On Sun, 15 Aug 2010 07:14:50 +0200 vto...@univ-evry.fr said: > Yes, I'm upset, because, now, the development is a mess. You, the lead dev, > announce a release (which is great), but you accept that, after that announce, > plenty of people commit plenty of stuff, with maybe a broken API, not > respecting the implicit naming rule (like in eet, for example... But it's too > late now for this lib). Cedric finished his changes on edje yesterday (btw, is > it really finished, tested, etc... ?).
i announced asking people to get anything in that may or WILL break api (or file formats) etc. as that can no longer be done once we go 1.0.0 - never again (until 2.x). if its buggy - it's buggy. it'll now get fixed. the API (and format) is by far the most important thing here. > I would have preferred that kind of announce: > > "Since 2 weeks, the points i mentioned (edje format, etc...) are now in > svn. Do > you guys think that it's mature ? One week for testing and cleaning could be > sufficient. I plan a release one week" we arent talking RELEASE. we are talking ALPHA. it gets to have bugs. i think this is the crux of things. alpha is the "api and format is now done. we think - or are damned sure". as i've already said - i'm not to crash hot on the eina thread stuff as has been argued on irc - it doesnt make eina threadsafe - it makes it rather complex and has to be used in special ways to maybe be threadsafe in some ways. it needs much more thought than that. i'm of the opinion it should be removed for eina 1.0.0. edje file format change is pretty fine - it relies on masses of existing infra that is tested and well known. also cedric has been on this for a while - it's not just a last second thing. i trust it. > Or something like that. And not annoucing the alpha release the day after the > last commit of cedric on the new edje format. I would have agreed without any > problem with that kind of announce. We wait for several years for a release. > One week to check the API and *potential big* problems after such a big change > like the new edje format is important. thats what we have alpha2... alpha3 etc. for. tough to tasn's q - i guess it's going to take me longer than 24-48hrs anyway. real life has sucked away most of my sunday and there's monday coming along to suck more away. if you have/find/know of bugs. squash them now. thats the point. no more adding of features. api's, fiddling with shit endlessly. its time to call and end to that. if something is broken for the windows build - fix it. (yes - i know epdf. i have to go add the things in but its not a 1.0.0 target - it can be done separately). i have a whole bunch of readme files to fix up install files. docs to fix for the build itself - others can too. but if you want5 me to delay this and allow everyone to go fiddle again endlessly for another 2 or 3 or 4 or 20 or 100 weeks, then that's not going to happen. we turn out an alpha as soon as practically possible. then we fix bugs and turn out either an alpha2 or a beta etc. etc. -- ------------- Codito, ergo sum - "I code, therefore I am" -------------- The Rasterman (Carsten Haitzler) ras...@rasterman.com ------------------------------------------------------------------------------ This SF.net email is sponsored by Make an app they can't live without Enter the BlackBerry Developer Challenge http://p.sf.net/sfu/RIM-dev2dev _______________________________________________ enlightenment-devel mailing list enlightenment-devel@lists.sourceforge.net https://lists.sourceforge.net/lists/listinfo/enlightenment-devel
