Re: [E-devel] straight exec'ing for ecore_exe_run
On the other hand, the rule out the presence of certain shell metacharacters part may be a problem. I can state that it will not be a problem for emu, as anything that emu executes will be written specifically for emu. Other uses may have a problem though. Well, my patch is pretty conservative currently (anything suspicious, pass over to sh), and I read through SUSv3 before coming up with the code. So, I hope there won't be many problems. Did much of the testing using -binding-key-add and exec. But since I can't weed out the bugs otherwise, I guess I will commit. All fingers crossed :) Ramkumar. -- WARN_(accel)(msg null; should hang here to be win compatible\n); -- WINE source code --- This SF.net email is sponsored by: Splunk Inc. Do you grep through log files for problems? Stop! Download the new AJAX search engine that makes searching your log files as easy as surfing the web. DOWNLOAD SPLUNK! http://ads.osdn.com/?ad_idv37alloc_id865op=click ___ enlightenment-devel mailing list enlightenment-devel@lists.sourceforge.net https://lists.sourceforge.net/lists/listinfo/enlightenment-devel
Re: [E-devel] specifying user environment in Entrance
On 28/11/05, Michael Jennings wrote: On Tuesday, 29 November 2005, at 12:25:54 (+0900), Carsten Haitzler wrote: as per xcomps own words - he is too busy to work on it. he would like to, but can't. So what? That doesn't make him any less a developer. It just makes him a busy developer. Something you and I know a little something about. considering it has bugs, issues, things like simply not working in xinerama, still managing to do re-start loops fo x, not handling failed loing scripts gracefully... i have listed all these before. and basically nothing really gets done. it's not goign anywhere. it's dormant. it's on minimal life support on a good day. That's not a fair assessment. Yes, there are specific problems. But not a single one of them is a show-stopper. I use entrance on every system I have, and the simple fact is that it performs flawlessly more than 99% of the time. Having bugs that don't get fixed right away doesn't make it dead. I still remember numerous conversations regarding a certain focus bug in E 0.16, even back when you were still actively maintaining it. And it's still not dead. frankly- this is where we differ. if it was alive, ibukun would be answering these mails, patching the code, improving it, working on current issues etc. etc. etc. fact is - that is not happening. i know he's busy and i dont blame him for it. Still doesn't make it dead. Dormant, perhaps, but not dead. whoever works on it the most takes over by defacto. there is no wonership system. welcome to open source. Bullshit. That's not open source. That's anarchy. Someone can fork the project, give it a new name, and continue onward (in accordance with the license, of course). But that doesn't mean they can take over the project itself. No more than someone could come along and fork Enlightenment and continue calling it Enlightenment while developing a conflicting product. anyone who takes the code ans runs with it, adnd if there is no competition (ie no one else is working on it) becomes the new owner/maintainer by defacto. Again, that's not your call. If you want to treat E that way, fine. But your word is not law for entrance, and copyright law is on Ibukun's side, not yours. this isn't the UN and some beurocracy where you need all the red tape and signing off on things. he who does the work gets the credit, and gets control. ibukun sure knows mroe about entrance than almsot anyone and thus is a good soruce of info and help,a nd is in the best position to work on it effectively, but anyone able to school up si then on equal footing, and thus he who does the work - wins. It's not a contest. Nobody wins or loses. Anyone is free to work on open source software, but that isn't to say anyone can do anything they want with it. It's what keeps Sun from calling their next desktop environment Enlightenment. so by your theory - ibukun could simply idle here for the next 10 years, and not do anything to entrance - and until he officially says i let go its all his and no one can do anything in terms of releases etc. that's utterly stupid. Whether it's stupid or not is irrelevant. That is reality. And it's not my theory either. Like I said, someone else can create a fork, but they cannot usurp control of entrance itself. one thing i will say is that *I* will never produce a release of entrance unless i take ownership. i will make no patches and fix no bugs, unless they annoy me enough to do so. one day i might take over if i get time - but i likely won't. fact is - by DEFACTO if you dont work on something AND someone else then does a tonne of work - the new person becomes maintainer. That is not a fact. That is an idealistic utopian pipe dream. That is an attempt to live in a world where no one owns anything and everyone contributes selflessly to the greater good. Even Stallman, the self-proclaimed deity of collaborative software and communal living, admits that *somebody* has to own the stuff or everybody loses. Taking over a project without the owner's consent is no different from regurgitating it under your own name without giving credit where it's due: you're taking actions which only the copyright holder has the legal right to take. Without respecting ownership, someone could quite easily release Enlightenment 0.13.3 as 1.0 without your consent, and there would be nothing you could do about it. i've had bad experiences. So have I. But those experiences don't change reality. do NOT rely on ibukun to won entrance and maintain it. dont rely on anyone - unless they are actively DOING stuff. Don't rely on anyone, ever. Even if they *are* doing stuff. They could just as easily disappear in an instant by taking their own life as they could fade away over time. it's just a general thing. perfect example si the whole crap we have about e.org now - someone never
[E-devel] 3rd party Module Internationalization
Gidday all, I have posted a couple of blog entries about 3rd party module internationalization over at edevelop. 3rd party Module Internationalization - Overview http://edevelop.org/node/1853 3rd party Module Internationalization - Proof of concept http://edevelop.org/node/1860 My autotools knowledge is rudimentary, but hopefully this can be useful for other 3rd party module developers wanting to internationalize :-) If anyone has suggestions for improvement please let me know. Regards! David
Re: [E-devel] specifying user environment in Entrance
This is funny indeed. What's the debate? The code is BSD licensed, and I own the rights to it. That said, I've already said before, I would like at least one person to help me maintain and manage the code. To say that I've disappeared altogether is not fair at all, or to compare me to others who left the project altogether. I do keep in touch, check the mailing list from time to time, make suggestions and contributions when I can. If anyone wants to actively work on/maintain Entrance, by all means contact me - I'd love to see some work being done on it. But this debate here is just ridiculous. I thought someone was offering to write a patch for an Entrance feature. Why not discuss THAT? --Ibukun On Tue, November 29, 2005 10:27 am, Nathan Ingersoll said: As only one person has even stepped forward to work on entrance recently with a couple small (but useful) patches, arguing ownership seems pretty pointless at this point. -- http://xcomputerman.com --- This SF.net email is sponsored by: Splunk Inc. Do you grep through log files for problems? Stop! Download the new AJAX search engine that makes searching your log files as easy as surfing the web. DOWNLOAD SPLUNK! http://ads.osdn.com/?ad_idv37alloc_id865op=click ___ enlightenment-devel mailing list enlightenment-devel@lists.sourceforge.net https://lists.sourceforge.net/lists/listinfo/enlightenment-devel
Re: [E-devel] specifying user environment in Entrance
On Tuesday, 29 November 2005, at 16:09:05 (+0100), Laszlo Treszkai wrote: I think this is the same situation, and we know xcomp is gone for a long time, No, we do not. and I'm sure he would give to anyone the right to maintain and own it, as I read him saying he CAN'T be active. I'd read it again, then. And he's not here and responsive. I think the person who's chatting on IRC right now using the nickname xcomp would disagree with that assertion. Michael -- Michael Jennings (a.k.a. KainX) http://www.kainx.org/ [EMAIL PROTECTED] n + 1, Inc., http://www.nplus1.net/ Author, Eterm (www.eterm.org) --- We're going to take a break and go outside for about 15 minutes. Security advises us that this would be a good idea. -- Bill Poucher --- This SF.net email is sponsored by: Splunk Inc. Do you grep through log files for problems? Stop! Download the new AJAX search engine that makes searching your log files as easy as surfing the web. DOWNLOAD SPLUNK! http://ads.osdn.com/?ad_id=7637alloc_id=16865op=click ___ enlightenment-devel mailing list enlightenment-devel@lists.sourceforge.net https://lists.sourceforge.net/lists/listinfo/enlightenment-devel
[E-devel] eVolume 0.0.17a has been released
Hi all! I again with you :), so made new release: http://paq.osdn.org.ua/~davinchi/evolume-0.0.17a.tar.bz2 as you can guess, `a' is stays for alpha, so don't surprise, if it segaultfs TODO list for now is: * Make internationalization * Write swapping and dragging code * Possibly use e_sliders. --- This SF.net email is sponsored by: Splunk Inc. Do you grep through log files for problems? Stop! Download the new AJAX search engine that makes searching your log files as easy as surfing the web. DOWNLOAD SPLUNK! http://ads.osdn.com/?ad_idv37alloc_id865op=click ___ enlightenment-devel mailing list enlightenment-devel@lists.sourceforge.net https://lists.sourceforge.net/lists/listinfo/enlightenment-devel
Re: [E-devel] specifying user environment in Entrance
On Tue, 29 Nov 2005 13:39:50 -0600 (CST) Ibukun Olumuyiwa [EMAIL PROTECTED] wrote: I thought someone was offering to write a patch for an Entrance feature. Why not discuss THAT? The, in my view, easiest solution for my problem would be to add an option to entrance_config.cfg specifiying whether or not the environment should be cleared. Would this (in general) be a security hazard? (I.e. not a good feature to add.) Thanks, Felix --- This SF.net email is sponsored by: Splunk Inc. Do you grep through log files for problems? Stop! Download the new AJAX search engine that makes searching your log files as easy as surfing the web. DOWNLOAD SPLUNK! http://ads.osdn.com/?ad_id=7637alloc_id=16865op=click ___ enlightenment-devel mailing list enlightenment-devel@lists.sourceforge.net https://lists.sourceforge.net/lists/listinfo/enlightenment-devel
Re: [E-devel] specifying user environment in Entrance
I'm afraid this thread is taking on a life of its own like some of the previous ones. So I'm going to try to trim a bit. On Tuesday, 29 November 2005, at 16:10:14 (+0900), Carsten Haitzler wrote: yes - but its not a temporary condition. xcomps life changed ina way where he'll be ultra-busy for many years to come. if you talk with him. i'll let him divulge his life's details. and it shows that entrance basically has fallen into dormant status ever since. My point is that it's his responsibility to give up the reigns or simply take a less active development role (and more management). It's his choice. and if the person who owns it vanishes? stops working on it? has no time? then it sits and rots (well just goes nowehere)? and no one is allowed to touch it without his express approval? come on! if someone wants to pick up the ball and keep runing wiht it - or more than 1 person. - why not? the original author doesnt lose credit. he still has the full respect and admiration for having picke dup the ball and run with it as far as they did. now a new set of people keep runing with it. you seem to have something against that? Not at all. Anyone with CVS access can contribute. I have no problem with that. All I'm saying is that the person in charge remains the person in charge until one of the following happens: (1) (S)he steps aside willingly; (2) (S)he is inexplicibly out of touch via e-mail and IRC for a significant period of time. Again, I'm not talking about contributors (i.e., people who commit changes to CVS). I'm talking about the lead developer. The person in charge. The designer, architect, and ultimate decision maker. For E, Evas, etc., that's you. For Eterm, it's me. For EWL, it's RbdPngn. For entrance, it's xcomp. For Etk, it's Simon. And so forth... well that reminds me. need to add myself to AUTHORS - as i have done work on entrance - a reasonable amount - enough to justify an AUTHORS entry... u can check cvs logs. thus as an AUTHOR i belive i have a little more weight to pull in this. i never saw you jump up and down when i worked on entrance beofre? do you see me making a fuss when YOU just commit things to spec and makefiles etc.? I think we're talking about two different things here. Anyone with commit access can contribute; I've never said anything about that. Hopefully my statements above will clarify. if they are the only one working on it - which they will be, You're making a rather large assumption there. they get to cal the shots as to when to make a release, what features to put in, how to code them, etc. what is wrong with that? What's wrong with it is that it's not their project, and as long as the original maintainer is still around, they do not forfeit their rights. If the original maintainer says no release, there will be no release. The contributor has two options at that point: accept the no release decree or fork the project. Otherwise things get very messy. so old owner says no - go away. don't touch it. ok. fine. now it gets no improvements as old owner is doing nothing. it gets no releases, no fixes. its DEAD. since any admin acts as benevolent dictator, its up to an ADMIN to call the final shot - should the fued just go on? shoudl someone be kicked out? (access removed). A project being committed to E's CVS tree does not give us (project admins) the right to overrule the original author. We can remove their access and kick them out of the E team, but that project remains theirs. They own the name, for lack of better terminology. The BSD license allows us to use the code and continue the work, but under a different name. i think you are taking this the wrong way. as if the new owner completely remvoes all copyright access. it's like a distribution. peolpe own a package. when a packager owns the kernel - they dont own copyright and all credit. they own the job of producing packages and then do it. if they dont do it - then someone else takes over as owner. if multiple people share co-ownership then they are a team. (ie they ALL are active). people dormant for long enough lose ownership of that project. they dont lose copyright or anything else, but they dont push the agenda of its code. Yes, we've established that you're referring to own in a different sense than I am. But Ibukun is still here and is still responsive. It's a very, very different situation. really? news to me. last commit by xcomp: nov 11 - it was just reverting a small change by seb. no new work. nov 9 - xcomp commits patch (not new work by him) nov 8 - xcomp puts in another path - not him either. nov 7 - xcomp adds patch again oct 12 - tiny 1 line change to turn off tcp on x oct 10 - tiny 1 line change for sh exec oct 10 - minimal key changes in config Commits are not the sole method of being responsive. Check e-mail. Check IRC. The bottom line is, it's up to him. So I'll let him say his piece. Michael -- Michael
[E-devel] Some basic contributing guidelins
Ok, it looks like we just need to lay out some simple guidelines to make everyone happy. Contributing code is ALWAYS welcome. Write your code, then use 'cvs diff -u' to create a patch. Attach it to an email to this list. If you haven't heard back after a few days, send another response. Or, send a personal email to the first listed AUTHOR for the package. (Sometimes people are busy and don't read the edevel stuff, but most of us feel naked when away from email for more than a day). By this point, if you still haven't heard from the author, most likely some one else will have reviewed/committed your patch. CVS access is easy to get around here. Just send a few good patches. We get tired of committing them for you, and give you access instead. If you break things too horribly (beyond the beginners cvs foul-ups that we all do) we can always revert things (and revoke your access ;-) ) If you see a project that seems dormant, chances are that the original author has too much else going on to devote time at the moment. If you want to see it move forward, by all means more it forward. Send patches. However, keep in mind that the original author/maintainer still holds this role. If you think things are getting to the point of a release, send them an email and let them know. They'll most likely say sure, go for it, or let you know what they think needs to be done before a release. After a bit, they may even say hey, i'm too busy these days, you've been doing well, its up to you. Now, if after a few months of contributing, and working on a project you STILL haven't heard from the original maintainer, then sure, we'll let you decide when to release, etc. If we just keep civil and communicate a bit, everyone will be happy. We'll get more young blood (which is always needed), without offending and pushing away the older workin' folk with wives and girlfriends and other commitments keeping their noses out of e-code. :) -- rephorm --- This SF.net email is sponsored by: Splunk Inc. Do you grep through log files for problems? Stop! Download the new AJAX search engine that makes searching your log files as easy as surfing the web. DOWNLOAD SPLUNK! http://ads.osdn.com/?ad_id=7637alloc_id=16865op=click ___ enlightenment-devel mailing list enlightenment-devel@lists.sourceforge.net https://lists.sourceforge.net/lists/listinfo/enlightenment-devel
Re: [E-devel] Some basic contributing guidelins
Brian Mattern wrote: Ok, it looks like we just need to lay out some simple guidelines to make everyone happy. Contributing code is ALWAYS welcome. Write your code, then use 'cvs diff -u' to create a patch. Attach it to an email to this list. If you haven't heard back after a few days, send another response. Or, send a personal email to the first listed AUTHOR for the package. (Sometimes people are busy and don't read the edevel stuff, but most of us feel naked when away from email for more than a day). By this point, if you still haven't heard from the author, most likely some one else will have reviewed/committed your patch. I never heard anything about the ecore_config patch I sent in.. noone even told me it sucked :) -- Morten --- This SF.net email is sponsored by: Splunk Inc. Do you grep through log files for problems? Stop! Download the new AJAX search engine that makes searching your log files as easy as surfing the web. DOWNLOAD SPLUNK! http://ads.osdn.com/?ad_id=7637alloc_id=16865op=click ___ enlightenment-devel mailing list enlightenment-devel@lists.sourceforge.net https://lists.sourceforge.net/lists/listinfo/enlightenment-devel
Re: [E-devel] Some basic contributing guidelins
On Tue, 29 Nov 2005 22:40:29 +0100 (CET) Vincent Torri [EMAIL PROTECTED] wrote: On Tue, 29 Nov 2005, Brian Mattern wrote: On Tuesday 29 November 2005 15:05, Morten Nilsen wrote: Brian Mattern wrote: Ok, it looks like we just need to lay out some simple guidelines to make everyone happy. Contributing code is ALWAYS welcome. Write your code, then use 'cvs diff -u' to create a patch. Attach it to an email to this list. If you haven't heard back after a few days, send another response. Or, send a personal email to the first listed AUTHOR for the package. (Sometimes people are busy and don't read the edevel stuff, but most of us feel naked when away from email for more than a day). By this point, if you still haven't heard from the author, most likely some one else will have reviewed/committed your patch. I never heard anything about the ecore_config patch I sent in.. noone even told me it sucked :) Oooh slipped by us. :) (Really, I think we need something better than just the list for tracking patches, but, that's another story...) bugzilla ? Enlightenment is a sourceforge project, they supply trackers for all sorts of things. pgpfArxiBf8O5.pgp Description: PGP signature
Re: [E-devel] Some basic contributing guidelins
On 11/29/05, Morten Nilsen [EMAIL PROTECTED] wrote: In my opinion, however, I don't see why one wouldn't want dir sort..Cheers,--Morten Are you referring to ecore_file_ls?
[E-devel] Maximize ideas
Hi, I've some ideas about maximize windows and I like to hear your ideas or if this is on a TODO list. I categorize my E17 modules into two categories. The navigating and application modules and the less important modules like e.g. volume or clock. So I've a problem with maximizing windows. I'm not happy the the currently available maximize policies. They all care about desktop modules. So if I've a big clock, maximizing a window doesn't enlarge the window very much. The fullscreen policy fits most to my needs, but the the ibar, ibox and pager are overlaped. So I like a fullscreen maximize with the exception to not overlap the lowest ~40 pixes of my screen. The same problem with window placement. I don't like E to place a window at the lowest ~40 pixel at my screen. This overlaps the navigation modules too. I know each person place this icons at another place, but many people like not to overlap their navigation modules. So how to solve this situation? I used xfce4 for some time, before I switched to E. There's a configuration dialog where you could set some never overlap areas for the window manager. Is this an idea for E? See this shot of the xfce4 configuration dialog: http://www.loculus.nl/xfce/documentation/docs-4.2/images/workspace_margins.png regards Andreas --- This SF.net email is sponsored by: Splunk Inc. Do you grep through log files for problems? Stop! Download the new AJAX search engine that makes searching your log files as easy as surfing the web. DOWNLOAD SPLUNK! http://ads.osdn.com/?ad_id=7637alloc_id=16865op=click ___ enlightenment-devel mailing list enlightenment-devel@lists.sourceforge.net https://lists.sourceforge.net/lists/listinfo/enlightenment-devel
Re: [E-devel] eVolume 0.0.17a has been released
Am Wed, 30 Nov 2005 00:17:16 +0400 schrieb Pavel Boldin: Hi all! I again with you :), so made new release: http://paq.osdn.org.ua/~davinchi/evolume-0.0.17a.tar.bz2 as you can guess, `a' is stays for alpha, so don't surprise, if it segaultfs Ahh, some minutes ago I build a Gentoo ebuild for 0.0.14. Is there a E17 Gentoo packager on this list? Should I send it to bugs.gentoo.org after upgrading to 0.0.17a? It seems to work great. Or should I paste the script here on the list? regards Andreas --- This SF.net email is sponsored by: Splunk Inc. Do you grep through log files for problems? Stop! Download the new AJAX search engine that makes searching your log files as easy as surfing the web. DOWNLOAD SPLUNK! http://ads.osdn.com/?ad_id=7637alloc_id=16865op=click ___ enlightenment-devel mailing list enlightenment-devel@lists.sourceforge.net https://lists.sourceforge.net/lists/listinfo/enlightenment-devel
Re: [E-devel] Some basic contributing guidelins
On Tue, Nov 29, 2005 at 10:05:42PM +0100, Morten Nilsen wrote: Brian Mattern wrote: Ok, it looks like we just need to lay out some simple guidelines to make everyone happy. Contributing code is ALWAYS welcome. Write your code, then use 'cvs diff -u' to create a patch. Attach it to an email to this list. If you haven't heard back after a few days, send another response. Or, send a personal email to the first listed AUTHOR for the package. (Sometimes people are busy and don't read the edevel stuff, but most of us feel naked when away from email for more than a day). By this point, if you still haven't heard from the author, most likely some one else will have reviewed/committed your patch. I never heard anything about the ecore_config patch I sent in.. noone even told me it sucked :) is this the ecore_config lib patch for ARGB parsing? or the ecore_config app patch for listing keys? One is my fault the other isn't ;) Andy -- Morten --- This SF.net email is sponsored by: Splunk Inc. Do you grep through log files for problems? Stop! Download the new AJAX search engine that makes searching your log files as easy as surfing the web. DOWNLOAD SPLUNK! http://ads.osdn.com/?ad_id=7637alloc_id=16865op=click ___ enlightenment-devel mailing list enlightenment-devel@lists.sourceforge.net https://lists.sourceforge.net/lists/listinfo/enlightenment-devel
Re: [E-devel] Some basic contributing guidelins
Many apologies, I do not remember which leak this is that you refer to. I know ecore_config lacks features etc and I will attend to it as soon as my TODO list reaches ecore_config. Engage also needs my love, and it does in the background, simply nothing committed yet. Not sure that asleep at the wheel counts. If nothing else I think that many folk agree that managing the E web site is now a lot easier, and XSM does take a lot of my time - if someone wants to help do let me know! Andy On Tue, Nov 29, 2005 at 04:11:01PM -0500, Michael Jennings wrote: On Tuesday, 29 November 2005, at 22:05:42 (+0100), Morten Nilsen wrote: I never heard anything about the ecore_config patch I sent in.. noone even told me it sucked :) Indeed! If you want to talk about showstoppers and maintainers who are asleep at the wheel, let's talk about HandyAndE and the memory leak he refuses to address that has been keeping Tilman from moving to ecore_config for the past who-knows-how-long! :-P Michael -- Michael Jennings (a.k.a. Moany argument maker) http://www.kainx.org/ [EMAIL PROTECTED] n + 1, Inc., http://www.nplus1.net/ Author, Eterm (www.eterm.org) --- This SF.net email is sponsored by: Splunk Inc. Do you grep through log files for problems? Stop! Download the new AJAX search engine that makes searching your log files as easy as surfing the web. DOWNLOAD SPLUNK! http://ads.osdn.com/?ad_id=7637alloc_id=16865op=click ___ enlightenment-devel mailing list enlightenment-devel@lists.sourceforge.net https://lists.sourceforge.net/lists/listinfo/enlightenment-devel
[E-devel] ecore_config list implementation (Was: Some basic contributing guidelins)
dan sinclair wrote:
no.. I'm referring to my patch implementing ecore_config list :)
Can you just resend the patch and make this easier for everyone.
sure thing..
Cheers,
--
Morten
Index: ecore_config.c
===
RCS file: /cvsroot/enlightenment/e17/libs/ecore/src/bin/ecore_config.c,v
retrieving revision 1.3
diff -u -r1.3 ecore_config.c
--- ecore_config.c 18 Sep 2005 12:48:24 - 1.3
+++ ecore_config.c 6 Nov 2005 16:34:14 -
@@ -8,6 +8,70 @@
#ifdef BUILD_ECORE_CONFIG
#include Ecore_Config.h
+
+#define OVERFLOW_CHECK
+
+#define LIMIT_SUPPORTED_KEYS 1000
+
+#if 1
+#define FILE_MAX 512
+
+#define BUBBLE_SORT_CMP pathcmp
+// strcmp for paths - for sorting folders before files
+int pathcmp(const char *s1, const char *s2)
+{
+ int i = 0, j = 0, d = 0;
+ char *s1d, *s2d;
+ s1d = malloc(FILE_MAX);
+ s2d = malloc(FILE_MAX);
+
+ while(*s1 *s2) {
+
+ if(*s1 == '/') s1++;
+ i = 0;
+ while(*s1 *s1 != '/') {
+ s1d[i++] = *s1++;
+#ifdef OVERFLOW_CHECK
+ if(i=FILE_MAX) {
+ while(*s1 *s1!='/') s2++;
+ break;
+ }
+#endif
+ }
+
+ if(*s2 == '/') s2++;
+ j = 0;
+ while(*s2 *s2 != '/') {
+ s2d[j++] = *s2++;
+#ifdef OVERFLOW_CHECK
+ if(j=FILE_MAX) {
+ while(*s2 *s2!='/') s2++;
+ break;
+ }
+#endif
+ }
+
+ // These returns order folders before files
+ if(!*s1 *s2 == '/') return 1;
+ if(!*s2 *s1 == '/') return -1;
+
+ // Perform regular strcmp of nodes
+ s1d[i] = 0;
+ s2d[j] = 0;
+ d = strcmp(s1d, s2d);
+
+ if(d != 0) {
+ return d;
+ }
+ }
+ // We should never come here...
+ fprintf(stderr,Equal path!);
+ return 0;
+}
+#else
+#define BUBBLE_SORT_CMP strcmp
+#endif
+
int
set(const char *key, int ec_type, const char *value)
{
@@ -49,22 +113,22 @@
printf(\n);
break;
case ECORE_CONFIG_INT:
- printf(%ld\n, ecore_config_int_get(key));
+ printf(int\t%ld\n, ecore_config_int_get(key));
break;
case ECORE_CONFIG_BLN:
- printf(%d\n, ecore_config_boolean_get(key));
+ printf(bool\t%d\n, ecore_config_boolean_get(key));
break;
case ECORE_CONFIG_FLT:
- printf(%lf\n, ecore_config_float_get(key));
+ printf(float\t%lf\n, ecore_config_float_get(key));
break;
case ECORE_CONFIG_STR:
- printf(%s\n, ecore_config_string_get(key));
+ printf(string\t\%s\\n, ecore_config_string_get(key));
break;
case ECORE_CONFIG_RGB:
- printf(%s\n, ecore_config_argbstr_get(key));
+ printf(rgb\t\%s\\n, ecore_config_argbstr_get(key));
break;
case ECORE_CONFIG_THM:
- printf(%s\n, ecore_config_theme_get(key));
+ printf(theme\t\%s\\n, ecore_config_theme_get(key));
break;
default:
fprintf(stderr, Property has unrecognised type);
@@ -76,8 +140,42 @@
int
list(const char *file)
{
- fprintf(stderr, Command not yet supported\n);
- return -1;
+ Ecore_Config_Prop *e;
+ int n = 0, x, y;
+ char **keys;
+ char *temp;
+
+ keys = malloc(LIMIT_SUPPORTED_KEYS);
+
+ e = __ecore_config_bundle_local-data;
+
+ do {
+ keys[n++] = e-key;
+#ifdef OVERFLOW_CHECK
+ if(n = LIMIT_SUPPORTED_KEYS) {
+ fprintf(stderr, Config file exceeds key limit\n);
+ break;
+ }
+#endif
+ } while((e = e-next));
+
+ // Bubble sort
+ for(x=0; x n; ++x) {
+ for(y=0; y n-1; ++y) {
+ if(BUBBLE_SORT_CMP(keys[y],keys[y+1])0) {
+ temp = keys[y+1];
+ keys[y+1] = keys[y];
+ keys[y] = temp;
+ }
+ }
+ }
+
+ for(x=0; x n; ++x) {
+ printf(%-28s\t, keys[x]);
+ get(keys[x]);
+ }
+
+ return 0;
}
int
@@ -154,6 +252,8 @@
int
main(int argc, const char **argv)
{
+ Ecore_Config_Bundle *t;
+ Ecore_Config_Prop *e;
const char *prog, *file, *cmd, *key, *type, *value;
int ec_type = -1;
int ret = 0;
@@ -194,8 +294,15 @@
}
}
- // Load configuration
Re: Re: [E-devel] Some basic contributing guidelins
On that note, here is a resend for something I sent back at October (trivial and easily forgotten :-))
This wee one liner just centers the optional digital text string for e's clock module (instead of left aligning it).
Index: apps/e/data/themes/default_clock.edc
===
RCS file: /cvsroot/enlightenment/e17/apps/e/data/themes/default_clock.edc,v
retrieving revision 1.8
diff -u -u -r1.8 default_clock.edc
---
apps/e/data/themes/default_clock.edc
24 Sep 2005 01:15:23 - 1.8
+++ apps/e/data/themes/default_clock.edc 21 Oct 2005 18:30:08 -
@@ -693,7 +693,7 @@
font: Edje-Vera;
size: 15;
fit: 0 1;
- align: 0.0 0.5;
+ align: 0.5 0.5;
}
}
description {
Regards!
David
Re: [E-devel] Some basic contributing guidelins
Andrew Williams wrote: is this the ecore_config lib patch for ARGB parsing? according to my archive, that patch came from Dylan Shell (and it got comments :) that patch also seems to be as simple as s/strtol/strtoul/ :) or the ecore_config app patch for listing keys? that one was by me, yes One is my fault the other isn't ;) we don't need any scapegoats, do we? ;) Cheers, -- Morten --- This SF.net email is sponsored by: Splunk Inc. Do you grep through log files for problems? Stop! Download the new AJAX search engine that makes searching your log files as easy as surfing the web. DOWNLOAD SPLUNK! http://ads.osdn.com/?ad_id=7637alloc_id=16865op=click ___ enlightenment-devel mailing list enlightenment-devel@lists.sourceforge.net https://lists.sourceforge.net/lists/listinfo/enlightenment-devel
Re: [E-devel] Some basic contributing guidelins
On Tuesday, 29 November 2005, at 21:49:24 (+), Andrew Williams wrote: Many apologies, I do not remember which leak this is that you refer to. Allow me to refresh your memory...again: 1251 KainX[#edevelop] tilman: I'd like to see it imported so that edb_gtk_ed in libs/edb can die die die die die. 1251 tilman[#edevelop] although. 1251 tilman[#edevelop] embrace is the only app that still needs edb : 1251 dj2[#edevelop] if we can kill of edb, no reason for an editor, heh 1251 KainX[#edevelop] EET doesn't use edb? Nor epeg? 1251 tilman[#edevelop] i won't switch to ecore_config until it works properly 1251 tilman[#edevelop] KainX: o_O 1251 tilman[#edevelop] KainX: they don't 1252 tilman[#edevelop] epeg modifies jpeg headers 1252 tilman[#edevelop] eet has its own data format 1252 KainX[#edevelop] And everything else uses eet... 1252 KainX[#edevelop] What does embrace use edb for? 1252 dj2[#edevelop] tilman, what doesnt' work for you? 1252 tilman[#edevelop] atmos: now look at the beautiful code and WEEP 1252 dj2[#edevelop] KainX, except etox 1252 tilman[#edevelop] KainX: configuration 1252 tilman[#edevelop] dj2: last i checked it still leaked 1252 KainX[#edevelop] Ah 1252 dj2[#edevelop] ah 1252 tilman[#edevelop] non-trivial to fix 1253 tilman[#edevelop] so i won't do it 1253 tilman[#edevelop] and i won't depend on a leaky lib 1253 atmos[#edevelop] tilman it's all redact right ? 1253 tilman[#edevelop] redact and of course embryo_cc 1255 KainX[#edevelop] tilman: Is raster aware of the leak? 1255 tilman[#edevelop] iirc, yes 1256 KainX[#edevelop] Does he plan to fix it? 1256 englebass[#edevelop] Isn't HandyAndE the ecore-config man? 1256 tilman[#edevelop] englebass: yes, but he kinda ignored me/shrugged off the issue 1257 KainX[#edevelop] tilman: HandyAndE did, or raster did? 1257 tilman[#edevelop] KainX: andy (from August 30th) Not sure that asleep at the wheel counts. If nothing else I think that many folk agree that managing the E web site is now a lot easier, and XSM does take a lot of my time Content management systems are a dime a dozen. Any CMS would've been an improvement, and I don't see anything XSM offers over the more mature alternatives. But that's really beside the point. Michael Jennings (a.k.a. Moany argument maker) That's just sad. Michael -- Michael Jennings (a.k.a. KainX) http://www.kainx.org/ [EMAIL PROTECTED] n + 1, Inc., http://www.nplus1.net/ Author, Eterm (www.eterm.org) --- Daddy, what is virginity? It's the state below Maryland. -- Courtney Chase and Dan Aykroyd, Soul Man --- This SF.net email is sponsored by: Splunk Inc. Do you grep through log files for problems? Stop! Download the new AJAX search engine that makes searching your log files as easy as surfing the web. DOWNLOAD SPLUNK! http://ads.osdn.com/?ad_id=7637alloc_id=16865op=click ___ enlightenment-devel mailing list enlightenment-devel@lists.sourceforge.net https://lists.sourceforge.net/lists/listinfo/enlightenment-devel
RE: [E-devel] ecore_config list implementation (Was: Some basic contributing guidelins)
dan sinclair wrote: no.. I'm referring to my patch implementing ecore_config list :) Can you just resend the patch and make this easier for everyone. sure thing.. Why the 1000 key limit? Is that an ecore_config limit or something that you just made up? Bubblesort may not be a good choice for this as it maybe too slow. One option is to use the ecore heap implementation and dump it to a list. EWL does this to sort files I believe. It's nice and fast. (And since we're in ecore we can probably use ecore_data.) Also looks like your forcing a max key length to 512 in the case of pathcmp correct? Is this an ecore_config limit or something your imposing. Would be nice to get rid of the 1000 key limit as all of the OVERFLOW code is pretty gross looking in there. dan --- This SF.net email is sponsored by: Splunk Inc. Do you grep through log files for problems? Stop! Download the new AJAX search engine that makes searching your log files as easy as surfing the web. DOWNLOAD SPLUNK! http://ads.osdn.com/?ad_id=7637alloc_id=16865op=click ___ enlightenment-devel mailing list enlightenment-devel@lists.sourceforge.net https://lists.sourceforge.net/lists/listinfo/enlightenment-devel
Re: [E-devel] Maximize ideas
On Tue, 29 Nov 2005 23:18:32 +0100 Andreas Volz [EMAIL PROTECTED] wrote: Hi, I've some ideas about maximize windows and I like to hear your ideas or if this is on a TODO list. I categorize my E17 modules into two categories. The navigating and application modules and the less important modules like e.g. volume or clock. So I've a problem with maximizing windows. I'm not happy the the currently available maximize policies. They all care about desktop modules. So if I've a big clock, maximizing a window doesn't enlarge the window very much. The fullscreen policy fits most to my needs, but the the ibar, ibox and pager are overlaped. I've had the sae problem... I think that being able to tell E to ignore certain modules when determining size for SMART maximize policy would be useful. So I like a fullscreen maximize with the exception to not overlap the lowest ~40 pixes of my screen. The same problem with window placement. I don't like E to place a window at the lowest ~40 pixel at my screen. This overlaps the navigation modules too. I know each person place this icons at another place, but many people like not to overlap their navigation modules. So how to solve this situation? I used xfce4 for some time, before I switched to E. There's a configuration dialog where you could set some never overlap areas for the window manager. Is this an idea for E? Something like -maximize-policy-set MANUAL, maybe, along with a few commands to set the margins should be relatively easy I would think. See this shot of the xfce4 configuration dialog: http://www.loculus.nl/xfce/documentation/docs-4.2/images/workspace_margins.png regards Andreas Jesse --- This SF.net email is sponsored by: Splunk Inc. Do you grep through log files for problems? Stop! Download the new AJAX search engine that makes searching your log files as easy as surfing the web. DOWNLOAD SPLUNK! http://ads.osdn.com/?ad_id=7637alloc_id=16865op=click ___ enlightenment-devel mailing list enlightenment-devel@lists.sourceforge.net https://lists.sourceforge.net/lists/listinfo/enlightenment-devel
Re: [E-devel] Some basic contributing guidelins
On Tue, Nov 29, 2005 at 05:57:47PM -0500, Michael Jennings wrote: On Tuesday, 29 November 2005, at 21:49:24 (+), Andrew Williams wrote: Many apologies, I do not remember which leak this is that you refer to. Allow me to refresh your memory...again: 1251 KainX[#edevelop] tilman: I'd like to see it imported so that edb_gtk_ed in libs/edb can die die die die die. 1251 tilman[#edevelop] although. 1251 tilman[#edevelop] embrace is the only app that still needs edb : 1251 dj2[#edevelop] if we can kill of edb, no reason for an editor, heh 1251 KainX[#edevelop] EET doesn't use edb? Nor epeg? 1251 tilman[#edevelop] i won't switch to ecore_config until it works properly 1251 tilman[#edevelop] KainX: o_O 1251 tilman[#edevelop] KainX: they don't 1252 tilman[#edevelop] epeg modifies jpeg headers 1252 tilman[#edevelop] eet has its own data format 1252 KainX[#edevelop] And everything else uses eet... 1252 KainX[#edevelop] What does embrace use edb for? 1252 dj2[#edevelop] tilman, what doesnt' work for you? 1252 tilman[#edevelop] atmos: now look at the beautiful code and WEEP 1252 dj2[#edevelop] KainX, except etox 1252 tilman[#edevelop] KainX: configuration 1252 tilman[#edevelop] dj2: last i checked it still leaked 1252 KainX[#edevelop] Ah 1252 dj2[#edevelop] ah 1252 tilman[#edevelop] non-trivial to fix 1253 tilman[#edevelop] so i won't do it 1253 tilman[#edevelop] and i won't depend on a leaky lib 1253 atmos[#edevelop] tilman it's all redact right ? 1253 tilman[#edevelop] redact and of course embryo_cc 1255 KainX[#edevelop] tilman: Is raster aware of the leak? 1255 tilman[#edevelop] iirc, yes 1256 KainX[#edevelop] Does he plan to fix it? 1256 englebass[#edevelop] Isn't HandyAndE the ecore-config man? 1256 tilman[#edevelop] englebass: yes, but he kinda ignored me/shrugged off the issue 1257 KainX[#edevelop] tilman: HandyAndE did, or raster did? 1257 tilman[#edevelop] KainX: andy (from August 30th) well, I sincerely apologise for not remembering this, but even after the refresher I do not recall what the problem was - does someone have a specific point, or is an in general it leaks issue? Not sure that asleep at the wheel counts. If nothing else I think that many folk agree that managing the E web site is now a lot easier, and XSM does take a lot of my time Content management systems are a dime a dozen. Any CMS would've been an improvement, and I don't see anything XSM offers over the more mature alternatives. But that's really beside the point. dime a dozen hmm, weird - I seem to remember that sf.net caused no end of trouble with their no web server can write to web space directive. List a dozen that work well with sf.net space and I might consider giving you a dime :-p Michael Jennings (a.k.a. Moany argument maker) That's just sad. nothing better to do with my time, clearly :-p Michael A --- This SF.net email is sponsored by: Splunk Inc. Do you grep through log files for problems? Stop! Download the new AJAX search engine that makes searching your log files as easy as surfing the web. DOWNLOAD SPLUNK! http://ads.osdn.com/?ad_id=7637alloc_id=16865op=click ___ enlightenment-devel mailing list enlightenment-devel@lists.sourceforge.net https://lists.sourceforge.net/lists/listinfo/enlightenment-devel
Re: [E-devel] Some basic contributing guidelins
On Tue, Nov 29, 2005 at 11:40:25PM +0100, Morten Nilsen wrote: Andrew Williams wrote: is this the ecore_config lib patch for ARGB parsing? according to my archive, that patch came from Dylan Shell (and it got comments :) that patch also seems to be as simple as s/strtol/strtoul/ :) aha, found in the archives - I was never sure how simple it was, I have applied it in bare simplicity with the expectation that I got it wrong, but at least it is in! or the ecore_config app patch for listing keys? that one was by me, yes aha, I remember looking at it and thinking who wrote that ecore_config app ??? One is my fault the other isn't ;) we don't need any scapegoats, do we? ;) strange, I thought that's what folk wanted ;) Cheers, -- Morten n.p. Andy --- This SF.net email is sponsored by: Splunk Inc. Do you grep through log files for problems? Stop! Download the new AJAX search engine that makes searching your log files as easy as surfing the web. DOWNLOAD SPLUNK! http://ads.osdn.com/?ad_id=7637alloc_id=16865op=click ___ enlightenment-devel mailing list enlightenment-devel@lists.sourceforge.net https://lists.sourceforge.net/lists/listinfo/enlightenment-devel
RE: Re: [E-devel] Some basic contributing guidelins
We already have Mantis set up that's not used :/ mantis.xcomputerman.com or something. xcomputerman.com/bugs --- This SF.net email is sponsored by: Splunk Inc. Do you grep through log files for problems? Stop! Download the new AJAX search engine that makes searching your log files as easy as surfing the web. DOWNLOAD SPLUNK! http://ads.osdn.com/?ad_id=7637alloc_id=16865op=click ___ enlightenment-devel mailing list enlightenment-devel@lists.sourceforge.net https://lists.sourceforge.net/lists/listinfo/enlightenment-devel
Re: [E-devel] Maximize ideas
Hi all, I've the same problem. Recently I implemented some maximization mechanizms. See my post :::))) fro this. I'm waiting for any reply still. Moreover, I'm waiting for my changes to be included into the recent cvs. If I see, that my code changes are accepted, I'll will start to improve maximization mechanizm. Best regards, aleksejOn 11/30/05, Jesse Luehrs [EMAIL PROTECTED] wrote: On Tue, 29 Nov 2005 23:18:32 +0100Andreas Volz [EMAIL PROTECTED] wrote: Hi, I've some ideas about maximize windows and I like to hear your ideas or if this is on a TODO list. I categorize my E17 modules into two categories. The navigating and application modules and the less important modules like e.g. volume or clock. So I've a problem with maximizing windows. I'm not happy the the currently available maximize policies. They all care about desktop modules. So if I've a big clock, maximizing a window doesn't enlarge the window very much. The fullscreen policy fits most to my needs, but the the ibar, ibox and pager are overlaped.I've had the sae problem... I think that being able to tell E to ignorecertain modules when determining size for SMART maximize policy would be useful. So I like a fullscreen maximize with the exception to not overlap the lowest ~40 pixes of my screen. The same problem with window placement. I don't like E to place a window at the lowest ~40 pixel at my screen. This overlaps the navigation modules too. I know each person place this icons at another place, but many people like not to overlap their navigation modules. So how to solve this situation? I used xfce4 for some time, before I switched to E. There's a configuration dialog where you could set some never overlap areas for the window manager. Is this an idea for E? Something like -maximize-policy-set MANUAL, maybe, along with a fewcommands to set the margins should be relatively easy I would think. See this shot of the xfce4 configuration dialog: http://www.loculus.nl/xfce/documentation/docs-4.2/images/workspace_margins.png regards AndreasJesse---This SF.net email is sponsored by: Splunk Inc. Do you grep through log files for problems?Stop!Download the new AJAX search engine that makessearching your log files as easy as surfing theweb.DOWNLOAD SPLUNK! http://ads.osdn.com/?ad_id=7637alloc_id=16865op=click___enlightenment-devel mailing listenlightenment-devel@lists.sourceforge.net https://lists.sourceforge.net/lists/listinfo/enlightenment-devel-- Aleksej Struk Master Degree StudentFree University of Bozen-BolzanoFaculty of Computer Sciencephone: +39-0471-061749cell phone: +39-3204627049[EMAIL PROTECTED] [EMAIL PROTECTED] - http://www.
Re: Re: [E-devel] Some basic contributing guidelins
as per bug #290 - do we want to link this bug tracker to the e website? or is it just internal and belonging on edevelop? /A On Tue, Nov 29, 2005 at 06:15:39PM -0500, dan sinclair wrote: We already have Mantis set up that's not used :/ mantis.xcomputerman.com or something. xcomputerman.com/bugs --- This SF.net email is sponsored by: Splunk Inc. Do you grep through log files for problems? Stop! Download the new AJAX search engine that makes searching your log files as easy as surfing the web. DOWNLOAD SPLUNK! http://ads.osdn.com/?ad_id=7637alloc_id=16865op=click ___ enlightenment-devel mailing list enlightenment-devel@lists.sourceforge.net https://lists.sourceforge.net/lists/listinfo/enlightenment-devel --- This SF.net email is sponsored by: Splunk Inc. Do you grep through log files for problems? Stop! Download the new AJAX search engine that makes searching your log files as easy as surfing the web. DOWNLOAD SPLUNK! http://ads.osdn.com/?ad_id=7637alloc_id=16865op=click ___ enlightenment-devel mailing list enlightenment-devel@lists.sourceforge.net https://lists.sourceforge.net/lists/listinfo/enlightenment-devel
Re: [E-devel] specifying user environment in Entrance
On Tue, 29 Nov 2005 15:47:11 -0500 Michael Jennings [EMAIL PROTECTED] babbled: I'm afraid this thread is taking on a life of its own like some of the previous ones. So I'm going to try to trim a bit. On Tuesday, 29 November 2005, at 16:10:14 (+0900), Carsten Haitzler wrote: yes - but its not a temporary condition. xcomps life changed ina way where he'll be ultra-busy for many years to come. if you talk with him. i'll let him divulge his life's details. and it shows that entrance basically has fallen into dormant status ever since. My point is that it's his responsibility to give up the reigns or simply take a less active development role (and more management). It's his choice. he HAS taken a less active role. thats my whole point! by sheer virtue of actions. and if the person who owns it vanishes? stops working on it? has no time? then it sits and rots (well just goes nowehere)? and no one is allowed to touch it without his express approval? come on! if someone wants to pick up the ball and keep runing wiht it - or more than 1 person. - why not? the original author doesnt lose credit. he still has the full respect and admiration for having picke dup the ball and run with it as far as they did. now a new set of people keep runing with it. you seem to have something against that? Not at all. Anyone with CVS access can contribute. I have no problem with that. All I'm saying is that the person in charge remains the person in charge until one of the following happens: (1) (S)he steps aside willingly; (2) (S)he is inexplicibly out of touch via e-mail and IRC for a significant period of time. OR in the event they just go minimally active (do no significant work on a project that needs work - see previous mails of mine), and thus become a block to that project moving forward. Again, I'm not talking about contributors (i.e., people who commit changes to CVS). I'm talking about the lead developer. The person in charge. The designer, architect, and ultimate decision maker. For E, Evas, etc., that's you. For Eterm, it's me. For EWL, it's RbdPngn. For entrance, it's xcomp. For Etk, it's Simon. And so forth... sure - but xcomp has - per his own words said that he doesnt have time. see my previous mails and comments. the proof of that is lookng at entrances commit history for the last 12 months. i read cvs commit logs. every day. i build a mental maqp of roughly who does what and where. please prove to me xcomp is active - that means non trivial 1 liners coming from email comments most likely or a quick fix for a problem found that annoys him or someone else, or a submitted patch? well that reminds me. need to add myself to AUTHORS - as i have done work on entrance - a reasonable amount - enough to justify an AUTHORS entry... u can check cvs logs. thus as an AUTHOR i belive i have a little more weight to pull in this. i never saw you jump up and down when i worked on entrance beofre? do you see me making a fuss when YOU just commit things to spec and makefiles etc.? I think we're talking about two different things here. Anyone with commit access can contribute; I've never said anything about that. Hopefully my statements above will clarify. and if you now have 1 person doing 90% of the work... do they not get to call when to release? or do you leave it to someone inactive? if they are the only one working on it - which they will be, You're making a rather large assumption there. ones based on history for the past 6+ months of entrance's life. see the last time there was a thread on this. they get to cal the shots as to when to make a release, what features to put in, how to code them, etc. what is wrong with that? What's wrong with it is that it's not their project, and as long as the original maintainer is still around, they do not forfeit their rights. If the original maintainer says no release, there will be no release. The contributor has two options at that point: accept the no release decree or fork the project. Otherwise things get very messy. so leadwer says no release, and then does nothing. doesnt say no release because i am working on x, y and z and it will be in soon or no release because person b and c over here are working on x, y, and z or no release because we want to syncronise with release of project h, i and j - just say no release... that to me - says dead project. 100% dead. what i'm saying is - entrance has dropped off getting any development (worth talking about) for at least 6-12 months now. xcomp does not have the time to work on it. until he comes back and lets say takes part in this and says hey - duded - look. i was busy - that will end, i will be active again really soon he has by sheer virtue of being inactive made room for people to take over. so old owner says no - go away. don't touch it. ok. fine. now it gets no improvements as old owner is doing nothing. it gets no
Re: [E-devel] Some basic contributing guidelins
On Tue, 29 Nov 2005 23:40:25 +0100 Morten Nilsen [EMAIL PROTECTED] babbled: Andrew Williams wrote: is this the ecore_config lib patch for ARGB parsing? according to my archive, that patch came from Dylan Shell (and it got comments :) that patch also seems to be as simple as s/strtol/strtoul/ :) or the ecore_config app patch for listing keys? that one was by me, yes One is my fault the other isn't ;) we don't need any scapegoats, do we? ;) we already have one. andy! :) i blame him for anything that takes my fancy. works well for me! :) hehehehe -- - Codito, ergo sum - I code, therefore I am -- The Rasterman (Carsten Haitzler)[EMAIL PROTECTED] 裸好多 Tokyo, Japan (東京 日本) --- This SF.net email is sponsored by: Splunk Inc. Do you grep through log files for problems? Stop! Download the new AJAX search engine that makes searching your log files as easy as surfing the web. DOWNLOAD SPLUNK! http://ads.osdn.com/?ad_id=7637alloc_id=16865op=click ___ enlightenment-devel mailing list enlightenment-devel@lists.sourceforge.net https://lists.sourceforge.net/lists/listinfo/enlightenment-devel
Re: [E-devel] Some basic contributing guidelins
On Tue, 29 Nov 2005 15:48:29 -0600 Brian Mattern [EMAIL PROTECTED] babbled: On Tuesday 29 November 2005 15:05, Morten Nilsen wrote: Brian Mattern wrote: Ok, it looks like we just need to lay out some simple guidelines to make everyone happy. Contributing code is ALWAYS welcome. Write your code, then use 'cvs diff -u' to create a patch. Attach it to an email to this list. If you haven't heard back after a few days, send another response. Or, send a personal email to the first listed AUTHOR for the package. (Sometimes people are busy and don't read the edevel stuff, but most of us feel naked when away from email for more than a day). By this point, if you still haven't heard from the author, most likely some one else will have reviewed/committed your patch. I never heard anything about the ecore_config patch I sent in.. noone even told me it sucked :) Oooh slipped by us. :) (Really, I think we need something better than just the list for tracking patches, but, that's another story...) you're telling me. my current system is to mark any mail with a patch or some important issue and leave it in my inboxes until i finally flush this stuff out. the bigger the patch, the less likely i am to get to flushing. it doesn't scale well - but however u look at it - i cant think of a better way. patches just take a lot of time to read. what i have been doing is trying to DROP working on anything i'm not mother-henning over (right now i mother-hen over eet, embryo, edje, evas, ecore, e17 - i pretty much leave everything else alone). within ecore, i leave ecore_config to andy as he pretty much knows the code the best. the problem is - if you are looking after a chunk of code - if you get patches it takes time. the only way to solve it is if multiple people mother-hen over the code, review the patch and then the workload is split. I'll take a look at it tonight if no one beats me to it. On a quick glance, sorting directory first should probably be made an options. I know I've seen contention on the list over which way is 'correct' before... -- rephorm --- This SF.net email is sponsored by: Splunk Inc. Do you grep through log files for problems? Stop! Download the new AJAX search engine that makes searching your log files as easy as surfing the web. DOWNLOAD SPLUNK! http://ads.osdn.com/?ad_id=7637alloc_id=16865op=click ___ 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] 裸好多 Tokyo, Japan (東京 日本) --- This SF.net email is sponsored by: Splunk Inc. Do you grep through log files for problems? Stop! Download the new AJAX search engine that makes searching your log files as easy as surfing the web. DOWNLOAD SPLUNK! http://ads.osdn.com/?ad_id=7637alloc_id=16865op=click ___ enlightenment-devel mailing list enlightenment-devel@lists.sourceforge.net https://lists.sourceforge.net/lists/listinfo/enlightenment-devel
Re: [E-devel] Some basic contributing guidelins
On Tue, 29 Nov 2005 23:09:02 +0100 Tilman Sauerbeck [EMAIL PROTECTED] babbled: Vincent Torri [EMAIL PROTECTED] [2005-11-29 22:40]: [patch] Oooh slipped by us. :) (Really, I think we need something better than just the list for tracking patches, but, that's another story...) bugzilla ? We already have Mantis set up that's not used :/ mantis.xcomputerman.com or something. yeah. i find it isn't THAT useful. well it has issues that make some things easier, and then other things harder. bugs assigned to the wrong thing (ok just bad bug reporting really) and my usual complain of anything put on a web gui that makes it slow and cumbersome by virtue of the medium :) -- - Codito, ergo sum - I code, therefore I am -- The Rasterman (Carsten Haitzler)[EMAIL PROTECTED] 裸好多 Tokyo, Japan (東京 日本) --- This SF.net email is sponsored by: Splunk Inc. Do you grep through log files for problems? Stop! Download the new AJAX search engine that makes searching your log files as easy as surfing the web. DOWNLOAD SPLUNK! http://ads.osdn.com/?ad_id=7637alloc_id=16865op=click ___ enlightenment-devel mailing list enlightenment-devel@lists.sourceforge.net https://lists.sourceforge.net/lists/listinfo/enlightenment-devel
Re: [E-devel] Maximize ideas
On Wed, 30 Nov 2005 00:17:51 +0100 Aleksej Struk [EMAIL PROTECTED] babbled: Hi all, I've the same problem. Recently I implemented some maximization mechanizms. See my post :::))) fro this. I'm waiting for any reply still. Moreover, I'm waiting for my changes to be included into the recent cvs. If I see, that my code changes are accepted, I'll will start to improve maximization mechanizm. like all patches... i've marked it and it's on my list... hopefulyl to get to... sometime... before i die :) Best regards, aleksej On 11/30/05, Jesse Luehrs [EMAIL PROTECTED] wrote: On Tue, 29 Nov 2005 23:18:32 +0100 Andreas Volz [EMAIL PROTECTED] wrote: Hi, I've some ideas about maximize windows and I like to hear your ideas or if this is on a TODO list. I categorize my E17 modules into two categories. The navigating and application modules and the less important modules like e.g. volume or clock. So I've a problem with maximizing windows. I'm not happy the the currently available maximize policies. They all care about desktop modules. So if I've a big clock, maximizing a window doesn't enlarge the window very much. The fullscreen policy fits most to my needs, but the the ibar, ibox and pager are overlaped. I've had the sae problem... I think that being able to tell E to ignore certain modules when determining size for SMART maximize policy would be useful. So I like a fullscreen maximize with the exception to not overlap the lowest ~40 pixes of my screen. The same problem with window placement. I don't like E to place a window at the lowest ~40 pixel at my screen. This overlaps the navigation modules too. I know each person place this icons at another place, but many people like not to overlap their navigation modules. So how to solve this situation? I used xfce4 for some time, before I switched to E. There's a configuration dialog where you could set some never overlap areas for the window manager. Is this an idea for E? Something like -maximize-policy-set MANUAL, maybe, along with a few commands to set the margins should be relatively easy I would think. See this shot of the xfce4 configuration dialog: http://www.loculus.nl/xfce/documentation/docs-4.2/images/workspace_margins.png regards Andreas Jesse --- This SF.net email is sponsored by: Splunk Inc. Do you grep through log files for problems? Stop! Download the new AJAX search engine that makes searching your log files as easy as surfing the web. DOWNLOAD SPLUNK! http://ads.osdn.com/?ad_id=7637alloc_id=16865op=click ___ enlightenment-devel mailing list enlightenment-devel@lists.sourceforge.net https://lists.sourceforge.net/lists/listinfo/enlightenment-devel -- Aleksej Struk Master Degree Student Free University of Bozen-Bolzano Faculty of Computer Science phone: +39-0471-061749 cell phone: +39-3204627049 [EMAIL PROTECTED] [EMAIL PROTECTED] - http://www. -- - Codito, ergo sum - I code, therefore I am -- The Rasterman (Carsten Haitzler)[EMAIL PROTECTED] 裸好多 Tokyo, Japan (東京 日本) --- This SF.net email is sponsored by: Splunk Inc. Do you grep through log files for problems? Stop! Download the new AJAX search engine that makes searching your log files as easy as surfing the web. DOWNLOAD SPLUNK! http://ads.osdn.com/?ad_id=7637alloc_id=16865op=click ___ enlightenment-devel mailing list enlightenment-devel@lists.sourceforge.net https://lists.sourceforge.net/lists/listinfo/enlightenment-devel
Re: [E-devel] ecore_config list implementation (take 3)
dan sinclair wrote:
Why the 1000 key limit? Is that an ecore_config limit or something that you
just made up? Bubblesort may not be a good choice for this as it maybe too
slow. One option is to use the ecore heap implementation and dump it to a list.
EWL does this to sort files I believe. It's nice and fast. (And since we're in
ecore we can probably use ecore_data.)
Also looks like your forcing a max key length to 512 in the case of pathcmp
correct? Is this an ecore_config limit or something your imposing.
Would be nice to get rid of the 1000 key limit as all of the OVERFLOW code is pretty gross looking in there.
here's take 3;
still using bubble sort, but it no longer has arbitrary limits on token
size or number of keys..
like before, comments appreciated :)
Cheers,
--
Morten
Index: ecore_config.c
===
RCS file: /cvsroot/enlightenment/e17/libs/ecore/src/bin/ecore_config.c,v
retrieving revision 1.3
diff -u -r1.3 ecore_config.c
--- ecore_config.c 18 Sep 2005 12:48:24 - 1.3
+++ ecore_config.c 30 Nov 2005 02:33:33 -
@@ -8,6 +8,56 @@
#ifdef BUILD_ECORE_CONFIG
#include Ecore_Config.h
+#ifndef BUBBLE_SORT_CMP
+#define BUBBLE_SORT_CMP pathcmp
+#endif
+
+// strcmp for paths - for sorting folders before files
+int pathcmp(const char *s1, const char *s2)
+{
+int i = 0;
+char *s1d, *s2d;
+
+// alloc potential maximum size of tokens
+s1d = malloc(strlen(s1) + 1);
+s2d = malloc(strlen(s2) + 1);
+
+while(*s1 *s2) {
+
+// skip leading /
+if(*s1 == '/') s1++;
+if(*s2 == '/') s2++;
+
+// get one token from s1
+i = 0;
+while(s1[i++] s1[i] != '/');
+strncpy(s1d, s1, i);
+s1d[i] = 0;
+s1 += i;
+
+// get one token from s2
+i = 0;
+while(s2[i++] s2[i] != '/');
+strncpy(s2d, s2, i);
+s2d[i] = 0;
+s2 += i;
+
+// These returns order folders before files
+if(!*s1 *s2 == '/') return 1;
+if(!*s2 *s1 == '/') return -1;
+
+// Perform regular strcmp of tokens
+i = strcmp(s1d, s2d);
+
+if(i != 0) {
+return i;
+}
+}
+// We should never come here...
+fprintf(stderr,Equal path!);
+return 0;
+}
+
int
set(const char *key, int ec_type, const char *value)
{
@@ -49,22 +99,22 @@
printf(\n);
break;
case ECORE_CONFIG_INT:
- printf(%ld\n, ecore_config_int_get(key));
+ printf(int\t%ld\n, ecore_config_int_get(key));
break;
case ECORE_CONFIG_BLN:
- printf(%d\n, ecore_config_boolean_get(key));
+ printf(bool\t%d\n, ecore_config_boolean_get(key));
break;
case ECORE_CONFIG_FLT:
- printf(%lf\n, ecore_config_float_get(key));
+ printf(float\t%lf\n, ecore_config_float_get(key));
break;
case ECORE_CONFIG_STR:
- printf(%s\n, ecore_config_string_get(key));
+ printf(string\t\%s\\n, ecore_config_string_get(key));
break;
case ECORE_CONFIG_RGB:
- printf(%s\n, ecore_config_argbstr_get(key));
+ printf(rgb\t\%s\\n, ecore_config_argbstr_get(key));
break;
case ECORE_CONFIG_THM:
- printf(%s\n, ecore_config_theme_get(key));
+ printf(theme\t\%s\\n, ecore_config_theme_get(key));
break;
default:
fprintf(stderr, Property has unrecognised type);
@@ -76,8 +126,47 @@
int
list(const char *file)
{
- fprintf(stderr, Command not yet supported\n);
- return -1;
+ Ecore_Config_Prop *e;
+ e = __ecore_config_bundle_local-data;
+
+ int c = 10, n = -1, x, y;
+ char **keys;
+ char *temp;
+
+keys = (char **)malloc(c * sizeof(char *));
+
+ // Collect keys in array for sorting
+ do {
+ // allocate blocks of 10 pointers as needed
+ if(++n c) {
+ c += 10;
+ keys = (char **)realloc(keys, c * sizeof(char *));
+ if(keys == NULL) {
+ fprintf(stderr, Out of memory);
+ return -1;
+ }
+ }
+ keys[n] = e-key;
+ } while((e = e-next));
+
+ // Bubble sort
+ for(x=0; x n; ++x) {
+ for(y=0; y n-1; ++y) {
+ if(BUBBLE_SORT_CMP(keys[y],keys[y+1])0) {
+ temp = keys[y+1];
+ keys[y+1] = keys[y];
+
Re: [E-devel] ecore_config list implementation (Was: Some basic contributing guidelins)
On 11/29/05, dan sinclair [EMAIL PROTECTED] wrote: Why the 1000 key limit? Is that an ecore_config limit or something that you just made up? Bubblesort may not be a good choice for this as it maybe too slow. One option is to use the ecore heap implementation and dump it to a list. EWL does this to sort files I believe. It's nice and fast. (And since we're in ecore we can probably use ecore_data.) It's ecore_file_ls in particular that uses the heap sort. If you know the number of data items prior to reading the whole dataset, you can avoid the initial list and just push the data directly into a heap, then pop off the maximum until it's empty to get a complete sorted list.
