 that's good enough there, it's good enough everywhere.

 In conclusion, I see no harm in using edbus2 all around the place.

it's not good enough there, even less everywhere.  The visible part to
the developer is the include path and the namespace. And these should
reflect the library name. If anything change everything to something
else, but don't put the 2 there.

Lucas De Marchi

--
Precog is a next-generation analytics platform capable of advanced
analytics on semi-structured data. The platform includes APIs for building
apps and a phenomenal toolset for data science. Developers can use
our toolset for easy data analysis  visualization. Get a free account!
http://www2.precog.com/precogplatform/slashdotnewsletter
___
enlightenment-devel mailing list
enlightenment-devel@lists.sourceforge.net
https://lists.sourceforge.net/lists/listinfo/enlightenment-devel

Re: [E-devel] edbus - Annoying inconsistency

On 16/04/13 16:01, Lucas De Marchi wrote:
 On Tue, Apr 16, 2013 at 11:54 AM, Tom Hacohen tom.haco...@samsung.com wrote:
 On 16/04/13 15:26, Lucas De Marchi wrote:

 On Tue, Apr 16, 2013 at 11:12 AM, Tom Hacohen tom.haco...@samsung.com
 wrote:

 Hey guys,

 How are you? I just noticed that 'edbus' installs headers under:
 prefix/include/edbus-1
 while using the pkgconfig name 'edbus2'
 the library name 'libedbus2'
 and an header called 'EDbus.h'.

 Luckily for us, it's not too late to fix it. Only question, what should
 we
 change it to? Everywhere with the 2 or everywhere without the 2?


 E_DBus installed edbus.pc and libedbus.so - that's why we needed to
 put a 2 in the name.

 However the name is edbus and has no relations with the previous
 library. It's not a new version, it's a complete different one. So
 appending a 2 in headers and include path is the wrong thing to do.

 If ebus was available I would vote for that. It seems we are short  in
 options for names.

 Lucas De Marchi



 I understand putting the 2 was needed, but if we add it once, we should add
 it everywhere. It's annoyingly inconsistent.

 Why did you choose 2 then? Why not 4? Or A? You chose 2 because it's the
 best explanation, as it is, although mostly unrelated, e_dbus v2. And if

 because it was not me is the best explanation.

 that's good enough there, it's good enough everywhere.

 In conclusion, I see no harm in using edbus2 all around the place.

 it's not good enough there, even less everywhere.  The visible part to
 the developer is the include path and the namespace. And these should
 reflect the library name. If anything change everything to something
 else, but don't put the 2 there.


Let's wait to see what others think, as we disagree about the source of 
the problem, and that's: is this a major inconsistency that should be 
solved.

Anyhow, regarding different names: I suggest edbus2.

--
Tom.


--
Precog is a next-generation analytics platform capable of advanced
analytics on semi-structured data. The platform includes APIs for building
apps and a phenomenal toolset for data science. Developers can use
our toolset for easy data analysis  visualization. Get a free account!
http://www2.precog.com/precogplatform/slashdotnewsletter
___
enlightenment-devel mailing list
enlightenment-devel@lists.sourceforge.net
https://lists.sourceforge.net/lists/listinfo/enlightenment-devel

Re: [E-devel] edbus - Annoying inconsistency

2013-04-16 Thread michael . blumenkrantz

Tom Hacohen tom.haco...@samsung.com wrote:

On 16/04/13 16:01, Lucas De Marchi wrote:
 On Tue, Apr 16, 2013 at 11:54 AM, Tom Hacohen
tom.haco...@samsung.com wrote:
 On 16/04/13 15:26, Lucas De Marchi wrote:

 On Tue, Apr 16, 2013 at 11:12 AM, Tom Hacohen
tom.haco...@samsung.com
 wrote:

 Hey guys,

 How are you? I just noticed that 'edbus' installs headers under:
 prefix/include/edbus-1
 while using the pkgconfig name 'edbus2'
 the library name 'libedbus2'
 and an header called 'EDbus.h'.

 Luckily for us, it's not too late to fix it. Only question, what
should
 we
 change it to? Everywhere with the 2 or everywhere without the 2?


 E_DBus installed edbus.pc and libedbus.so - that's why we needed to
 put a 2 in the name.

 However the name is edbus and has no relations with the previous
 library. It's not a new version, it's a complete different one. So
 appending a 2 in headers and include path is the wrong thing to
do.

 If ebus was available I would vote for that. It seems we are short 
in
 options for names.

 Lucas De Marchi



 I understand putting the 2 was needed, but if we add it once, we
should add
 it everywhere. It's annoyingly inconsistent.

 Why did you choose 2 then? Why not 4? Or A? You chose 2 because it's
the
 best explanation, as it is, although mostly unrelated, e_dbus v2.
And if

 because it was not me is the best explanation.

 that's good enough there, it's good enough everywhere.

 In conclusion, I see no harm in using edbus2 all around the place.

 it's not good enough there, even less everywhere.  The visible part
to
 the developer is the include path and the namespace. And these should
 reflect the library name. If anything change everything to something
 else, but don't put the 2 there.


Let's wait to see what others think, as we disagree about the source of

the problem, and that's: is this a major inconsistency that should be 
solved.

Anyhow, regarding different names: I suggest edbus2.

--
Tom.


--
Precog is a next-generation analytics platform capable of advanced
analytics on semi-structured data. The platform includes APIs for
building
apps and a phenomenal toolset for data science. Developers can use
our toolset for easy data analysis  visualization. Get a free account!
http://www2.precog.com/precogplatform/slashdotnewsletter
___
enlightenment-devel mailing list
enlightenment-devel@lists.sourceforge.net
https://lists.sourceforge.net/lists/listinfo/enlightenment-devel

I agree with whatever demarchi says. 
--
Precog is a next-generation analytics platform capable of advanced
analytics on semi-structured data. The platform includes APIs for building
apps and a phenomenal toolset for data science. Developers can use
our toolset for easy data analysis  visualization. Get a free account!
http://www2.precog.com/precogplatform/slashdotnewsletter
___
enlightenment-devel mailing list
enlightenment-devel@lists.sourceforge.net
https://lists.sourceforge.net/lists/listinfo/enlightenment-devel

Re: [E-devel] edbus - Annoying inconsistency

On Tue, Apr 16, 2013 at 4:13 PM,  michael.blumenkra...@gmail.com wrote:
 Tom Hacohen tom.haco...@samsung.com wrote:

 On 16/04/13 16:01, Lucas De Marchi wrote:

 On Tue, Apr 16, 2013 at 11:54 AM, Tom Hacohen tom.haco...@samsung.com
 wrote:

 On 16/04/13 15:26, Lucas De Marchi wrote:

 On Tue, Apr 16, 2013 at 11:12 AM, Tom Hacohen tom.haco...@samsung.com
 wrote:

 Hey guys,

 How are you? I just noticed that 'edbus' installs headers under:
 prefix/include/edbus-1
 while using the
 pkgconfig name 'edbus2'
 the library name 'libedbus2'
 and an header called 'EDbus.h'.

 Luckily for us, it's not too late to fix it. Only question, what
 should
 we
 change it to? Everywhere with the 2 or everywhere without the 2?



 E_DBus installed edbus.pc and libedbus.so - that's why we needed to
 put a 2 in the name.

 However the name is edbus and has no relations with the previous
 library. It's not a new version, it's a complete different one. So
 appending a 2 in headers and include path is the wrong thing to do.

 If ebus was available I would vote for that. It seems we are short  in
 options for names.

 Lucas De Marchi




 I understand putting the 2 was needed, but if we add it once, we should
 add
 it everywhere. It's annoyingly inconsistent.

 Why did you choose 2 then? Why not 4? Or A? You chose 2 because
 it's the
 best explanation, as it is, although mostly unrelated, e_dbus v2. And if


 because it was not me is the best explanation.

 that's good enough there, it's good enough everywhere.

 In conclusion, I see no harm in using edbus2 all around the place.


 it's not good enough there, even less everywhere.  The visible part to
 the developer is the include path and the namespace. And these should
 reflect the library name. If anything change everything to something
 else, but don't put the 2 there.



 Let's wait to see what others think, as we disagree about the source of
 the problem, and that's: is this a major inconsistency that should be
 solved.

 Anyhow, regarding different names: I suggest edbus2.

 --
 Tom.


 

 Precog is a next-generation analytics platform capable of advanced
 analytics on semi-structured data. The platform includes APIs for building
 apps and a phenomenal toolset for data science. Developers can use
 our toolset for easy data analysis  visualization. Get a free account!
 http://www2.precog.com/precogplatform/slashdotnewsletter
 

 enlightenment-devel mailing list
 enlightenment-devel@lists.sourceforge.net
 https://lists.sourceforge.net/lists/listinfo/enlightenment-devel


 I agree with whatever demarchi says.

I disagree.


oh, wait...

Lucas De Marchi

--
Precog is a next-generation analytics platform capable of advanced
analytics on semi-structured data. The platform includes APIs for building
apps and a phenomenal toolset for data science. Developers can use
our toolset for easy data analysis  visualization. Get a free account!
http://www2.precog.com/precogplatform/slashdotnewsletter
___
enlightenment-devel mailing list
enlightenment-devel@lists.sourceforge.net
https://lists.sourceforge.net/lists/listinfo/enlightenment-devel

Re: [E-devel] new bugs. in e (border releated)

2013-04-16 Thread Gustavo Lima Chaves

* Michael Blumenkrantz michael.blumenkra...@gmail.com [2013-04-16 11:15:30 
+0100]:

 Okay, fixed now

Not for me :/

 
 
 On Tue, Apr 16, 2013 at 9:33 AM, Carsten Haitzler ras...@rasterman.comwrote:
 
  On Tue, 16 Apr 2013 07:53:11 +0100 Michael Blumenkrantz
  michael.blumenkra...@gmail.com said:
 
   This sounds like a localized config bug. Try changing your desk flip
   transition.
 
  all transitions except nonebreak here. i tried all of them.
 
  
  
   On Tue, Apr 16, 2013 at 3:51 AM, Carsten Haitzler ras...@rasterman.com
  wrote:
  
On Tue, 16 Apr 2013 11:46:04 +0900 Carsten Haitzler (The Rasterman)
ras...@rasterman.com said:
   
bug #7:
   
very often mouse events just stop working (going to clients). mouse
press/move/release/wheel etc. after flipping virtual desktops a few
  times
or alt
+left dragging a window about.
   
 On Tue, 16 Apr 2013 09:37:54 +0900 Carsten Haitzler (The Rasterman)
 ras...@rasterman.com said:

 bug #6:

 xdaliclock -transparent is nicely broken (shape mask is applied and
offset by
 title height). this used to work. it's one of my comp shape test
  cases.
:)

  On Tue, 16 Apr 2013 09:27:21 +0900 Carsten Haitzler (The Rasterman)
  ras...@rasterman.com said:
 
  bug #4:
 
  moving windows around no matter how small or large they are is now
really
  jerky... or it is almost all of the time... terminology seems to be
smooth
  all the time, but xterm seems to be 50/50 and anything else
  (sylpheed,
  chromium, gedit, kcachegrind...) is jerky as all hell. it seems to
  be a
  timing issue of some sort...
 
  bug #5:
 
  resizing of a window produces swimming of content as it is
  stretched
to
  fill the new border UNTIL the client can render an updated set of
content
  (and then fix it up). the effect basically is that content
  squashes and
  expands back and forth during resizing as opposed to remaining
unchanged
  in scaling. comp used to address this nicely by enforcing e_comp
  size
==
  size of pixmap dimensions. :)
 
   bug #1:
  
   1. start terminal (xterm/terminology - any window actually)
   2. type cmd: sleep 5; exit
   3. alt+left click and drag to move window. keep moving it until
window
   closes/exits
   4. release mouse.
   5. segfault happiness! :)
  
   (a quick look says to me action_border is still pointing to a
  border
that
   has now been deleted - crashdump attached).
  
   bug #2:
  
   something has happened to the shadow in comp... it's now wrong.
  it
has
   moved down from the title:
  
  
   
  http://www.enlightenment.org/ss/display.php?image=e-516c92a8841755.22448970.png
  
   vs old:
  
  
   
  http://www.enlightenment.org/ss/display.php?image=e-515c83bebe2212.05354344.jpg
  
   look carefully at shadow on left/right size of the window near
  the
   titlebar top. now the shadow begins much lower.. just below
  client
start.
  
   bug #3.
  
   double-click a window to shade it. it USED to just become a nice
simple
   titlebar... no longer. :)
  
   --
   - Codito, ergo sum - I code, therefore I am
--
   The Rasterman (Carsten Haitzler)ras...@rasterman.com
  
 
 
  --
  - Codito, ergo sum - I code, therefore I am
--
  The Rasterman (Carsten Haitzler)ras...@rasterman.com
 
 
 
   
  --
  Precog is a next-generation analytics platform capable of advanced
  analytics on semi-structured data. The platform includes APIs for
building
  apps and a phenomenal toolset for data science. Developers can use
  our toolset for easy data analysis  visualization. Get a free
  account!
  http://www2.precog.com/precogplatform/slashdotnewsletter
  ___
  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



   
  --
 Precog is a next-generation analytics platform capable of advanced
 analytics on semi-structured data. The platform includes APIs for
building
 apps and a phenomenal toolset for data science. Developers can use
 our toolset for easy data analysis  visualization. Get a free
  account!
 http://www2.precog.com/precogplatform/slashdotnewsletter
 ___

Re: [E-devel] evas crashing in scale_rgba_in_to_out_clip_sample_internal() on intel Gen2 GPU

2013-04-16 Thread Bruno

Hi Raster,

On Tue, 16 April 2013 Carsten Haitzler wrote:
 On Mon, 15 Apr 2013 23:08:41 +0200 Bruno said:
 could you give more details - like it crashes immediately on start?

Yes, the crash happens during early startup of enlightenment, right after
painting background and setting up cursor.

 you are is the depth 16bpp BUT the visual is still 24bpp with 24bpp rgb
 masks? xdpyinfo will tell you this. i.e.

According to xdpyinfo, with driver prior to bisected commit root window
is 24bit but with 2.21.6 it is 15bit.

Also of interest, with 2.21.6 there are much fewer visuals listed: 15bit
default, 15bit and 32bit while last working one show lots of 24bit visuals.
Maybe the difference in number comes from the moment when xdpyinfo is called.
For 2.21.6 right after starting Xorg, as only X client.
For last good commit from xterm while running under enlightenment.

See attached xpdyinfo outputs attached to the bug report.
  https://bugs.freedesktop.org/show_bug.cgi?id=63572

 more information needed. also consider using valgrind to catch the problem
 earlier.

A valgrind trace is not very informative. It reports a bad write and later
a segfault quite a few bytes after the address of bad write.

==5361== Invalid write of size 4
==5361==at 0x40FFE90: evas_common_scale_rgba_in_to_out_clip_smooth_mmx 
(evas_scale_smooth_scaler_downx_downy.c:239)
==5361==by 0x4102D6C: evas_common_scale_rgba_in_to_out_clip_smooth 
(evas_scale_smooth.c:513)
==5361==by 0x40F9314: evas_common_rgba_image_scalecache_do 
(evas_image_scalecache.c:796)
==5361==by 0x4110153: eng_image_draw (evas_engine.c:915)
==5361==by 0x4077A6B: evas_object_image_render (evas_object_image.c:3042)
==5361==by 0x40BF34D: evas_render_mapped (evas_render.c:1160)
==5361==by 0x40C1ABE: evas_render_updates_internal (evas_render.c:1575)
==5361==by 0x41EBC83: _ecore_evas_x_render (ecore_evas_x.c:288)
==5361==by 0x41E81AD: _ecore_evas_idle_enter (ecore_evas.c:59)
==5361==by 0x4173D11: _ecore_idle_enterer_call (ecore_private.h:281)
==5361==by 0x41758F4: _ecore_main_loop_iterate_internal (ecore_main.c:1833)
==5361==by 0x4175DAE: ecore_main_loop_begin (ecore_main.c:956)
==5361==  Address 0x7f9c000 is not stack'd, malloc'd or (recently) free'd
==5361== 
==5361== 
==5361== Process terminating with default action of signal 11 (SIGSEGV)
==5361==  Access not within mapped region at address 0x7F9D770
==5361==at 0x40FFE90: evas_common_scale_rgba_in_to_out_clip_smooth_mmx 
(evas_scale_smooth_scaler_downx_downy.c:239)
==5361==by 0x4102D6C: evas_common_scale_rgba_in_to_out_clip_smooth 
(evas_scale_smooth.c:513)
==5361==by 0x40F9314: evas_common_rgba_image_scalecache_do 
(evas_image_scalecache.c:796)
==5361==by 0x4110153: eng_image_draw (evas_engine.c:915)
==5361==by 0x4077A6B: evas_object_image_render (evas_object_image.c:3042)
==5361==by 0x40BF34D: evas_render_mapped (evas_render.c:1160)
==5361==by 0x40C1ABE: evas_render_updates_internal (evas_render.c:1575)
==5361==by 0x41EBC83: _ecore_evas_x_render (ecore_evas_x.c:288)
==5361==by 0x41E81AD: _ecore_evas_idle_enter (ecore_evas.c:59)
==5361==by 0x4173D11: _ecore_idle_enterer_call (ecore_private.h:281)
==5361==by 0x41758F4: _ecore_main_loop_iterate_internal (ecore_main.c:1833)
==5361==by 0x4175DAE: ecore_main_loop_begin (ecore_main.c:956)
==5361==  If you believe this happened as a result of a stack
==5361==  overflow in your program's main thread (unlikely but
==5361==  possible), you can try to increase the size of the
==5361==  main thread stack using the --main-stacksize= flag.
==5361==  The main thread stack size used in this run was 8388608.
==5361== 

Full valgrind log attached to the bug.

Bruno

  Since intel driver xf86-video-intel-2.21.4 evas crashes, preventing
  enlightenment to start on my laptop.
  
  I bisected it to a change in intel driver:
a19da0ea517127052ae49cdd6441e8b6077ca523
sna: Reduce DefaultDepth to 16 on older chipsets
  
  Is the crash caused by wrong assumption on EFL/Enlightenment side?
  
  See freedesktop bug report
 https://bugs.freedesktop.org/show_bug.cgi?id=63572
  for enlightenment backtrace and 
 
  http://cgit.freedesktop.org/xorg/driver/xf86-video-intel/commit/?id=a19da0ea517127052ae49cdd6441e8b6077ca523
  for the triggering commit.

--
Precog is a next-generation analytics platform capable of advanced
analytics on semi-structured data. The platform includes APIs for building
apps and a phenomenal toolset for data science. Developers can use
our toolset for easy data analysis  visualization. Get a free account!
http://www2.precog.com/precogplatform/slashdotnewsletter
___
enlightenment-devel mailing list
enlightenment-devel@lists.sourceforge.net
https://lists.sourceforge.net/lists/listinfo/enlightenment-devel

Re: [E-devel] edbus - Annoying inconsistency

On Tue, 16 Apr 2013 16:24:08 -0300 Lucas De Marchi
lucas.demar...@profusion.mobi said:

 On Tue, Apr 16, 2013 at 4:13 PM,  michael.blumenkra...@gmail.com wrote:
  Tom Hacohen tom.haco...@samsung.com wrote:
 
  On 16/04/13 16:01, Lucas De Marchi wrote:
 
  On Tue, Apr 16, 2013 at 11:54 AM, Tom Hacohen tom.haco...@samsung.com
  wrote:
 
  On 16/04/13 15:26, Lucas De Marchi wrote:
 
  On Tue, Apr 16, 2013 at 11:12 AM, Tom Hacohen tom.haco...@samsung.com
  wrote:
 
  Hey guys,
 
  How are you? I just noticed that 'edbus' installs headers under:
  prefix/include/edbus-1
  while using the
  pkgconfig name 'edbus2'
  the library name 'libedbus2'
  and an header called 'EDbus.h'.
 
  Luckily for us, it's not too late to fix it. Only question, what
  should
  we
  change it to? Everywhere with the 2 or everywhere without the 2?
 
 
 
  E_DBus installed edbus.pc and libedbus.so - that's why we needed to
  put a 2 in the name.
 
  However the name is edbus and has no relations with the previous
  library. It's not a new version, it's a complete different one. So
  appending a 2 in headers and include path is the wrong thing to do.
 
  If ebus was available I would vote for that. It seems we are short  in
  options for names.
 
  Lucas De Marchi
 
 
 
 
  I understand putting the 2 was needed, but if we add it once, we should
  add
  it everywhere. It's annoyingly inconsistent.
 
  Why did you choose 2 then? Why not 4? Or A? You chose 2 because
  it's the
  best explanation, as it is, although mostly unrelated, e_dbus v2. And if
 
 
  because it was not me is the best explanation.
 
  that's good enough there, it's good enough everywhere.
 
  In conclusion, I see no harm in using edbus2 all around the place.
 
 
  it's not good enough there, even less everywhere.  The visible part to
  the developer is the include path and the namespace. And these should
  reflect the library name. If anything change everything to something
  else, but don't put the 2 there.
 
 
 
  Let's wait to see what others think, as we disagree about the source of
  the problem, and that's: is this a major inconsistency that should be
  solved.
 
  Anyhow, regarding different names: I suggest edbus2.
 
  --
  Tom.
 
 
  
 
  Precog is a next-generation analytics platform capable of advanced
  analytics on semi-structured data. The platform includes APIs for building
  apps and a phenomenal toolset for data science. Developers can use
  our toolset for easy data analysis  visualization. Get a free account!
  http://www2.precog.com/precogplatform/slashdotnewsletter
  
 
  enlightenment-devel mailing list
  enlightenment-devel@lists.sourceforge.net
  https://lists.sourceforge.net/lists/listinfo/enlightenment-devel
 
 
  I agree with whatever demarchi says.
 
 I disagree.
 
 
 oh, wait...
 
 Lucas De Marchi

:-P ... i have to say though - consistency is good. not just for devs, but
packagers and anyone looking at their filesystem. if things are named in a
consistent pattern it is easy to know what belongs where without needing
package databases or other forensics tools.

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


--
Precog is a next-generation analytics platform capable of advanced
analytics on semi-structured data. The platform includes APIs for building
apps and a phenomenal toolset for data science. Developers can use
our toolset for easy data analysis  visualization. Get a free account!
http://www2.precog.com/precogplatform/slashdotnewsletter
___
enlightenment-devel mailing list
enlightenment-devel@lists.sourceforge.net
https://lists.sourceforge.net/lists/listinfo/enlightenment-devel

Re: [E-devel] evas crashing in scale_rgba_in_to_out_clip_sample_internal() on intel Gen2 GPU

On Tue, 16 Apr 2013 22:33:26 +0200 Bruno bonbon...@internet.lu said:

 Hi Raster,
 
 On Tue, 16 April 2013 Carsten Haitzler wrote:
  On Mon, 15 Apr 2013 23:08:41 +0200 Bruno said:
  could you give more details - like it crashes immediately on start?
 
 Yes, the crash happens during early startup of enlightenment, right after
 painting background and setting up cursor.
 
  you are is the depth 16bpp BUT the visual is still 24bpp with 24bpp rgb
  masks? xdpyinfo will tell you this. i.e.
 
 According to xdpyinfo, with driver prior to bisected commit root window
 is 24bit but with 2.21.6 it is 15bit.
 
 Also of interest, with 2.21.6 there are much fewer visuals listed: 15bit
 default, 15bit and 32bit while last working one show lots of 24bit visuals.
 Maybe the difference in number comes from the moment when xdpyinfo is called.
 For 2.21.6 right after starting Xorg, as only X client.
 For last good commit from xterm while running under enlightenment.
 
 See attached xpdyinfo outputs attached to the bug report.
   https://bugs.freedesktop.org/show_bug.cgi?id=63572
 
  more information needed. also consider using valgrind to catch the problem
  earlier.
 
 A valgrind trace is not very informative. It reports a bad write and later
 a segfault quite a few bytes after the address of bad write.
 
 ==5361== Invalid write of size 4
 ==5361==at 0x40FFE90: evas_common_scale_rgba_in_to_out_clip_smooth_mmx
 (evas_scale_smooth_scaler_downx_downy.c:239) ==5361==by 0x4102D6C:
 evas_common_scale_rgba_in_to_out_clip_smooth (evas_scale_smooth.c:513)
 ==5361==by 0x40F9314: evas_common_rgba_image_scalecache_do
 (evas_image_scalecache.c:796) ==5361==by 0x4110153: eng_image_draw
 (evas_engine.c:915) ==5361==by 0x4077A6B: evas_object_image_render
 (evas_object_image.c:3042) ==5361==by 0x40BF34D: evas_render_mapped
 (evas_render.c:1160) ==5361==by 0x40C1ABE: evas_render_updates_internal
 (evas_render.c:1575) ==5361==by 0x41EBC83: _ecore_evas_x_render
 (ecore_evas_x.c:288) ==5361==by 0x41E81AD: _ecore_evas_idle_enter
 (ecore_evas.c:59) ==5361==by 0x4173D11: _ecore_idle_enterer_call
 (ecore_private.h:281) ==5361==by 0x41758F4:
 _ecore_main_loop_iterate_internal (ecore_main.c:1833) ==5361==by
 0x4175DAE: ecore_main_loop_begin (ecore_main.c:956) ==5361==  Address
 0x7f9c000 is not stack'd, malloc'd or (recently) free'd ==5361== ==5361== 
 ==5361== Process terminating with default action of signal 11 (SIGSEGV)
 ==5361==  Access not within mapped region at address 0x7F9D770
 ==5361==at 0x40FFE90: evas_common_scale_rgba_in_to_out_clip_smooth_mmx
 (evas_scale_smooth_scaler_downx_downy.c:239) ==5361==by 0x4102D6C:
 evas_common_scale_rgba_in_to_out_clip_smooth (evas_scale_smooth.c:513)
 ==5361==by 0x40F9314: evas_common_rgba_image_scalecache_do
 (evas_image_scalecache.c:796) ==5361==by 0x4110153: eng_image_draw
 (evas_engine.c:915) ==5361==by 0x4077A6B: evas_object_image_render
 (evas_object_image.c:3042) ==5361==by 0x40BF34D: evas_render_mapped
 (evas_render.c:1160) ==5361==by 0x40C1ABE: evas_render_updates_internal
 (evas_render.c:1575) ==5361==by 0x41EBC83: _ecore_evas_x_render
 (ecore_evas_x.c:288) ==5361==by 0x41E81AD: _ecore_evas_idle_enter
 (ecore_evas.c:59) ==5361==by 0x4173D11: _ecore_idle_enterer_call
 (ecore_private.h:281) ==5361==by 0x41758F4:
 _ecore_main_loop_iterate_internal (ecore_main.c:1833) ==5361==by
 0x4175DAE: ecore_main_loop_begin (ecore_main.c:956) ==5361==  If you believe
 this happened as a result of a stack ==5361==  overflow in your program's
 main thread (unlikely but ==5361==  possible), you can try to increase the
 size of the ==5361==  main thread stack using the --main-stacksize= flag.
 ==5361==  The main thread stack size used in this run was 8388608. ==5361== 
 
 Full valgrind log attached to the bug.

that's a little more useful. it gives me an actual address. the address looks
sane - ie it's writing to and invalid memory region. it'd be goot to get a
backtrace there: valgrind --db-attach=yes ... that will ask to attach gdb per
complaint - if you say yes, you can get a backtrace (bt command) and print
variables/values.

 Bruno
 
   Since intel driver xf86-video-intel-2.21.4 evas crashes, preventing
   enlightenment to start on my laptop.
   
   I bisected it to a change in intel driver:
 a19da0ea517127052ae49cdd6441e8b6077ca523
 sna: Reduce DefaultDepth to 16 on older chipsets
   
   Is the crash caused by wrong assumption on EFL/Enlightenment side?
   
   See freedesktop bug report
  https://bugs.freedesktop.org/show_bug.cgi?id=63572
   for enlightenment backtrace and 
  
   http://cgit.freedesktop.org/xorg/driver/xf86-video-intel/commit/?id=a19da0ea517127052ae49cdd6441e8b6077ca523
   for the triggering commit.
 
 --
 Precog is a next-generation analytics platform capable of advanced
 analytics on semi-structured data.

Re: [E-devel] edbus - Annoying inconsistency

On Tue, Apr 16, 2013 at 8:34 PM, Carsten Haitzler ras...@rasterman.com wrote:
 On Tue, 16 Apr 2013 16:24:08 -0300 Lucas De Marchi
 lucas.demar...@profusion.mobi said:

 On Tue, Apr 16, 2013 at 4:13 PM,  michael.blumenkra...@gmail.com wrote:
  Tom Hacohen tom.haco...@samsung.com wrote:
 
  On 16/04/13 16:01, Lucas De Marchi wrote:
 
  On Tue, Apr 16, 2013 at 11:54 AM, Tom Hacohen tom.haco...@samsung.com
  wrote:
 
  On 16/04/13 15:26, Lucas De Marchi wrote:
 
  On Tue, Apr 16, 2013 at 11:12 AM, Tom Hacohen tom.haco...@samsung.com
  wrote:
 
  Hey guys,
 
  How are you? I just noticed that 'edbus' installs headers under:
  prefix/include/edbus-1
  while using the
  pkgconfig name 'edbus2'
  the library name 'libedbus2'
  and an header called 'EDbus.h'.
 
  Luckily for us, it's not too late to fix it. Only question, what
  should
  we
  change it to? Everywhere with the 2 or everywhere without the 2?
 
 
 
  E_DBus installed edbus.pc and libedbus.so - that's why we needed to
  put a 2 in the name.
 
  However the name is edbus and has no relations with the previous
  library. It's not a new version, it's a complete different one. So
  appending a 2 in headers and include path is the wrong thing to do.
 
  If ebus was available I would vote for that. It seems we are short  in
  options for names.
 
  Lucas De Marchi
 
 
 
 
  I understand putting the 2 was needed, but if we add it once, we should
  add
  it everywhere. It's annoyingly inconsistent.
 
  Why did you choose 2 then? Why not 4? Or A? You chose 2 because
  it's the
  best explanation, as it is, although mostly unrelated, e_dbus v2. And if
 
 
  because it was not me is the best explanation.
 
  that's good enough there, it's good enough everywhere.
 
  In conclusion, I see no harm in using edbus2 all around the place.
 
 
  it's not good enough there, even less everywhere.  The visible part to
  the developer is the include path and the namespace. And these should
  reflect the library name. If anything change everything to something
  else, but don't put the 2 there.
 
 
 
  Let's wait to see what others think, as we disagree about the source of
  the problem, and that's: is this a major inconsistency that should be
  solved.
 
  Anyhow, regarding different names: I suggest edbus2.
 
  --
  Tom.
 
 
  
 
  Precog is a next-generation analytics platform capable of advanced
  analytics on semi-structured data. The platform includes APIs for building
  apps and a phenomenal toolset for data science. Developers can use
  our toolset for easy data analysis  visualization. Get a free account!
  http://www2.precog.com/precogplatform/slashdotnewsletter
  
 
  enlightenment-devel mailing list
  enlightenment-devel@lists.sourceforge.net
  https://lists.sourceforge.net/lists/listinfo/enlightenment-devel
 
 
  I agree with whatever demarchi says.

 I disagree.


 oh, wait...

 Lucas De Marchi

 :-P ... i have to say though - consistency is good. not just for devs, but
 packagers and anyone looking at their filesystem. if things are named in a
 consistent pattern it is easy to know what belongs where without needing
 package databases or other forensics tools.

Sure it's good, no doubt. Just that renaming to edbus2 for the sake of
consistency is even worse, not because of the rename, but because of
the name chosen.

ebus? anyone?

Lucas De Marchi

--
Precog is a next-generation analytics platform capable of advanced
analytics on semi-structured data. The platform includes APIs for building
apps and a phenomenal toolset for data science. Developers can use
our toolset for easy data analysis  visualization. Get a free account!
http://www2.precog.com/precogplatform/slashdotnewsletter
___
enlightenment-devel mailing list
enlightenment-devel@lists.sourceforge.net
https://lists.sourceforge.net/lists/listinfo/enlightenment-devel

Re: [E-devel] edbus - Annoying inconsistency

2013-04-16 Thread David Seikel

On Tue, 16 Apr 2013 20:57:33 -0300 Lucas De Marchi
lucas.demar...@profusion.mobi wrote:

 On Tue, Apr 16, 2013 at 8:34 PM, Carsten Haitzler
 ras...@rasterman.com wrote:
  On Tue, 16 Apr 2013 16:24:08 -0300 Lucas De Marchi
  lucas.demar...@profusion.mobi said:
 
  On Tue, Apr 16, 2013 at 4:13 PM,  michael.blumenkra...@gmail.com
  wrote:
   Tom Hacohen tom.haco...@samsung.com wrote:
  
   On 16/04/13 16:01, Lucas De Marchi wrote:
  
   On Tue, Apr 16, 2013 at 11:54 AM, Tom Hacohen
   tom.haco...@samsung.com wrote:
  
   On 16/04/13 15:26, Lucas De Marchi wrote:
  
   On Tue, Apr 16, 2013 at 11:12 AM, Tom Hacohen
   tom.haco...@samsung.com wrote:
  
   Hey guys,
  
   How are you? I just noticed that 'edbus' installs headers
   under: prefix/include/edbus-1
   while using the
   pkgconfig name 'edbus2'
   the library name 'libedbus2'
   and an header called 'EDbus.h'.
  
   Luckily for us, it's not too late to fix it. Only question,
   what should
   we
   change it to? Everywhere with the 2 or everywhere without
   the 2?
  
  
  
   E_DBus installed edbus.pc and libedbus.so - that's why we
   needed to put a 2 in the name.
  
   However the name is edbus and has no relations with the
   previous library. It's not a new version, it's a complete
   different one. So appending a 2 in headers and include
   path is the wrong thing to do.
  
   If ebus was available I would vote for that. It seems we are
   short  in options for names.
  
   Lucas De Marchi
  
  
  
  
   I understand putting the 2 was needed, but if we add it once,
   we should add
   it everywhere. It's annoyingly inconsistent.
  
   Why did you choose 2 then? Why not 4? Or A? You chose 2
   because it's the
   best explanation, as it is, although mostly unrelated, e_dbus
   v2. And if
  
  
   because it was not me is the best explanation.
  
   that's good enough there, it's good enough everywhere.
  
   In conclusion, I see no harm in using edbus2 all around the
   place.
  
  
   it's not good enough there, even less everywhere.  The visible
   part to the developer is the include path and the namespace.
   And these should reflect the library name. If anything change
   everything to something else, but don't put the 2 there.
  
  
  
   Let's wait to see what others think, as we disagree about the
   source of the problem, and that's: is this a major
   inconsistency that should be solved.
  
   Anyhow, regarding different names: I suggest edbus2.
  
   --
   Tom.
  
  
   
  
   Precog is a next-generation analytics platform capable of
   advanced analytics on semi-structured data. The platform
   includes APIs for building apps and a phenomenal toolset for
   data science. Developers can use our toolset for easy data
   analysis  visualization. Get a free account!
   http://www2.precog.com/precogplatform/slashdotnewsletter
   
  
   enlightenment-devel mailing list
   enlightenment-devel@lists.sourceforge.net
   https://lists.sourceforge.net/lists/listinfo/enlightenment-devel
  
  
   I agree with whatever demarchi says.
 
  I disagree.
 
 
  oh, wait...
 
  Lucas De Marchi
 
  :-P ... i have to say though - consistency is good. not just for
  devs, but packagers and anyone looking at their filesystem. if
  things are named in a consistent pattern it is easy to know what
  belongs where without needing package databases or other forensics
  tools.
 
 Sure it's good, no doubt. Just that renaming to edbus2 for the sake of
 consistency is even worse, not because of the rename, but because of
 the name chosen.
 
 ebus? anyone?

The name of the ebus goes 'round and 'round.

-- 
A big old stinking pile of genius that no one wants
coz there are too many silver coated monkeys in the world.


signature.asc
Description: PGP signature
--
Precog is a next-generation analytics platform capable of advanced
analytics on semi-structured data. The platform includes APIs for building
apps and a phenomenal toolset for data science. Developers can use
our toolset for easy data analysis  visualization. Get a free account!
http://www2.precog.com/precogplatform/slashdotnewsletter___
enlightenment-devel mailing list
enlightenment-devel@lists.sourceforge.net
https://lists.sourceforge.net/lists/listinfo/enlightenment-devel

Re: [E-devel] edbus - Annoying inconsistency

On Tue, 16 Apr 2013 20:57:33 -0300 Lucas De Marchi
lucas.demar...@profusion.mobi said:

 On Tue, Apr 16, 2013 at 8:34 PM, Carsten Haitzler ras...@rasterman.com
 wrote:
  On Tue, 16 Apr 2013 16:24:08 -0300 Lucas De Marchi
  lucas.demar...@profusion.mobi said:
 
  On Tue, Apr 16, 2013 at 4:13 PM,  michael.blumenkra...@gmail.com wrote:
   Tom Hacohen tom.haco...@samsung.com wrote:
  
   On 16/04/13 16:01, Lucas De Marchi wrote:
  
   On Tue, Apr 16, 2013 at 11:54 AM, Tom Hacohen tom.haco...@samsung.com
   wrote:
  
   On 16/04/13 15:26, Lucas De Marchi wrote:
  
   On Tue, Apr 16, 2013 at 11:12 AM, Tom Hacohen
   tom.haco...@samsung.com wrote:
  
   Hey guys,
  
   How are you? I just noticed that 'edbus' installs headers under:
   prefix/include/edbus-1
   while using the
   pkgconfig name 'edbus2'
   the library name 'libedbus2'
   and an header called 'EDbus.h'.
  
   Luckily for us, it's not too late to fix it. Only question, what
   should
   we
   change it to? Everywhere with the 2 or everywhere without the 2?
  
  
  
   E_DBus installed edbus.pc and libedbus.so - that's why we needed to
   put a 2 in the name.
  
   However the name is edbus and has no relations with the previous
   library. It's not a new version, it's a complete different one. So
   appending a 2 in headers and include path is the wrong thing to do.
  
   If ebus was available I would vote for that. It seems we are short
   in options for names.
  
   Lucas De Marchi
  
  
  
  
   I understand putting the 2 was needed, but if we add it once, we
   should add
   it everywhere. It's annoyingly inconsistent.
  
   Why did you choose 2 then? Why not 4? Or A? You chose 2 because
   it's the
   best explanation, as it is, although mostly unrelated, e_dbus v2. And
   if
  
  
   because it was not me is the best explanation.
  
   that's good enough there, it's good enough everywhere.
  
   In conclusion, I see no harm in using edbus2 all around the place.
  
  
   it's not good enough there, even less everywhere.  The visible part to
   the developer is the include path and the namespace. And these should
   reflect the library name. If anything change everything to something
   else, but don't put the 2 there.
  
  
  
   Let's wait to see what others think, as we disagree about the source of
   the problem, and that's: is this a major inconsistency that should be
   solved.
  
   Anyhow, regarding different names: I suggest edbus2.
  
   --
   Tom.
  
  
   
  
   Precog is a next-generation analytics platform capable of advanced
   analytics on semi-structured data. The platform includes APIs for
   building apps and a phenomenal toolset for data science. Developers can
   use our toolset for easy data analysis  visualization. Get a free
   account! http://www2.precog.com/precogplatform/slashdotnewsletter
   
  
   enlightenment-devel mailing list
   enlightenment-devel@lists.sourceforge.net
   https://lists.sourceforge.net/lists/listinfo/enlightenment-devel
  
  
   I agree with whatever demarchi says.
 
  I disagree.
 
 
  oh, wait...
 
  Lucas De Marchi
 
  :-P ... i have to say though - consistency is good. not just for devs, but
  packagers and anyone looking at their filesystem. if things are named in a
  consistent pattern it is easy to know what belongs where without needing
  package databases or other forensics tools.
 
 Sure it's good, no doubt. Just that renaming to edbus2 for the sake of
 consistency is even worse, not because of the rename, but because of
 the name chosen.
 
 ebus? anyone?
 
 Lucas De Marchi

whatever the name chosen is, if .pc file, header include dir, header names, and
library name are all consistent in terms of pattern/convention then we are
good.:)


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


--
Precog is a next-generation analytics platform capable of advanced
analytics on semi-structured data. The platform includes APIs for building
apps and a phenomenal toolset for data science. Developers can use
our toolset for easy data analysis  visualization. Get a free account!
http://www2.precog.com/precogplatform/slashdotnewsletter
___
enlightenment-devel mailing list
enlightenment-devel@lists.sourceforge.net
https://lists.sourceforge.net/lists/listinfo/enlightenment-devel

Re: [E-devel] edbus - Annoying inconsistency

2013-04-16 Thread Iván Briano

On Wed, Apr 17, 2013 at 12:18 AM, Carsten Haitzler ras...@rasterman.com wrote:
 On Tue, 16 Apr 2013 20:57:33 -0300 Lucas De Marchi
 lucas.demar...@profusion.mobi said:

 On Tue, Apr 16, 2013 at 8:34 PM, Carsten Haitzler ras...@rasterman.com
 wrote:
  On Tue, 16 Apr 2013 16:24:08 -0300 Lucas De Marchi
  lucas.demar...@profusion.mobi said:
 
  On Tue, Apr 16, 2013 at 4:13 PM,  michael.blumenkra...@gmail.com wrote:
   Tom Hacohen tom.haco...@samsung.com wrote:
  
   On 16/04/13 16:01, Lucas De Marchi wrote:
  
   On Tue, Apr 16, 2013 at 11:54 AM, Tom Hacohen 
   tom.haco...@samsung.com
   wrote:
  
   On 16/04/13 15:26, Lucas De Marchi wrote:
  
   On Tue, Apr 16, 2013 at 11:12 AM, Tom Hacohen
   tom.haco...@samsung.com wrote:
  
   Hey guys,
  
   How are you? I just noticed that 'edbus' installs headers under:
   prefix/include/edbus-1
   while using the
   pkgconfig name 'edbus2'
   the library name 'libedbus2'
   and an header called 'EDbus.h'.
  
   Luckily for us, it's not too late to fix it. Only question, what
   should
   we
   change it to? Everywhere with the 2 or everywhere without the 2?
  
  
  
   E_DBus installed edbus.pc and libedbus.so - that's why we needed to
   put a 2 in the name.
  
   However the name is edbus and has no relations with the previous
   library. It's not a new version, it's a complete different one. So
   appending a 2 in headers and include path is the wrong thing to 
   do.
  
   If ebus was available I would vote for that. It seems we are short
   in options for names.
  
   Lucas De Marchi
  
  
  
  
   I understand putting the 2 was needed, but if we add it once, we
   should add
   it everywhere. It's annoyingly inconsistent.
  
   Why did you choose 2 then? Why not 4? Or A? You chose 2 because
   it's the
   best explanation, as it is, although mostly unrelated, e_dbus v2. And
   if
  
  
   because it was not me is the best explanation.
  
   that's good enough there, it's good enough everywhere.
  
   In conclusion, I see no harm in using edbus2 all around the place.
  
  
   it's not good enough there, even less everywhere.  The visible part to
   the developer is the include path and the namespace. And these should
   reflect the library name. If anything change everything to something
   else, but don't put the 2 there.
  
  
  
   Let's wait to see what others think, as we disagree about the source of
   the problem, and that's: is this a major inconsistency that should be
   solved.
  
   Anyhow, regarding different names: I suggest edbus2.
  
   --
   Tom.
  
  
   
  
   Precog is a next-generation analytics platform capable of advanced
   analytics on semi-structured data. The platform includes APIs for
   building apps and a phenomenal toolset for data science. Developers can
   use our toolset for easy data analysis  visualization. Get a free
   account! http://www2.precog.com/precogplatform/slashdotnewsletter
   
  
   enlightenment-devel mailing list
   enlightenment-devel@lists.sourceforge.net
   https://lists.sourceforge.net/lists/listinfo/enlightenment-devel
  
  
   I agree with whatever demarchi says.
 
  I disagree.
 
 
  oh, wait...
 
  Lucas De Marchi
 
  :-P ... i have to say though - consistency is good. not just for devs, but
  packagers and anyone looking at their filesystem. if things are named in a
  consistent pattern it is easy to know what belongs where without needing
  package databases or other forensics tools.

 Sure it's good, no doubt. Just that renaming to edbus2 for the sake of
 consistency is even worse, not because of the rename, but because of
 the name chosen.

 ebus? anyone?

 Lucas De Marchi

 whatever the name chosen is, if .pc file, header include dir, header names, 
 and
 library name are all consistent in terms of pattern/convention then we are
 good.:)



Platypus.h
platypus.pc
libplatypus.so

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


 --
 Precog is a next-generation analytics platform capable of advanced
 analytics on semi-structured data. The platform includes APIs for building
 apps and a phenomenal toolset for data science. Developers can use
 our toolset for easy data analysis  visualization. Get a free account!
 http://www2.precog.com/precogplatform/slashdotnewsletter
 ___
 enlightenment-devel mailing list
 enlightenment-devel@lists.sourceforge.net
 https://lists.sourceforge.net/lists/listinfo/enlightenment-devel

--
Precog is a next-generation analytics platform capable of advanced
analytics on semi-structured data. The platform includes APIs for building
apps and a phenomenal toolset for data science. Developers can use
our toolset for easy data analysis

Re: [E-devel] edbus - Annoying inconsistency

2013-04-16 Thread Jérôme Pinot

On 04/17/13 10:11, David Seikel wrote:
 On Tue, 16 Apr 2013 20:57:33 -0300 Lucas De Marchi
 lucas.demar...@profusion.mobi wrote:
 
  On Tue, Apr 16, 2013 at 8:34 PM, Carsten Haitzler
  ras...@rasterman.com wrote:
   On Tue, 16 Apr 2013 16:24:08 -0300 Lucas De Marchi
   lucas.demar...@profusion.mobi said:
  
   On Tue, Apr 16, 2013 at 4:13 PM,  michael.blumenkra...@gmail.com
   wrote:
Tom Hacohen tom.haco...@samsung.com wrote:
   
On 16/04/13 16:01, Lucas De Marchi wrote:
   
On Tue, Apr 16, 2013 at 11:54 AM, Tom Hacohen
tom.haco...@samsung.com wrote:
   
On 16/04/13 15:26, Lucas De Marchi wrote:
   
On Tue, Apr 16, 2013 at 11:12 AM, Tom Hacohen
tom.haco...@samsung.com wrote:
   
Hey guys,
   
How are you? I just noticed that 'edbus' installs headers
under: prefix/include/edbus-1
while using the
pkgconfig name 'edbus2'
the library name 'libedbus2'
and an header called 'EDbus.h'.
   
Luckily for us, it's not too late to fix it. Only question,
what should
we
change it to? Everywhere with the 2 or everywhere without
the 2?
   
   
   
E_DBus installed edbus.pc and libedbus.so - that's why we
needed to put a 2 in the name.
   
However the name is edbus and has no relations with the
previous library. It's not a new version, it's a complete
different one. So appending a 2 in headers and include
path is the wrong thing to do.
   
If ebus was available I would vote for that. It seems we are
short  in options for names.
   
Lucas De Marchi
   
   
   
   
I understand putting the 2 was needed, but if we add it once,
we should add
it everywhere. It's annoyingly inconsistent.
   
Why did you choose 2 then? Why not 4? Or A? You chose 2
because it's the
best explanation, as it is, although mostly unrelated, e_dbus
v2. And if
   
   
because it was not me is the best explanation.
   
that's good enough there, it's good enough everywhere.
   
In conclusion, I see no harm in using edbus2 all around the
place.
   
   
it's not good enough there, even less everywhere.  The visible
part to the developer is the include path and the namespace.
And these should reflect the library name. If anything change
everything to something else, but don't put the 2 there.
   
   
   
Let's wait to see what others think, as we disagree about the
source of the problem, and that's: is this a major
inconsistency that should be solved.
   
Anyhow, regarding different names: I suggest edbus2.
   
--
Tom.
   
   

   
Precog is a next-generation analytics platform capable of
advanced analytics on semi-structured data. The platform
includes APIs for building apps and a phenomenal toolset for
data science. Developers can use our toolset for easy data
analysis  visualization. Get a free account!
http://www2.precog.com/precogplatform/slashdotnewsletter

   
enlightenment-devel mailing list
enlightenment-devel@lists.sourceforge.net
https://lists.sourceforge.net/lists/listinfo/enlightenment-devel
   
   
I agree with whatever demarchi says.
  
   I disagree.
  
  
   oh, wait...
  
   Lucas De Marchi
  
   :-P ... i have to say though - consistency is good. not just for
   devs, but packagers and anyone looking at their filesystem. if
   things are named in a consistent pattern it is easy to know what
   belongs where without needing package databases or other forensics
   tools.
  
  Sure it's good, no doubt. Just that renaming to edbus2 for the sake of
  consistency is even worse, not because of the rename, but because of
  the name chosen.
  
  ebus? anyone?
 
 The name of the ebus goes 'round and 'round.

As we all like E, I suggest edeebus

-- 
Jérôme Pinot
http://ngc891.blogdns.net/


signature.asc
Description: Digital signature
--
Precog is a next-generation analytics platform capable of advanced
analytics on semi-structured data. The platform includes APIs for building
apps and a phenomenal toolset for data science. Developers can use
our toolset for easy data analysis  visualization. Get a free account!
http://www2.precog.com/precogplatform/slashdotnewsletter___
enlightenment-devel mailing list
enlightenment-devel@lists.sourceforge.net
https://lists.sourceforge.net/lists/listinfo/enlightenment-devel

Re: [E-devel] edbus - Annoying inconsistency