From mboxrd@z Thu Jan 1 00:00:00 1970 Path: news.gmane.io!.POSTED.blaine.gmane.org!not-for-mail From: Philip Kaludercic Newsgroups: gmane.emacs.help Subject: Re: Use package-vc.el for declarative, source-based package management? Date: Mon, 15 Apr 2024 19:45:49 +0000 Message-ID: <87wmoyl89e.fsf@posteo.net> References: <87plurnecb.fsf@posteo.de> Mime-Version: 1.0 Content-Type: text/plain Injection-Info: ciao.gmane.io; posting-host="blaine.gmane.org:116.202.254.214"; logging-data="35515"; mail-complaints-to="usenet@ciao.gmane.io" Cc: help-gnu-emacs@gnu.org To: Mekeor Melire Original-X-From: help-gnu-emacs-bounces+geh-help-gnu-emacs=m.gmane-mx.org@gnu.org Mon Apr 15 21:46:48 2024 Return-path: Envelope-to: geh-help-gnu-emacs@m.gmane-mx.org Original-Received: from lists.gnu.org ([209.51.188.17]) by ciao.gmane.io with esmtps (TLS1.2:ECDHE_RSA_AES_256_GCM_SHA384:256) (Exim 4.92) (envelope-from ) id 1rwSHw-00093v-4I for geh-help-gnu-emacs@m.gmane-mx.org; Mon, 15 Apr 2024 21:46:48 +0200 Original-Received: from localhost ([::1] helo=lists1p.gnu.org) by lists.gnu.org with esmtp (Exim 4.90_1) (envelope-from ) id 1rwSHN-0002wq-6M; Mon, 15 Apr 2024 15:46:13 -0400 Original-Received: from eggs.gnu.org ([2001:470:142:3::10]) by lists.gnu.org with esmtps (TLS1.2:ECDHE_RSA_AES_256_GCM_SHA384:256) (Exim 4.90_1) (envelope-from ) id 1rwSHE-0002vq-WA for help-gnu-emacs@gnu.org; Mon, 15 Apr 2024 15:46:05 -0400 Original-Received: from mout01.posteo.de ([185.67.36.65]) by eggs.gnu.org with esmtps (TLS1.2:ECDHE_RSA_AES_256_GCM_SHA384:256) (Exim 4.90_1) (envelope-from ) id 1rwSH7-0002qG-GO for help-gnu-emacs@gnu.org; Mon, 15 Apr 2024 15:45:59 -0400 Original-Received: from submission (posteo.de [185.67.36.169]) by mout01.posteo.de (Postfix) with ESMTPS id 3325A240027 for ; Mon, 15 Apr 2024 21:45:51 +0200 (CEST) DKIM-Signature: v=1; a=rsa-sha256; c=relaxed/simple; d=posteo.net; s=2017; t=1713210351; bh=jg/yzHP7x5Amqe2yC3J0lptW5FnoVUXRmJnsqulBNIE=; h=From:To:Cc:Subject:OpenPGP:Date:Message-ID:MIME-Version: Content-Type:From; b=CgNXnuxI31i7SabqbNObkyZW5AgVC3lOElQrNecjlsjLbw/qusIW8JT9vFWQINuJS ioHd9X0/xIAJk7AnD01LRFUyMnSdhKmmA6f80ZiTpCuKVval0izRuqLv6YuEj+Z1DA oQ0Yp7Z6t/mz3Xb5W/rzmR3QFV3shV4lE4xdR5DhwgpvtmJW8O0Ol1ZDdjX5+5zXMB KiSRT7SmjUekwBcuqlFSls5v2rfVgZHB1y+IkxpQfliFnuGj5dLyr616yvH1jYGNK1 Ay4nfppO+iSyFUkbNtXDBSMtx/uLqfJdakDgRGaDcOuebXeJdDNdwrLfJ7D18huJFG TNQL1FwDvXmFQ== Original-Received: from customer (localhost [127.0.0.1]) by submission (posteo.de) with ESMTPSA id 4VJHht4J8cz6twQ; Mon, 15 Apr 2024 21:45:50 +0200 (CEST) In-Reply-To: <87plurnecb.fsf@posteo.de> (Mekeor Melire's message of "Mon, 15 Apr 2024 09:51:32 +0000") OpenPGP: id=7126E1DE2F0CE35C770BED01F2C3CC513DB89F66; url="https://keys.openpgp.org/vks/v1/by-fingerprint/7126E1DE2F0CE35C770BED01F2C3CC513DB89F66"; preference=signencrypt Received-SPF: pass client-ip=185.67.36.65; envelope-from=philipk@posteo.net; helo=mout01.posteo.de X-Spam_score_int: -43 X-Spam_score: -4.4 X-Spam_bar: ---- X-Spam_report: (-4.4 / 5.0 requ) BAYES_00=-1.9, DKIM_SIGNED=0.1, DKIM_VALID=-0.1, DKIM_VALID_AU=-0.1, DKIM_VALID_EF=-0.1, RCVD_IN_DNSWL_MED=-2.3, RCVD_IN_MSPIKE_H3=0.001, RCVD_IN_MSPIKE_WL=0.001, SPF_HELO_NONE=0.001, SPF_PASS=-0.001 autolearn=ham autolearn_force=no X-Spam_action: no action X-BeenThere: help-gnu-emacs@gnu.org X-Mailman-Version: 2.1.29 Precedence: list List-Id: Users list for the GNU Emacs text editor List-Unsubscribe: , List-Archive: List-Post: List-Help: List-Subscribe: , Errors-To: help-gnu-emacs-bounces+geh-help-gnu-emacs=m.gmane-mx.org@gnu.org Original-Sender: help-gnu-emacs-bounces+geh-help-gnu-emacs=m.gmane-mx.org@gnu.org Xref: news.gmane.io gmane.emacs.help:146332 Archived-At: Mekeor Melire writes: > Hello :) > > I'm trying to use Emacs' built-in package-vc.el for an opinionated way > of declarative, source-based package management. I use Emacs from master > branch and I would like to use some Elpa-packages from their respective > development branch too. I also want local checkouts of their > repositories so that I can quickly contribute to them. And I do not want > to use Emacs' built-in use-package.el. Ideally, my init.el would begin > like this: > > ;; Here, I'd maintain a list of packages that I want to be installed: > (setopt package-vc-...-list-of-packages '(consult embark vertico)) > > ;; Then I'd instruct package-vc.el to ensure that these packages are > ;; installed, i.e. install missing packages with local repo > ;; checkouts: > (package-vc-...-ensure) > > > But so far, I was only able to approximate this ideal with several > caveats in my init.el, as follows. First, I load relevant modules. > > (require 'package) > (require 'package-vc) > > Here I maintain a list of packages that I want to be installed. > > (setopt package-vc-selected-packages > '( (consult > :url "https://github.com/minad/consult.git") > (embark > :url "https://github.com/oantolin/embark.git" > :ignored-files ("embark-consult.el" "avy-embark-collect.el")) > (vertico > :url "https://github.com/minad/vertico.git") > (vertico-buffer > :url "https://github.com/minad/vertico.git" > :lisp-dir "extensions" > :main-file "vertico-buffer.el"))) > > As you can see, one caveat is that I need to specify :url, :lisp-dir and > similar values manually. Can I gather these somehow automatically > instead? ELPA should already provide this information. > Another caveat is that some repositories, like Vertico's, contain > multiple packages, like vertico-buffer. Can I avoid having multiple > local checkouts of the same repository (while still keeping all > (sub)packages marked as installed)? Not really, by checking out repositories directly and adding their directories to `load-path' there are certain restrictions. Usually this is best fixed upstream. > Now, I'd like to ensure that these packages are installed. But if I put > (package-vc-install-selected-packages) in my init.el, Emacs > interactively asks me "Overwrite previous checkout for package > `consult'? (yes or no)" during startup. That's why I currently evaluate > that s-expression manually whenever I change the value of > package-vc-selected-packages. Is there a way to to make > package-vc-install-selected-packages (1.) not ask anything > interactively; and (2.) simply not overwrite already installed packages; > and (3.) still install packages that not yet installed? In the current state, `package-vc-install-selected-packages' should be marked as interactive-only, since this kind of usage just isn't reasonably supported, as you point out. We should perhaps add a NOCONFORM argument, like what `package-install-selected-packages' has. > After having ensured that packages are installed, the following > initializes load-path: > > (package-initialize) > > There's another problem though: When I put a package foo in > package-vc-selected-packages, and foo depends on another external > package bar, and I have not put bar in -selected-packages, and I call > package-vc-install-selected-packages, then Emacs will try to install to > install bar from a package-archive without any local repository checkout > but rather in the usual package.el way. That is because package-vc is never meant to be a replacement for package.el or the default way of installing things. One should emphasise, that from-source package management has the downside of increasing the complexity of upgrading and that one might always land on an unreliable commit that the author never meant to publish. > That's why I also have this > setting in my init.el: > > (setopt package-archives nil) > > This forces package.el to fail when I forgot to list a dependency in > -selected-packages variable. As an unfortunate consequence, I also end > up doing dependency management manually. But I'd love to have > dependencies vc-installed too! Thus this question arises: Is there a way > to tell package-vc-install-selected-packages to vc-install the selected > packages *including* their dependencies? That would require a modification to `package-vc-install-dependencies', so that it dosn't call `package-install-from-archive'. But once again, I would urge to rethink the approach. What is the reason you want to have everything installed using package-vc? Frequently, I find that there are some wrong assumptions behind the wish that I hope to clarify. -- Philip Kaludercic on peregrine