Return-Path: X-Spam-Checker-Version: SpamAssassin 3.4.0 (2014-02-07) on aws-us-west-2-korg-lkml-1.web.codeaurora.org Received: from vger.kernel.org (vger.kernel.org [23.128.96.18]) by smtp.lore.kernel.org (Postfix) with ESMTP id E4424C7618A for ; Thu, 16 Mar 2023 18:28:02 +0000 (UTC) Received: (majordomo@vger.kernel.org) by vger.kernel.org via listexpand id S230146AbjCPS2B (ORCPT ); Thu, 16 Mar 2023 14:28:01 -0400 Received: from lindbergh.monkeyblade.net ([23.128.96.19]:57606 "EHLO lindbergh.monkeyblade.net" rhost-flags-OK-OK-OK-OK) by vger.kernel.org with ESMTP id S229589AbjCPS17 (ORCPT ); Thu, 16 Mar 2023 14:27:59 -0400 Received: from ms.lwn.net (ms.lwn.net [IPv6:2600:3c01:e000:3a1::42]) by lindbergh.monkeyblade.net (Postfix) with ESMTPS id 4E5BE4B829; Thu, 16 Mar 2023 11:27:57 -0700 (PDT) Received: from localhost (unknown [IPv6:2601:281:8300:73::5f6]) (using TLSv1.2 with cipher ECDHE-RSA-AES256-GCM-SHA384 (256/256 bits)) (No client certificate requested) by ms.lwn.net (Postfix) with ESMTPSA id D14FA823; Thu, 16 Mar 2023 18:27:56 +0000 (UTC) DKIM-Filter: OpenDKIM Filter v2.11.0 ms.lwn.net D14FA823 DKIM-Signature: v=1; a=rsa-sha256; c=relaxed/relaxed; d=lwn.net; s=20201203; t=1678991277; bh=OwZBhz9VMbfwLaD6vDW+gcJbWhV8x+C3/Q50X51+JhM=; h=From:To:Cc:Subject:In-Reply-To:References:Date:From; b=QFstPU4oaTDsTc2nbT7WD/XIDOTQ92knb79vNLSDyNlTuLSVTuHbnrG33v4DZG6/z ScKv9Fw6tZhUWrmKPBuKeNC4Wpwpk0A5ELxE8awvHZjIzaiKu424sgXKxjQxVAv8HA wuazJA60F7mewuumNZaL1xo0aBfcRbqanhnSCYU95qNms0+0XCx5BFul00AazkRzSy LfTfAzHdZ3p3BFXcWxBJhKtd37+ajbeRa3UzS+OZ2lTaMWc4FG60rYHnTgdQc+xGr6 An7PAhN7AOZ0di42WKOrdrk20ZDtt9KB3z5jNPPKOmKpbKDg50D6KwDbs1bh63do1Q G0lZ0aIk1TO+w== From: Jonathan Corbet To: Thorsten Leemhuis Cc: Randy Dunlap , Lukas Bulwahn , linux-doc@vger.kernel.org, linux-kernel@vger.kernel.org, regressions@lists.linux.dev Subject: Re: [PATCH v3] docs: describe how to quickly build a trimmed kernel In-Reply-To: References: <1a788a8e7ba8a2063df08668f565efa832016032.1678021408.git.linux@leemhuis.info> <87a60frxk0.fsf@meer.lwn.net> Date: Thu, 16 Mar 2023 12:27:56 -0600 Message-ID: <87edpomtzn.fsf@meer.lwn.net> MIME-Version: 1.0 Content-Type: text/plain Precedence: bulk List-ID: X-Mailing-List: linux-kernel@vger.kernel.org Thorsten Leemhuis writes: > Documentation/doc-guide/contributing.rst says that "books" are meant to > "group documentation for specific readers"; creating a new book for > tutorials would work against that, as readers (users and administrators > in this case) then would have to consult two books. The idea behind that guideline is that readers should be able to know where to look and to not have to dig through a lot of material that was not intended for them. Not that, for any given reader, there should be exactly one book that has everything they might want. One could also argue, of course, that readers seeking tutorials are a different group than those seeking reference material. > And isn't for example Documentation/process/submitting-patches.rst also > more of a tutorial than reference material (which we also have in the > form of Documentation/process/development-process.rst)? It's a pretty clear example of what happens when you try to combine both types of documentation - you get something that isn't ideal for either type of reader. It tries to take people through the process, but it is also the only reference document we have on how patches should be submitted. > Does that mean > it should be moved? Into the same book or a separate book, as it has a > different target audience? I fear that might quickly get confusing for > readers without any real benefits No, I wouldn't move it. We could, someday, consider splitting it into two more focused documents, one of which could (say) go under tutorials/. > Or did I understand the idea of a new book wrong and you meant something > else? Like creating Documentation/admin-guide/tutorials/ and putting the > text there? That might work and would help future authors to get the > right mental model when writing new texts. But I'm not sure that's worth it. I wasn't thinking of doing it that way, but we could certainly consider it. It doesn't seem like we would have vast numbers of these, though, and they would mostly cover relatively elementary topics, so a single, top-level directory might be better if we decide to take this path. Thanks, jon