Newsgroups: php.internals Path: news.php.net Xref: news.php.net php.internals:101405 Return-Path: Mailing-List: contact internals-help@lists.php.net; run by ezmlm Delivered-To: mailing list internals@lists.php.net Received: (qmail 67577 invoked from network); 21 Dec 2017 20:19:10 -0000 Received: from unknown (HELO lists.php.net) (127.0.0.1) by localhost with SMTP; 21 Dec 2017 20:19:10 -0000 Authentication-Results: pb1.pair.com header.from=johannes@schlueters.de; sender-id=unknown Authentication-Results: pb1.pair.com smtp.mail=johannes@schlueters.de; spf=permerror; sender-id=unknown Received-SPF: error (pb1.pair.com: domain schlueters.de from 84.19.169.162 cause and error) X-PHP-List-Original-Sender: johannes@schlueters.de X-Host-Fingerprint: 84.19.169.162 mail.experimentalworks.net Received: from [84.19.169.162] ([84.19.169.162:54794] helo=mail.experimentalworks.net) by pb1.pair.com (ecelerity 2.1.1.9-wez r(12769M)) with ESMTP id CF/9F-10479-B371C3A5 for ; Thu, 21 Dec 2017 15:19:09 -0500 Received: from kuechenschabe.fritz.box (ppp-46-244-175-146.dynamic.mnet-online.de [46.244.175.146]) by mail.experimentalworks.net (Postfix) with ESMTPSA id D716B6AB08; Thu, 21 Dec 2017 21:19:04 +0100 (CET) Message-ID: <1513887544.16381.12.camel@schlueters.de> To: Anatol Belski , "Christoph M. Becker" , Pedro =?ISO-8859-1?Q?Magalh=E3es?= Cc: PHP internals list Date: Thu, 21 Dec 2017 21:19:04 +0100 In-Reply-To: References: <1512478149.2069.40.camel@schlueters.de> <19331e5c-d82a-2d60-4f26-755da11fc576@gmx.de> <1512483692.2069.58.camel@schlueters.de> Content-Type: text/plain; charset="UTF-8" X-Mailer: Evolution 3.18.5.2-0ubuntu3.2 Mime-Version: 1.0 Content-Transfer-Encoding: 8bit Subject: Re: [PHP-DEV] Outstanding php.net account requests From: johannes@schlueters.de (Johannes =?ISO-8859-1?Q?Schl=FCter?=) On Di, 2017-12-19 at 21:17 +0000, Anatol Belski wrote: > IMO it is a strategical question and we should keep up with the real > world. It was one of the reasons for Pickle to come to life, which > however didn't come to the optimal end implementation meanwhile. > There's github, bitbucket, etc. As a usage example - an extension > could be accepted on PECL, but indeed be released on GitHub only. The > PECL site could sync that release to PECL, the pecl tool could be > aware there were some release outside at the places we'd define as > safe. Similar thoughts would be if it came to the Composer > integration - the more it'd ease access to the extensions, the more > profit it would be in the end. Perhaps it's just me, but reading on > an arbitrary PHP project page something like "no extensions required" > sounds really amiss.  The issue with composer integration is, that composer is build around a "project" whereas extensions are installed for a runtime and those not necessarily overlap. i.e. it is possible to run "composer install foo" via CLI from one PHP runtime and then execute the installed code using a php-fpm with different ini locations. If we build a pecl installer replacement we should see how we can integrate with composer on one side and the php7enmod stuff from Ubuntu on the other side. (and maybe also Windows ;-) ) > Basically, I'd suggest we don't pretend there's no world outta PECL > and make things easy for use case scenarios where fe people only work > on GitHub. For instance, we could also accept authentication with > GitHub and others on PECL, with all the logical consequences. There's > already quite some work with the regard to PR handling on GitHub for > qa.php.net. Of course, it wouldn't deny the classic PECL use case > scenario. The underlying questions have to be looked at. PECL once aimed to be (among other things) an incubator for PHP extensions, which would later be bundled. This has higher requirements (on license, copyright etc.) also the old model (back when CVS/SVN were the source of truth) allowed us to take over maintenance of "important" extensions once the author went away. If we're just a frontend to external repos we loose that to some degree (while we could change the repository pointer) This needs some thinking. I'd really like to see more opinions on this. johannes