[E-devel] Ecore patch: ecore_file_download memory leak
Hi, This fixes a memory leak in the ecore_file_download function. The string returned by ecore_file_dir_get (that was strdup'd) was not being freed. Also, removing whitespaces at the end of some lines. -- Vinicius - This SF.net email is sponsored by: Splunk Inc. Still grepping through log files to find problems? Stop. Now Search log events and configuration files using AJAX and a browser. Download your FREE copy of Splunk now http://get.splunk.com/___ enlightenment-devel mailing list enlightenment-devel@lists.sourceforge.net https://lists.sourceforge.net/lists/listinfo/enlightenment-devel
Re: [E-devel] News from the E stables
Carsten wrote: First. I need to offer some big apologies. I have been pretty bad at being . Well, I'll give you my own thoughts on this. These new developments in E may well be exciting new opportunities and whatnot, but for some e-developers at least, there are long-standing historical issues with this project. What you call your 'lack of leadership', some saw rather differently. Indeed, some saw it as a good approach, a benevolent- dictator kind of approach. Maybe they're right, and that is a 'good' way to develop something like a wm plus its supporting libs.. but it's not likely to scale well much beyond that - eg. to scale to something like a serious development platform (unless perhaps you envision the wm+modules as a kind of limited platform itself). On several ocassions, some developers here have brought up some of the things you seem to be addressing now - essentially, to have the project run as a true 'community' project. But, they were either ignored, brushed-off, or even treated with some hostility. Busy or not, you've never expressed any interest or support for seeing this project become a true community one. Why the sudden change of heart? Because you have a new employer and can now be the leader you say you haven't had the chance to be? While I applaud your statements to (finally) want to see this project move in the direction of a true community effort, and I believe the things you've proposed are good... Do you really expect everyone to feel warm and fuzzy all over, ignore the past, and joyfully embrace this 'new age of enlightenment'? I hope they do, honestly - but I think that it's not going to be that easy.. not for those who were satisfied with the old order, and not for those who've questioned things and called for change before. jose. _ Click for your daily horoscope, learn about money, love amp; family. http://thirdpartyoffers.juno.com/TGL2121/fc/Ioyw6i3ly8kw6Kq0SdgSaubx9a4J7TZF52NdjtfrVt8447rKt2qTQc/ - This SF.net email is sponsored by: Splunk Inc. Still grepping through log files to find problems? Stop. Now Search log events and configuration files using AJAX and a browser. Download your FREE copy of Splunk now http://get.splunk.com/ ___ enlightenment-devel mailing list enlightenment-devel@lists.sourceforge.net https://lists.sourceforge.net/lists/listinfo/enlightenment-devel
[E-devel] Nightly build log for E17 on 2007-11-07 07:04:53 -0800
Build log for Enlightenment DR 0.17 on 2007-11-07 07:04:53 -0800 Build logs are available at http://download.enlightenment.org/tests/logs Packages that failed to build: engage http://download.enlightenment.org/tests/logs/engage.log epdf http://download.enlightenment.org/tests/logs/epdf.log evolve http://download.enlightenment.org/tests/logs/evolve.log Packages with no supported build system: esmart_rsvg, exorcist, python-efl, Packages skipped: camE, enotes, enscribe, epbb, eplay, erss, etk_server, etox, e_utils, Evas_Perl, evoak, gfx_routines, lvs-gui, med, nexus, notgame, ruby-efl, webcam, Packages that build OK: alarm, bling, cpu, deskshow, eclair, ecore, edb, e_dbus, edje_editor, edje, edje_viewer, edvi, eet, eflame, eflpp, efreet, elapse, elation, elicit, elitaire, e, embrace, embryo, emotion, emphasis, empower, emu, engrave, engycad, enhance, enity, enterminus, enthrall, entice, entrance_edit_gui, entrance, entropy, envision, epeg, ephoto, e_phys, epsilon, equate, esmart, estickies, etk_extra, etk, etk-perl, evas, evfs, ewl, examine, exhibit, exml, expedite, express, extrackt, feh, flame, forecasts, gevas2, iconbar, imlib2_loaders, imlib2, Imlib2_Perl, imlib2_tools, language, mail, mem, mixer, moon, net, news, pesh, photo, rage, rain, screenshot, scrot, slideshow, snow, taskbar, tclock, uptime, weather, winselector, wlan, Debian GNU/Linux 4.0 \n \l Linux enlightenment2 2.6.18-4-686 #1 SMP Wed May 9 23:03:12 UTC 2007 i686 GNU/Linux See http://download.enlightenment.org/tests/ for details. - This SF.net email is sponsored by: Splunk Inc. Still grepping through log files to find problems? Stop. Now Search log events and configuration files using AJAX and a browser. Download your FREE copy of Splunk now http://get.splunk.com/ ___ enlightenment-devel mailing list enlightenment-devel@lists.sourceforge.net https://lists.sourceforge.net/lists/listinfo/enlightenment-devel
Re: [E-devel] Ecore patch: ecore_file_download memory leak
Sorry folks, my firefox thought the patch was a binary and it was ignored.
On 11/7/07, Vinicius Gomes [EMAIL PROTECTED] wrote:
Hi,
This fixes a memory leak in the ecore_file_download function. The
string returned by ecore_file_dir_get (that was strdup'd) was not
being freed. Also, removing whitespaces at the end of some lines.
--
Vinicius
--
Vinicius
Index: src/lib/ecore_file/ecore_file_download.c
===
RCS file: /var/cvs/e/e17/libs/ecore/src/lib/ecore_file/ecore_file_download.c,v
retrieving revision 1.16
diff -u -r1.16 ecore_file_download.c
--- src/lib/ecore_file/ecore_file_download.c6 Aug 2007 20:00:28 -
1.16
+++ src/lib/ecore_file/ecore_file_download.c7 Nov 2007 20:05:41 -
@@ -90,7 +90,7 @@
if (!_job_list)
return;
-
+
ecore_list_first_goto(_job_list);
while ((job = ecore_list_next(_job_list)))
{
@@ -116,7 +116,7 @@
*
* You must provide the full url, including 'http://', 'ftp://' or 'file://'.\n
* If @p dst already exist it will not be overwritten and the function will
fail.\n
- * Ecore must be compiled with CURL to download using http and ftp protocols.
+ * Ecore must be compiled with CURL to download using http and ftp protocols.
*/
EAPI int
ecore_file_download(const char *url, const char *dst,
@@ -124,7 +124,17 @@
int (*progress_cb)(void *data, const char *file, long int
dltotal, long int dlnow, long int ultotal, long int ulnow),
void *data)
{
- if (!ecore_file_is_dir(ecore_file_dir_get((char *)dst))) return 0;
+ if (dst == NULL) return 0;
+
+ char *path = ecore_file_dir_get((char *) dst);
+
+ if (!ecore_file_is_dir(path))
+ {
+free(path);
+return 0;
+ }
+ free(path);
+
if (ecore_file_exists(dst)) return 0;
/* FIXME: Add handlers for http and ftp! */
@@ -145,7 +155,7 @@
{
/* download */
Ecore_File_Download_Job *job;
-
+
job = _ecore_file_download_curl(url, dst, completion_cb, progress_cb,
data);
if (job)
return 1;
@@ -170,7 +180,7 @@
* @return 1 if protocol is handled or 0 if not
*
* @p protocol can be 'http://', 'ftp://' or 'file://'.\n
- * Ecore must be compiled with CURL to handle http and ftp protocols.
+ * Ecore must be compiled with CURL to handle http and ftp protocols.
*/
EAPI int
ecore_file_download_protocol_available(const char *protocol)
@@ -185,15 +195,15 @@
}
#ifdef HAVE_CURL
-/* this reports the downloads progress. if we return 0, then download
+/* this reports the downloads progress. if we return 0, then download
* continues, if we return anything else, then the download stops */
static int
_ecore_file_download_curl_progress_func(void *clientp, double dltotal, double
dlnow, double ultotal, double ulnow)
-{
+{
Ecore_File_Download_Job *job;
-
+
job = clientp;
-
+
if(job-progress_cb)
return job-progress_cb(job-data, job-dst, (long int)dltotal, (long
int)dlnow, (long int)ultotal, (long int)ulnow);
return 0;
@@ -203,9 +213,9 @@
_ecore_file_download_curl(const char *url, const char *dst,
void (*completion_cb)(void *data, const char *file,
int status),
- int (*progress_cb)(void *data, const char *file,
+ int (*progress_cb)(void *data, const char *file,
long int dltotal, long int dlnow,
-long int ultotal,
+long int ultotal,
long int ulnow),
void *data)
{
@@ -234,17 +244,17 @@
free(job);
return NULL;
}
-
+
curl_easy_setopt(job-curl, CURLOPT_URL, url);
curl_easy_setopt(job-curl, CURLOPT_WRITEDATA, job-file);
-
+
if (progress_cb)
{
- curl_easy_setopt(job-curl, CURLOPT_NOPROGRESS, FALSE);
- curl_easy_setopt(job-curl, CURLOPT_PROGRESSDATA, job);
+ curl_easy_setopt(job-curl, CURLOPT_NOPROGRESS, FALSE);
+ curl_easy_setopt(job-curl, CURLOPT_PROGRESSDATA, job);
curl_easy_setopt(job-curl, CURLOPT_PROGRESSFUNCTION,
_ecore_file_download_curl_progress_func);
}
-
+
job-data = data;
job-completion_cb = completion_cb;
job-progress_cb = progress_cb;
-
This SF.net email is sponsored by: Splunk Inc.
Still grepping through log files to find problems? Stop.
Now Search log events and configuration files using AJAX and a browser.
Download your FREE copy of Splunk now http://get.splunk.com/___
enlightenment-devel mailing list
enlightenment-devel@lists.sourceforge.net
https://lists.sourceforge.net/lists/listinfo/enlightenment-devel
Re: [E-devel] [e-users] Integrating an OpenGL render Window within an Evas canevas.
Thanks for the hints. I will provide the E-mailing list with my code/hacks or else as soon as I have something working. By 3D, I mean 3D as in OpenGL object made of voxels, not 3D desktop like XGL/compiz or whatever the new things are which, anyways, I don't use ;-) thanks all for your time. Flo. On 6-Nov-07, at 7:40 PM, Gustavo Sverzut Barbieri wrote: On Nov 6, 2007 9:26 PM, Flo [EMAIL PROTECTED] wrote: well, thanks for your answers and ideas. I have to find my way to do this anyways otherwise no more E17/Evas and beauty for me ! Indeed, for the kind of applications I am writing (research in 3D Imaging processing - filtering and so on) there is no way I can keep going without 3D openGL Renderers ! I thought of maybe attaching my OpenGL renderer window to a GTK canvas, through the gtk-opengl extension, and as well integrating my Evas UI through gevas2. - that is lots of dirty hacks for an end result that might be way too far from the kinda really nice Evas UI I was looking for ! No, don't go this route. Given your software, seems that your best bet is to attach a new X window with ecore_x_window_new(), it's quite easy to do so. I, personally, don't understand why there are not more ppl interested in doing so. To me, this would definitely bring the EFLs to the top UI option, at least for ppl with a little bit of taste. well, Linux 3d is still far from good, so it's far from being really used... so ... you got it. But sure, it deserves more attention. We really plan to integrate 3d as a first class citizen, but it will not happen that soon. Anyways, thanks for your time, let me know if you have other ideas ... and more importantly, I will let you know if I come up with a solution, then an elegant solution ;-) ! Ok, if you develop a simple helper around ecore_x_window_new(), then let us know, publish the code somewhere. Hints for implementation: X is async, you need to wait for commands to complete and given back to your process, so you need to use ecore_event_handler_add() for some events, otherwise you'll get errors from X. But ecore makes this quite simple. -- Gustavo Sverzut Barbieri -- Jabber: [EMAIL PROTECTED] MSN: [EMAIL PROTECTED] ICQ#: 17249123 Skype: gsbarbieri Mobile: +55 (81) 9927 0010 - This SF.net email is sponsored by: Splunk Inc. Still grepping through log files to find problems? Stop. Now Search log events and configuration files using AJAX and a browser. Download your FREE copy of Splunk now http://get.splunk.com/ ___ enlightenment-devel mailing list enlightenment-devel@lists.sourceforge.net https://lists.sourceforge.net/lists/listinfo/enlightenment-devel
Re: [E-devel] [e-users] Integrating an OpenGL render Window within an Evas canevas.
That's good to hear that. I'm glad you were able to achieve to include/integrate anyhow a rendering space for openGL renderings. Although I need mouse interactions to be able to rotate my 3D openGL volumes, that is what I'm looking for. If I'm able to do this all, I will send some Code for any future users. On a same note: If anybody wants to start a project/module or related project to integrate External Renderer windows within Evas/GEvas canvas I'm up for it. regards, Flo. On 6-Nov-07, at 6:49 PM, Jorge Luis Zapata Muga wrote: On Nov 4, 2007 2:21 AM, Flo [EMAIL PROTECTED] wrote: Dear all: My name's florent and although I'm rather new to the EFLs world, and thanks to some great examples over the internet (from Gustavo Sverzut Barbieri and Raoul Hecky), I already managed to achieved nice results using Evas. However, after investigating evas's (endless) possibility as a GUI I'm hitting a wall while trying to do the following: I want to integrate an OpenGL render window within my Evas UI, this Evas UI handling every callbacks (mouse and keyboard events) whether from the evas canevas or the openGL render window. The example I tried to achieved, but failed so far, would be to integrate a glxgears render window (from the famous linux glxgears OpenGL render app) in a simple Evas canevas contaning a button that would launch the openGL render window. I would appreciate any idea/code/direction on how to achieve this as, from my internet digging, I did not find any example. thanks for your help. florent. Hi, don't know if this might help you, but i did something similar with the fb engine (if i understood correctly): i make the engine export the private engine data (i.e the fb pointer) to my evas application, then i do an obscure rectangle in the part of the fb that i will access so evas wont draw there (the mouse and everyhting wont be drawn too) and use my own gfx routines to draw there (the obscure / redraw code in evas isnt perfect, some small parts of the obscure rectangle are redrawn, but its ok). With this you can use opengl functions on that area and still use evas' ones on the other part of the window. This is a good reason to change how evas behaves with engines, some engines allocate the data by themselves and you can't access it from outside and to others you provide them with the required engine data. regards - This SF.net email is sponsored by: Splunk Inc. Still grepping through log files to find problems? Stop. Now Search log events and configuration files using AJAX and a browser. Download your FREE copy of Splunk now http://get.splunk.com/ ___ enlightenment-users mailing list [EMAIL PROTECTED] https://lists.sourceforge.net/lists/listinfo/enlightenment-users - This SF.net email is sponsored by: Splunk Inc. Still grepping through log files to find problems? Stop. Now Search log events and configuration files using AJAX and a browser. Download your FREE copy of Splunk now http://get.splunk.com/ ___ enlightenment-devel mailing list enlightenment-devel@lists.sourceforge.net https://lists.sourceforge.net/lists/listinfo/enlightenment-devel
Re: [E-devel] E17's Direction and Development
On Sun, 4 Nov 2007 07:43:40 -0500 Hisham Mardam Bey [EMAIL PROTECTED] babbled: it shouldn't take a tonne of time to complete. though from what you say you are suggesting a freeze to fix bugs - not finish the TODO ? Not entirely - just avoid any items on the TODO (if any) that are big-ish new features as opposed to fixing bugs or completing currently open (and incomplete) features. agreed - though these 2 biggish ones are imho needed, but that's it. i agree - the wizard i know i only want certain functionality - it's even modular in and of itself, so we only put the modules in we need. i even have a proposed list of modules we need to do - it's not long. it's in comments in the wizard code. most are simple. i'd be happy to drop some (xrender vs software test and just use software anyway), drop default ibar apps (just ship with a default set anyway - but make that a separate file for the list so it can be customised per distribution), drop the keybindings question and default wallpaper and theme q's - most of the rest are really trivial question pages. it just needs momentum and examples. the fm has more work - but not a lot more. it's so close i don't think it should be all dropped. it's been on the todo forever. after that we have done all major todo items left - the rest are minor things, cleanups, fixing gaps in complete functionality etc. Sounds good. :) we can set a schedule - and i will bet we will slip. if we set milestones/tasks and just do those - then we can get it done at a pace that allows that. if you want a schedule from me - i will currently put in 0 hours per week for the next 2 months as an allocation - why? i'm moving countries and who knows what time i will have. in essence those milestones are IN the TODO. i'm removing fm todo items i think can be dropped right now. i've dropped some other items. see cvs commit. Noted. I guess that after your last commit to the TODO, we can grab that file and split it up into milestones. I agree that although it would be nice to time things (set a schedule), it would also be hard since we have very (VERY) few people who can work on E17 on a daily basis giving in solid hours every (other) day. yup. thus N tasks to complete and dependencies will do just that for you. the tasks are in TODO. just no order for them or dependency info. What I want us to try to get to is a point where we can associate milestones with release candidates of some sort, and a final release for the last milestone. When we have something like that going we'll not only feel like we're getting somewhere, but we can have snapshots / announcements on the ML and let people feel like we're approaching a final release. So at this point, the next task is to create those milestones, associate them with release candidates, split up the TODO items between them, and just hack hack hack at them. hmm - for me a release candidate - ie first one being alpha would be all our feature todo items done. we are down to bugs (and maybe some spit and polish on dialog layouts, theme, etc.). thats alpha. beta is spit and polish stuff improved and most bugs gone if not all. and RC1,2,3 come after that based on more bug feedback etc. etc. once we get down to a very small # of bugs (eg 0) we do a release. -- Hisham Mardam Bey http://hisham.cc/ +1-514-713-9312 Codito Ergo Sum (I Code Therefore I Am) - This SF.net email is sponsored by: Splunk Inc. Still grepping through log files to find problems? Stop. Now Search log events and configuration files using AJAX and a browser. Download your FREE copy of Splunk now http://get.splunk.com/ ___ 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)[EMAIL PROTECTED] - This SF.net email is sponsored by: Splunk Inc. Still grepping through log files to find problems? Stop. Now Search log events and configuration files using AJAX and a browser. Download your FREE copy of Splunk now http://get.splunk.com/ ___ enlightenment-devel mailing list enlightenment-devel@lists.sourceforge.net https://lists.sourceforge.net/lists/listinfo/enlightenment-devel
Re: [E-devel] Ecore + XIM working
On Tue, 6 Nov 2007 23:53:31 +0900 Stafford Horne [EMAIL PROTECTED] babbled: as per IRC - lets just drop the preedit_type - lets just make it callback based (ie the ecore_x events to notify you of info from the IM and if you need to send anything - ecore_x calls to tell it stuff).otherwise looks ok to me. Hello, It seems my original patch did not go out. The files are here: http://www.shorne-pla.net/uploads/ecore_xim.diff http://www.shorne-pla.net/uploads/xim_ecore.c New function ecore_x_window_input_context_init(win, Root); Lately I have been working on shutdown code. I still need to fix it so that if the XIM server dies and comes back only all input_context objects will be recreated automatically. Still working on other preedit types, only Root is working now. I am thinking now that IM callbacks should be handled with ecore_events instead of having to actually register callbacks. -Stafford On Sat, 3 Nov 2007 13:00:13 +0900 Stafford Horne [EMAIL PROTECTED] wrote: Hi All, Just want to let you know that I got the first steps of ecore XIM integration done and basic functionality is working. Attached is a diff for ecore_x and a small sample application. Currently there is only one extra method needed to be called to add an input_context to your window. The method takes one argument which should be the type for preedit style (OverTheSpot, OnTheSpot, Root, Callbacks, etc.) but only Root works for now. void ecore_x_window_input_method_init(const char *input_style); Please, if someone can just check my approach and api and give comments I will be happy. I have a few ideas for how to handle setting up position callbacks, rendering callbacks and other things, but ill try to catch up in IRC later. -Stafford - This SF.net email is sponsored by: Splunk Inc. Still grepping through log files to find problems? Stop. Now Search log events and configuration files using AJAX and a browser. Download your FREE copy of Splunk now http://get.splunk.com/ ___ 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)[EMAIL PROTECTED] - This SF.net email is sponsored by: Splunk Inc. Still grepping through log files to find problems? Stop. Now Search log events and configuration files using AJAX and a browser. Download your FREE copy of Splunk now http://get.splunk.com/ ___ enlightenment-devel mailing list enlightenment-devel@lists.sourceforge.net https://lists.sourceforge.net/lists/listinfo/enlightenment-devel
