# APC in 5.4 - Thread: https://externals.io/message/55323 - Original archive: https://news-web.php.net/php.internals/55323 - Started by Kalle Sommer Nielsen on 2011-09-09T01:04:27+00:00 - Messages: 10 - Reply by email to internals@lists.php.net - More for agents: search with https://externals.io/search.md?q={query} · JSON API at https://externals.io/api · documentation at https://externals.io/llms.txt --- ## #55323 — Kalle Sommer Nielsen — 2011-09-09T01:04:27+00:00 URL: https://externals.io/message/55323#55323 Hi This have been a while since we had the discussions about APC in 5.4 (or in general extensions to move in and out of the Core, but more about that in another thread). http://markmail.org/message/4w6lcbunw3qfof3c I bought this thread up a while back and it had a general approval feeling among the core guys and those who participate in internal discussions. So I was wondering if we finally agree on moving APC into the core, or is it too "late" for that? As Rasmus has spoken of, then the newly built-in webserver will greatly help improving the code coverage for APC, while also bringing some more attention to APC from a developer perspective (other than those who dwell into pecl/). -- regards, Kalle Sommer Nielsen kalle@php.net --- ## #55324 — Stas Malyshev — 2011-09-09T01:07:30+00:00 URL: https://externals.io/message/55323#55324 · In reply to: #55323 Hi! > This have been a while since we had the discussions about APC in 5.4 > (or in general extensions to move in and out of the Core, but more > about that in another thread). Not saying anything specifically about APC, I think it is way too late to propose major features for 5.4 when we're days before beta. So if it's 5.4, it's not 5.4.0 for sure. -- Stanislav Malyshev, Software Architect SugarCRM: http://www.sugarcrm.com/ (408)454-6900 ext. 227 --- ## #55340 — Kalle Sommer Nielsen — 2011-09-09T17:38:04+00:00 URL: https://externals.io/message/55323#55340 · In reply to: #55324 Hi 2011/9/9 Stas Malyshev : > Not saying anything specifically about APC, I think it is way too late to > propose major features for 5.4 when we're days before beta. So if it's 5.4, > it's not 5.4.0 for sure. Well all in all I won't mind adding it in a "patch" release, much like we added the FPM SAPI in 5.3. Like it was raised in the original thread, bringing more awareness about an opcode cacher into a standard distribution of PHP will benefit us all in the long run. -- regards, Kalle Sommer Nielsen kalle@php.net --- ## #55383 — Ilia Alshanetsky — 2011-09-12T12:09:08+00:00 URL: https://externals.io/message/55323#55383 · In reply to: #55324 The agreement to include apc in 5.4 is an old one, unfortunately the action of doing was just missed. Also, inclusion of the extension won't break any code since it is self contained... > Hi! > > > > […] > > Not saying anything specifically about APC, I think it is way too late to > propose major features for 5.4 when we're days before beta. So if it's 5.4, > it's not 5.4.0 for sure. > > -- > Stanislav Malyshev, Software Architect > SugarCRM: http://www.sugarcrm.com/ > (408)454-6900 ext. 227 --- ## #55389 — Stas Malyshev — 2011-09-12T17:42:09+00:00 URL: https://externals.io/message/55323#55389 · In reply to: #55383 Hi! > The agreement to include apc in 5.4 is an old one, unfortunately the > action of doing was just missed. Also, inclusion of the extension > won't break any code since it is self contained... If it's an "old one", how comes nobody ever mentioned it during all the extensive process of naming features, voting, etc. and didn't bother to propose it? I don't feel good about first agreeing to a set of rules and then immediately starting to throw them away "just this time, just for this occasion". Why have rules and release cycles then? -- Stanislav Malyshev, Software Architect SugarCRM: http://www.sugarcrm.com/ (408)454-6900 ext. 227 --- ## #55390 — Pierre Joye — 2011-09-12T17:59:15+00:00 URL: https://externals.io/message/55323#55390 · In reply to: #55389 > Hi! > > > > […] > > If it's an "old one", how comes nobody ever mentioned it during all the > extensive process of naming features, voting, etc. and didn't bother to > propose it? There was a discussion and the conclusion was that APC was not ready for inclusion. See my other reply. > I don't feel good about first agreeing to a set of rules and then > immediately starting to throw them away "just this time, just for this > occasion". Why have rules and release cycles then? -- Pierre @pierrejoye | http://blog.thepimp.net | http://www.libgd.org --- ## #55397 — Derick Rethans — 2011-09-13T13:35:22+00:00 URL: https://externals.io/message/55323#55397 · In reply to: #55383 > The agreement to include apc in 5.4 is an old one, unfortunately the > action of doing was just missed. Also, inclusion of the extension > won't break any code since it is self contained... I also thought that APC was going to part of PHP 5.4. And I agree with Ilia that it doesn't matter to the overal stability if we bundle it or not. It is self contained code, and having it part of the distribution means more people play with it and use it, and problems are found earlier. Release early, release often. Having to delay APC, or basically any other self-contained bit of code (extensions and new functions/classes/methods) for at least a year, and perhaps sometimes up to even 2 years, makes very little sense to me. The more ticker tape we get, the slower we innovate. Of course, it's likely that things will have bugs, but if we mark it as "exprimental" in the docs than I see no problems. It makes very little sense to not allow any "new" code for a whole new year to filter into PHP. It makes developing new exiciting things for PHP a pain as it simply sucks to have to wait up to a year and a half for new shiny things to appear in PHP. Derick -- http://derickrethans.nl | http://xdebug.org Like Xdebug? Consider a donation: http://xdebug.org/donate.php twitter: @derickr and @xdebug --- ## #55401 — johannes@schlueters.de — 2011-09-13T14:52:07+00:00 URL: https://externals.io/message/55323#55401 · In reply to: #55323 Hi, > Hi > > This have been a while since we had the discussions about APC in 5.4 > (or in general extensions to move in and out of the Core, but more > about that in another thread). > > http://markmail.org/message/4w6lcbunw3qfof3c I think the conclusion was that we should include it (that's btw. a thought since the beginning of PHP 6) but anytime it was discussed it was premature. In general I'd tend to drop stuff from PHP instead of adding it. We ship quite a few extensions where I claim that we don't do any testing and little maintenance. I would reduce the distribution size and include things we consider "core PHP" and enable these by default in the hope that gives developers a clear guidance and what they should be able to expect. (hoping we can convince distributors to have the same things in their "PHP" meta packages) But that's nothing for 5.4 but for 5.5. Anybody wants to write an RFC by chance? :-) johannes --- ## #55402 — Nathaniel Catchpole — 2011-09-13T15:06:50+00:00 URL: https://externals.io/message/55323#55402 · In reply to: #55397 I didn't closely follow the earlier discussions about including APC, but I (and at least some other people working a lot on Drupal) had been mostly assuming it'd be included in 5.4 based on those. The assumption that APC is available by default (if not necessarily enabled) would allow us to start making more assumptions in Drupal as well over time, so while I have no stake in the decision, it would not at all be a surprise, and it'd be very very welcome. If it came in 5.4.1 or something then that'd make no odds though. Nat > > > > […] > > I also thought that APC was going to part of PHP 5.4. And I agree with > Ilia that it doesn't matter to the overal stability if we bundle it or > not. It is self contained code, and having it part of the distribution > means more people play with it and use it, and problems are found > earlier. Release early, release often. > > Having to delay APC, or basically any other self-contained bit of code > (extensions and new functions/classes/methods) for at least a year, and > perhaps sometimes up to even 2 years, makes very little sense to me. The > more ticker tape we get, the slower we innovate. > > Of course, it's likely that things will have bugs, but if we mark it as > "exprimental" in the docs than I see no problems. It makes very little > sense to not allow any "new" code for a whole new year to filter into > PHP. It makes developing new exiciting things for PHP a pain as it > simply sucks to have to wait up to a year and a half for new shiny > things to appear in PHP. > > Derick > -- > http://derickrethans.nl | http://xdebug.org > Like Xdebug? Consider a donation: http://xdebug.org/donate.php > twitter: @derickr and @xdebug --- ## #55403 — Lester Caine — 2011-09-13T15:29:52+00:00 URL: https://externals.io/message/55323#55403 · In reply to: #55401 Johannes Schlüter wrote: > In general I'd tend to drop stuff from PHP instead of adding it. We ship > quite a few extensions where I claim that we don't do any testing and > little maintenance. I would reduce the distribution size and include > things we consider "core PHP" and enable these by default in the hope > that gives developers a clear guidance and what they should be able to > expect. (hoping we can convince distributors to have the same things in > their "PHP" meta packages) But that's nothing for 5.4 but for 5.5. > Anybody wants to write an RFC by chance?:-) Since most of the linux distributors already strip the extra packages and ship them self contained in their own packages it IS about time that the non-essential extensions were kept separate from the core, so a move to a more modular management system is overdue? We can then simply select what we want to include with the core rather than the current method of simply ignoring the unused extensions? APC is optional ... like many other extensions ... so only needs to be supplied as an option anyway. -- Lester Caine - G8HFL ----------------------------- Contact - http://lsces.co.uk/wiki/?page=contact L.S.Caine Electronic Services - http://lsces.co.uk EnquirySolve - http://enquirysolve.com/ Model Engineers Digital Workshop - http://medw.co.uk// Firebird - http://www.firebirdsql.org/index.php