X-Original-To: alpine-devel@lists.alpinelinux.org Received: from outgoing.fripost.org (giraff.fripost.org [193.234.15.44]) by lists.alpinelinux.org (Postfix) with ESMTP id 19474F83256 for ; Tue, 19 Mar 2019 13:08:51 +0000 (UTC) Received: from localhost (localhost [127.0.0.1]) by outgoing.fripost.org (Postfix) with ESMTP id 56D671577F75 for ; Tue, 19 Mar 2019 14:08:50 +0100 (CET) DKIM-Signature: v=1; a=rsa-sha256; c=relaxed/simple; d=x.fripost.org; h= in-reply-to:content-disposition:content-type:content-type :mime-version:references:message-id:subject:subject:from:from :date:date; s=9df9cdc7e101629b5003b587945afa70; t=1553000930; x= 1554815331; bh=LV5pG61Dm1L9em5MCur0/Invf4Qexa4TSQmPac18u94=; b=M 7Vm117taW9r12FLDFLVPkh9zGJNzLXuQtODkbWryHTAhUJZicxunuJthHaclyD1K OwJBPKYg3Xxiqsh9du08v+TP2qLEoAfeiBtN0QkVemLJVsthrA4NGyrA2UIGp6nY l2xmz10bee/fTJZwfB2SkKLNxZxPZIKSyWU/4lqr6NzbaZLeKvNMvDQjxzeVhNrf A6vy3ujUQTjrHbTJr8IqrpkZFrWT6YU217Yf7QRLYIDJ9kJCTFZFIQ63cWtvDv8U 0bue/GLxaaFrThNTTzebIS3q/biblXoxqWW2XSvHcLIk5fPiuFgJQlVb05516UnK RXPiJkgj4wvKm+3TR7i7A== X-Virus-Scanned: Debian amavisd-new at fripost.org Received: from outgoing.fripost.org ([127.0.0.1]) by localhost (giraff.fripost.org [127.0.0.1]) (amavisd-new, port 10040) with LMTP id LtBI3kDIPRNS for ; Tue, 19 Mar 2019 14:08:50 +0100 (CET) Received: from smtp.fripost.org (unknown [172.16.0.6]) by outgoing.fripost.org (Postfix) with ESMTP id 3B3C31577F72 for ; Tue, 19 Mar 2019 14:08:50 +0100 (CET) Received: from [127.0.0.1] (localhost [127.0.0.1]) (using TLSv1.2 with cipher ECDHE-RSA-AES256-GCM-SHA384 (256/256 bits)) by smtp.fripost.org (Postfix) with ESMTPSA id B52E34D4A42F for ; Tue, 19 Mar 2019 14:08:47 +0100 (CET) Received: (qmail 8101 invoked from network); 19 Mar 2019 13:08:19 -0000 Received: from localhost (HELO aetey.se) (eh1ba719@127.0.0.1) by mail with ESMTPA; 19 Mar 2019 13:08:19 -0000 Date: Tue, 19 Mar 2019 14:08:24 +0100 From: u-3qpt@aetey.se To: Natanael Copa Cc: Drew DeVault , alpine-devel@lists.alpinelinux.org Subject: Re: [alpine-devel] Normalizing Python packages in aports Message-ID: <20190319130824.GN23721@example.net> References: <20190306212501.GD2800@cirno.localdomain> <20190314112413.29dd9816@ncopa-desktop.copa.dup.pw> <20190314191241.GF1544@homura.localdomain> <20190319103153.3964ded5@ncopa-desktop.copa.dup.pw> <20190319113230.GM23721@example.net> <20190319125738.73f31222@ncopa-desktop.copa.dup.pw> X-Mailinglist: alpine-devel Precedence: list List-Id: Alpine Development List-Unsubscribe: List-Post: List-Help: List-Subscribe: MIME-Version: 1.0 Content-Type: text/plain; charset=us-ascii Content-Disposition: inline In-Reply-To: <20190319125738.73f31222@ncopa-desktop.copa.dup.pw> On Tue, Mar 19, 2019 at 12:57:38PM +0100, Natanael Copa wrote: > On Tue, 19 Mar 2019 12:32:30 +0100 > u-3qpt@aetey.se wrote: > > From this point of view, it is important to (also) name all python3 > > packages as py3-*. > > always name all python3 packages py3-* would also make it possible to > install py4-* and py3-* in parallel if for some reason that would be > needed in future. It is always useful to be able to do parallell installations, this avoids artificial collisions when different packages need different packages which want to occupy the same naming slot. > python3.6 and python3.7 will not be possible to install in parallel. Why not? As long as "python3" is the only one which is named "python3" and, say, depends on a certain one of 3.6 or 3.7. Then the semantics of "python3" and the corresponding contract against other packages is clear: "only the features which are present in 3.x for all x<=N". Then the consumer can either accept this contract or explicitly refer to, say "python3.7" if it needs a feature not present in the "pyhton3" contract. Of course there will be still possibilities to choose wrong, but at least the semantics can be specified unambiguously and there will be no collisions. Redundancy may appear if different consumers pull in both 3.6 and 3.7. But as well this can mean that really both 3.6 and 3.7 are needed on the computer in question. No breakage in either case. > > Please do not perpetuate the others' badly chosen practices - they can > > not be made consistent other than only temporary and only with a large > > effort each time, if at all. > > what about `python-dev`? should you get latest python or failure? If you ask me - this is clearly underspecified and _must_ fail. How do I as a packager know which of the "python-family" languages the local sysadmin will want his users to develop in? "Latest" has no stable mapping, how can a consumer (another package or a local administrator) rely on it without checking what "latest" means in the context of "python-dev"? Somewhat later this may no longer be the same? No, thanks. Of course nothing prevents having a package "pythonLatest-dev" with a description "this will correspond to yet unknown versions of "Python" depending on the pace of the development of the language and on how fast we will update the corresponding packages, updates will change what you get via this package, possibly to something incompatible". This might be exactly what someone needs for a day, but probably not what someone else needs to specify dependencies in his python-related package. Regards, Rune --- Unsubscribe: alpine-devel+unsubscribe@lists.alpinelinux.org Help: alpine-devel+help@lists.alpinelinux.org ---