From mboxrd@z Thu Jan 1 00:00:00 1970 From: Kaushal Modi Subject: Re: bug#32722: bug#32722: bug#32722: 26.1; Org-publish depend on non-free platform ? Date: Fri, 21 Sep 2018 17:30:27 -0400 Message-ID: References: <531a7f93882cc0d5d09700457d06ea082f47569b.camel@flqt.fr> <83sh2ck42e.fsf@gnu.org> <87va73gn90.fsf@nicolasgoaziou.fr> <83efdqg32d.fsf@gnu.org> <87lg7yfraw.fsf@nicolasgoaziou.fr> <874leme9vp.fsf@aminb.org> <83h8ikenbc.fsf@gnu.org> <87in30geqo.fsf__12383.8069885553$1537471022$gmane$org@nicolasgoaziou.fr> <875zyzg3mw.fsf@alphapapa.net> <87zhwbehtn.fsf@alphapapa.net> Mime-Version: 1.0 Content-Type: text/plain; charset="UTF-8" Content-Transfer-Encoding: quoted-printable Return-path: Received: from eggs.gnu.org ([2001:4830:134:3::10]:59182) by lists.gnu.org with esmtp (Exim 4.71) (envelope-from ) id 1g3TEo-0005JV-KJ for emacs-orgmode@gnu.org; Fri, 21 Sep 2018 17:45:23 -0400 Received: from Debian-exim by eggs.gnu.org with spam-scanned (Exim 4.71) (envelope-from ) id 1g3T10-0003ql-PM for emacs-orgmode@gnu.org; Fri, 21 Sep 2018 17:31:07 -0400 Received: from mail-lj1-x235.google.com ([2a00:1450:4864:20::235]:42524) by eggs.gnu.org with esmtps (TLS1.0:RSA_AES_128_CBC_SHA1:16) (Exim 4.71) (envelope-from ) id 1g3T10-0003pm-GU for emacs-orgmode@gnu.org; Fri, 21 Sep 2018 17:31:06 -0400 Received: by mail-lj1-x235.google.com with SMTP id f1-v6so13201728ljc.9 for ; Fri, 21 Sep 2018 14:31:06 -0700 (PDT) In-Reply-To: <87zhwbehtn.fsf@alphapapa.net> List-Id: "General discussions about Org-mode." List-Unsubscribe: , List-Archive: List-Post: List-Help: List-Subscribe: , Errors-To: emacs-orgmode-bounces+geo-emacs-orgmode=m.gmane.org@gnu.org Sender: "Emacs-orgmode" To: Adam Porter Cc: emacs-org list On Thu, Sep 20, 2018 at 9:55 PM Adam Porter wrote: > 1. Why not? I just git-blamed this line in ox-html.el: > > (declare-function htmlize-region "ext:htmlize" (beg end)) > > It's from February, 2012. That's 6 and a half years, at least, that > that code has been present. Why are we suddenly concerned about it? I just so happens that someone recently filed this bug report: https://debbugs.gnu.org/cgi/bugreport.cgi?bug=3D32722. htmlize.el was living in the Org contrib directory until Aug 2017: =3D=3D=3D=3D=3D commit d0ced98943da0e5851ba1145515db27e939bff08 Author: Bastien Date: Fri Aug 18 09:54:19 2017 +0200 Delete htmlize.el from Org=E2=80=99s contrib directory * lisp/ox-org.el (org-org-publish-to-org): * lisp/ox-html.el (org-html-htmlize-generate-css): (org-html-fontify-code): * lisp/org-agenda.el (org-agenda-write): Throw an error requesting the user to install htmlize.el. * doc/org.texi (Exporting agenda views, Literal examples): Don=E2=80=99t assume htmlize.el is available. You need to install it from https://github.com/hniksic/emacs-htmlize See https://github.com/hniksic/emacs-htmlize/issues/7 for this issue. diff --git a/lisp/ox-html.el b/lisp/ox-html.el --- a/lisp/ox-html.el +++ b/lisp/ox-html.el @@ -1746,1 +1747,1 @@ - (require 'htmlize) + (error "Please install htmlize from https://github.com/hniksic/emacs-htmlize")) =3D=3D=3D=3D=3D That change lived in Org master branch (devel) for a while, then propagated to Org 9.1 and eventually to Emacs 26.1 release. So users switching from Emacs 25.3 to 26.1. would see as an abrupt change, when in fact it's just that htmlize.el got removed from contrib and instead started being correctly referenced to its source on GitHub. > 2. Is Org part of "Emacs core"? I didn't think that was how "Emacs > core" was defined, but I may be wrong. It is officially part of Emacs, > of course. So is there a distinction between "part of Emacs" and "Emacs > core"? Sorry, please don't quote me on that. By "Emacs core", I meant all packages that you can get from an Emacs installation without having to fetch one from internet. > Are there any other "time bombs" > in Org that we should be concerned about? Not that I know of. I think this is the only one. > a. That is not the originally stated problem. Yes, I know. The original problem was GitHub, and it snowballed to "htmlize cannot be fetched from outside Emacs". It's just so happens that someone noticed it and filed a bug report. In Emacs 25.3, if htmlize weren't installed, it would have thrown an error when (require 'htmlize) got executed (or not, if people were using org-plus-contrib which htmlize was part of, until Aug 2017). > 4. I'm certainly in favor of using built-in libraries as much as > possible. If htmlfontify is a better or equivalent solution, and > someone's willing to write the code and ensure there are no regressions, > that would be great, because it would save users from having to manually > install other packages to get expected functionality. +1 > 5. Having a passing familiarity with the complexity of the Org code > base, I am concerned about the potential for regressions in > functionality and usability. I'm also a bit disappointed to see this > burden potentially thrust upon Nicolas and other Org maintainers, to > replace working code that's existed for over 6 years, for little-to-no > technical reason. Their time available for working on Org is very > valuable. That's my concern too. And I think that it's mainly Nicolas, and in another thread, he mentioned that he will do this refactoring, but without an urgency (which makes sense).