Re: 2.0.0 (Re: [PEAR-DEV] PEAR 1.9.5 testing)
| From: | Alexey Borzov | Date: | Tue, 15 Jul 2014 16:28:54 +0000 |
| Subject: | Re: 2.0.0 (Re: [PEAR-DEV] PEAR 1.9.5 testing) | ||
| References: | 1 2 3 4 | Groups: | php.pear.dev |
| Request: | Send a blank email to pear-dev+get-55382@lists.php.net to get a copy of this message | ||
Hi Till,
On 15.07.2014 20:11, till wrote:
"Not keeping BC / going 2.0.0 / whatever" is only needed if we rewrite from scratch.I agree with Christian. My thoughts were not to throw it all away and write the installer from scratch.I don't see a future for Pyrus. Nobody is working on it, and it is not on-par with the features PEAR currently has - as well as a plethora of open bugs I opened years ago. See https://github.com/pyrus/Pyrus/issueswas wondering what you thought about going 2.0.0-alpha1 and dropping PHP4 support once and for all.Where does that leave Pyrus?
An alpha1 could have bugfixes and so on from the roadmap and maybe one or two dependencies swapped out/updated. E.g. I'd rather use HTTP_Request2 in there than to rewrite another HTTP layer, etc.. I'd swap out pecl-related code for pickle[1] etc. pp.. Why — because it's code that's maintained.Yes, I'm going to change PEAR to use HTTP_Request2. This does not require version 2.0.0 and will not break BC.
Just granularly tackle current issues and modernize the code base. Not a complete rewrite. (cc Alexey)What prevents doing it with version 1.10?
A 2.0 would allow us BC breaks, e.g. to drop PHP4 support, which might even make it fun to work on it.We don't need a BC break to drop PHP4 support, this was discussed quite a bit, I think. I'd suggest going with Pyrus if we want a brand new version, its problems are not code-related but [lack-of-]doc-related and people-related mostly.