From mboxrd@z Thu Jan 1 00:00:00 1970 From: Bastien Subject: Re: org-plus-contrib: Invalid function: org-no-popups Date: Sat, 26 Jan 2013 11:45:24 +0100 Message-ID: <871ud88bp7.fsf@bzg.ath.cx> References: <87ham6wbwx.fsf@Rainer.invalid> <87622mkzti.fsf@bzg.ath.cx> <87pq0tzrk0.fsf@Rainer.invalid> <871ud9ir2e.fsf@bzg.ath.cx> <874ni5owxt.fsf@Rainer.invalid> Mime-Version: 1.0 Content-Type: text/plain; charset=utf-8 Content-Transfer-Encoding: 8bit Return-path: Received: from eggs.gnu.org ([208.118.235.92]:60216) by lists.gnu.org with esmtp (Exim 4.71) (envelope-from ) id 1Tz3GJ-0002Tn-68 for emacs-orgmode@gnu.org; Sat, 26 Jan 2013 05:45:28 -0500 Received: from Debian-exim by eggs.gnu.org with spam-scanned (Exim 4.71) (envelope-from ) id 1Tz3GG-0003Ps-Ph for emacs-orgmode@gnu.org; Sat, 26 Jan 2013 05:45:27 -0500 Received: from mail-wi0-f173.google.com ([209.85.212.173]:58460) by eggs.gnu.org with esmtp (Exim 4.71) (envelope-from ) id 1Tz3GG-0003Pg-Ir for emacs-orgmode@gnu.org; Sat, 26 Jan 2013 05:45:24 -0500 Received: by mail-wi0-f173.google.com with SMTP id hn17so244471wib.6 for ; Sat, 26 Jan 2013 02:45:23 -0800 (PST) In-Reply-To: <874ni5owxt.fsf@Rainer.invalid> (Achim Gratz's message of "Fri, 25 Jan 2013 20:59:58 +0100") 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-bounces+geo-emacs-orgmode=m.gmane.org@gnu.org To: Achim Gratz Cc: emacs-orgmode@gnu.org Hi Achim, Achim Gratz writes: > My reading comprehension may be at an all-time low, but this is what you > wrote in another thread: > >>>> This [removing Org from ELPA -ed] is using Org users as hostages, I >>>> don't want to do this. > >> I think this is an option. > > Hm… I'm obviously missing some piece. Yes :) This is were misunderstanding started: ,---- | > I've to admit that I've had that thought too, but it seems that doing | > this would perhaps even lower the attention that Emacs' maintainers give | > to that problem. | | This is using Org users as hostages, I don't want to do this. `---- What I meant is this: "Using the fact that Org/ELPA users suffer from a bug in package.el in hope that this will make Emacs maintainers fix the package.el bug is using Org/ELPA users as hostages." (Obviously my "this" where not equivalent to your "that" in the passage quoted above.) >> How can you make sure that Org-mode has not been used before people >> want to install an updated version through GNU/Org ELPA? > > That's what the patch does that you didn't like. I don't like it for at least these reasons: - it introduces an advice in the code -- we should work toward removing them, not adding them. - it advices `require', which is a very core function in Emacs. This is completely hackish. >> I can't think of any good way. >> >> Until someone have an idea about this, I'm very serious about removing >> the possibility of installing Org through ELPA. But as I said I won't >> do this before getting a sense of what the users use and expect. > > There are exactly two possibilities to get into this situation: > > The first is using Org anything before opening the package manager. I > think it should be possible to educate users that it's better t use a > fresh Emacs instance for updating or installing packages until this has > been fixed in package manager. This is like telling Windows users to reboot their system so that it can works correctly (remember the old days?). It feels wrong. And I doubt we can educate the users -- there are many of them, and many don't have the time to tell they have a problem. > The second are things like starter-kit. I'm thinking that it should be > the responsibility of such software to load and unload Org if they > really need to use it before activating packages and ask for support > from Emacs devs in how to do this properly. Package manager is a first > class citizen of Emacs by now which means any add-on has to cope with > its existence. It's interesting to see that the bugs seem to get > chalked up to the packages that are in ELPA rather than the mechanisms > that install them (or prevent proper initialization of Emacs) and it > kind of seems to fly under the radar for this reason even though it > seems to affect many users and has had for quite some time already. I'm not sure what solution you suggest here. In general, I think the benefit of having several ELPA distributions of Org is not worth the trouble of dealing with installation issues. -- Bastien