Hi!
I've been wondering why there are ways to append to a strbuf but no way to
create one from scratch providing a format string and arguments. Actually,
I've found a few places where a strbuf is created (with its default size) and
then appended to, possibly causing a reallocation. I think the
On Friday 02 September 2011 23:25:01 Gustavo Sverzut Barbieri wrote:
> Recently we added something similar to Evas' generic loaders to
> Emotion. Initially Antognolli installed them to $(bindir) (/usr/bin)
> and I complained and changed it to $(libdir)/emotion/utils similar to
> Enlightenment's uti
On Friday 02 September 2011 09:12:07 Vincent Torri wrote:
> using typedef enum _foo
> { *** } foo;
>
> would be cleaner. I'm wondering if the c++ standard forbid such
> construction
Yes, but only because it uses a name with a leading underscore, which is
reserved in the global namespace. That sa
On Saturday 20 August 2011 23:13:23 Quentin Gibeaux wrote:
> The images downloaded from the web are stored in the /tmp folder (is that
> linuxproof only?).
The canonical dir for temporary files is $TMPDIR, according to POSIX, and that
should be used on anything but MS Windows, which has a special
On Saturday 16 July 2011 09:11:15 Vincent Torri wrote:
[hexdump]
> what's the difference with
>
> diff -u file1 file2
>
> ?
"diff" tries to be convenient, just like many text-based tools, so it hides
different line-endings from you. "hexdump" works with bytes and doesn't try to
make things eas
On Saturday 16 July 2011 01:04:29 Vincent Torri wrote:
> On Fri, 15 Jul 2011, Ulrich Eckhardt wrote:
> > Does it have multiple lines at all?
>
> what do you mean ?
What is the file content? Is it
- multiple lines terminated by CR or CRLF?
- a single line terminated by CR or
On Thursday 14 July 2011 16:36:58 Vincent Torri wrote:
> raster added a 'dos2unix' function in epp to remove trailing \r (rev
> 59799).
As I understand it, it converts '\r\n' to '\n' in a string.
> However, when i compare the files generated by epp on Windows and
> linux (just after line 751 of
On Tuesday 12 July 2011 16:25:21 Christopher Michael wrote:
> The old evas/ecore_x code used xcb_image_create_native to create basic x
> images, but the code for xcb_image_create_native does not check for
> endian-ness (it creates everything using MSB_FIRST) which is incorrect.
If you need somethi
On Saturday 11 June 2011 16:43:55 Christopher Michael wrote:
> On 06/11/2011 06:11 AM, Ulrich Eckhardt wrote:
> > On Saturday 11 June 2011 10:19:40 Ulrich Eckhardt wrote:
> >> I'm trying to create a patch which does just that, which I'll post
> >> lateron.
&
On Wednesday 15 June 2011 22:27:06 Cedric BAIL wrote:
[...]
> /**
> + * @brief Redefine the callback that clean the data of a hash
> + *
> + * @param hash The given hash table
> + * @param data_free_cb The function called on each value when the
> + * hash table is freed. @c
On Wednesday 15 June 2011 18:06:03 Cedric BAIL wrote:
> On Wed, Jun 15, 2011 at 5:53 PM, Mike Blumenkrantz
wrote:
> > On Wed, 15 Jun 2011 08:46:11 -0700
> >
> >> /**
> >> + * @brief Redefine the callback that clean the data of a hash
> >> + *
> >> + * @param hash The given hash table
> >> + * @
On Wednesday 15 June 2011 07:56:25 Mike Blumenkrantz wrote:
> On Wed, 15 Jun 2011 07:55:27 +0200 (CEST)
> Vincent Torri wrote:
> > On Wed, 15 Jun 2011, Mike Blumenkrantz wrote:
> > > I like the idea of better error messages, but using access creates a
> > > permissions loophole where the file coul
Hi!
If you start edje_player or edje_inspector with a path to a file that doesn't
exist you get a bogus error message saying that the file doesn't contain any
groups. The attached patch uses access() to check if the program can read the
file, giving a sensible error message if not.
I have chec
Hi!
I'm trying to use the OpenGL renderer of Evas on an X11 display. As a test
program, I'm using edje_player. Normally it works, but this doesn't:
$ edje_player -e=opengl_x11 ../../test1.edj
ERR<5244>:evas-gl_x11 evas_x_main.c:802 eng_best_visual_get()
glXChooseFBConfig returned no configs
On Saturday 11 June 2011 10:19:40 Ulrich Eckhardt wrote:
> I'm trying to create a patch which does just that, which I'll post lateron.
Here it comes.
There are a few observations which I'm not sure how to resolve:
1. configure.ac
There is a check for the Xcursor support, whic
Greetings!
The ecore tests fail to build because they need to be linked to eina, the
error is that eina_log_domain_register is not found by the linker. The
attached patch fixes that.
Uli
Index: Makefile.am
===
--- Makefile.am (revi
Hi!
The error is
Evas_Engine_XRender_X11.h: No such file or directory
while building ecore. The problem is that this engine was removed from evas
but not yet completely from ecore. I was on IRC with Vincent Torri (vtorri)
and Daniel Juyung Seo (SeoZ) and the consensus was to remove the code
Hi!
I hope I'm not asking too many stupid questions, but here's another one that
came up while reading Eina's buffer implementation. In e.g. the function
"eina_strbuf_common_new(size_t csize)", there is this undocumented parameter
"csize" that is passed to various other functions, too. This see
I noticed some weirdness in eina_ustrbuf.h, there is a macro called
eina_ustrbuf_prepend_Eina_Unicode, which looks as if some search'n'replace
operation on it had gone wrong. Looking at its documentation seems to confirm
this. Some invokations of "svn blame", I find that this weirdness is presen
Hi!
I have a question regarding the thread-safety and reentrancy of Evas. Firstly,
are these actually goals or is Evas intended by design to be used in a single
thread for rendering? If so, you can ignore my other questions
Now, what got me started on this issue is the commit in
http://tra
On Tuesday 07 June 2011 22:53:59 Vincent Torri wrote:
> It's nice to see someone interested in the win32 code :)
Thank you for writing it! Actually, I'm interested in the WinCE port even,
because that's the platform I work with for a living, doing embedded
programming in an industrial environmen
Hi!
The following code is an example from that file:
EAPI void
ecore_win32_window_free(Ecore_Win32_Window *window)
{
Ecore_Win32_Window *wnd = window;
if (!window) return;
INF("destroying window");
if (wnd->shape.mask)
free(wnd->shape.mask);
DestroyWindow(((Ecore_Win32_Wi
Hi!
The functions looks like this:
int _ecore_thread_win32_join(win32_thread *x, void **res)
{
if (!PHE(x, PHS()))
{
WaitForSingleObject(x->thread, INFINITE);
CloseHandle(x->thread);
}
if (res) *res = x->val;
return 0;
}
What I'm wondering is if there shouldn't be a "f
Hi!
I've been reading commits recently, and found a small buglet committed in
revision 59823. The line
ECORE_MAGIC_FAIL(p, ECORE_MAGIC_PIPE, "ecore_pipe_read_close");
should probably contain the function name as string, but it doesn't for the
freeze/thaw functions.
Uli
---
24 matches
Mail list logo
