Re: Criticism of "version numbering"-document
| From: | Tobias Schlitt | Date: | Mon, 08 Mar 2004 19:31:13 +0000 |
| Subject: | Re: Criticism of "version numbering"-document | ||
| References: | 1 2 | Groups: | php.pear.qa |
| Request: | Send a blank email to pear-qa+get-880@lists.php.net to get a copy of this message | ||
Stefan Neufeind wrote:
Well, once you bring up that idea (which I also thought about) there is one more problem: Linux (and others) use even numbers for stable and uneven number for unstable. This way your 2.0.0 would also be a "stable" release, whereas 2.1.0 might be a non-stablee and 2.2.0 the next stable. While this works fine we still got Lukas' demand that we can issue releases of 2.0.0 without it being stable. That's not possible this way. then we would need to demand that for all major version the numbering would start from 2.1.0 (in unstable state) and that 2.2.0 would be the first stable of the 2.x-branch. Ok, since Linux is one of the first (I guess it is the first) OS project, we should follow it's example. But since PEAR packages are not such a huge project than an operating system and Linus did not care in about any versioning standards in the first time, we should make that fit a bit more to our needs. Linux uses the first minor version number for indicating stability, what is not feasible for PEAR. But I guess we run much better in taking 0.x packages as unstable, 1.x as stable, 2.x as unstable and 3.x as stable. This is not exactly, what Linux does, but it's pretty clean and understandable.On the other hand, since there are some packages which have releases in 2.x marked stable this will bring up some inconsistancies, too. But these are much smaller than the current system we are recommending. -- Tobias Schlitt a passion for php http://www.schlitt.info