From mboxrd@z Thu Jan 1 00:00:00 1970 Path: news.gmane.org!.POSTED!not-for-mail From: Eli Zaretskii Newsgroups: gmane.emacs.devel Subject: Re: Preview: portable dumper Date: Fri, 02 Dec 2016 09:54:57 +0200 Message-ID: <83inr2oje6.fsf@gnu.org> References: <047a67ec-9e29-7e4e-0fb0-24c3e59b5886@dancol.org> <83zikjxt1j.fsf@gnu.org> <8360n6ruzu.fsf@gnu.org> <834m2nplmb.fsf@gnu.org> Reply-To: Eli Zaretskii NNTP-Posting-Host: blaine.gmane.org X-Trace: blaine.gmane.org 1480665332 28469 195.159.176.226 (2 Dec 2016 07:55:32 GMT) X-Complaints-To: usenet@blaine.gmane.org NNTP-Posting-Date: Fri, 2 Dec 2016 07:55:32 +0000 (UTC) Cc: emacs-devel@gnu.org To: Stefan Monnier Original-X-From: emacs-devel-bounces+ged-emacs-devel=m.gmane.org@gnu.org Fri Dec 02 08:55:28 2016 Return-path: Envelope-to: ged-emacs-devel@m.gmane.org Original-Received: from lists.gnu.org ([208.118.235.17]) by blaine.gmane.org with esmtp (Exim 4.84_2) (envelope-from ) id 1cCigp-0005z6-98 for ged-emacs-devel@m.gmane.org; Fri, 02 Dec 2016 08:55:27 +0100 Original-Received: from localhost ([::1]:33052 helo=lists.gnu.org) by lists.gnu.org with esmtp (Exim 4.71) (envelope-from ) id 1cCigq-00020Z-6t for ged-emacs-devel@m.gmane.org; Fri, 02 Dec 2016 02:55:28 -0500 Original-Received: from eggs.gnu.org ([2001:4830:134:3::10]:42294) by lists.gnu.org with esmtp (Exim 4.71) (envelope-from ) id 1cCigG-00020H-K7 for emacs-devel@gnu.org; Fri, 02 Dec 2016 02:54:53 -0500 Original-Received: from Debian-exim by eggs.gnu.org with spam-scanned (Exim 4.71) (envelope-from ) id 1cCigD-0007Q1-GM for emacs-devel@gnu.org; Fri, 02 Dec 2016 02:54:52 -0500 Original-Received: from fencepost.gnu.org ([2001:4830:134:3::e]:36683) by eggs.gnu.org with esmtp (Exim 4.71) (envelope-from ) id 1cCigD-0007Px-DK; Fri, 02 Dec 2016 02:54:49 -0500 Original-Received: from 84.94.185.246.cable.012.net.il ([84.94.185.246]:1193 helo=home-c4e4a596f7) by fencepost.gnu.org with esmtpsa (TLS1.2:RSA_AES_256_CBC_SHA1:256) (Exim 4.82) (envelope-from ) id 1cCigC-0007Jg-Gl; Fri, 02 Dec 2016 02:54:49 -0500 In-reply-to: (message from Stefan Monnier on Thu, 01 Dec 2016 21:18:04 -0500) X-detected-operating-system: by eggs.gnu.org: GNU/Linux 2.2.x-3.x [generic] X-Received-From: 2001:4830:134:3::e X-BeenThere: emacs-devel@gnu.org X-Mailman-Version: 2.1.21 Precedence: list List-Id: "Emacs development discussions." List-Unsubscribe: , List-Archive: List-Post: List-Help: List-Subscribe: , Errors-To: emacs-devel-bounces+ged-emacs-devel=m.gmane.org@gnu.org Original-Sender: "Emacs-devel" Xref: news.gmane.org gmane.emacs.devel:209904 Archived-At: > From: Stefan Monnier > Date: Thu, 01 Dec 2016 21:18:04 -0500 > > > It's a new and significant mechanism, which is central to building and > > starting Emacs. It needs to be studied, understood, and maintained. > > FWIW, I looked at Daniel's code and found it understandable, whereas > I never managed to make any sense of the unexec code, which still just looks > like magical voodoo to me, despite my trying twice to understand it. I understand both (though, admittedly, that excludes unexelf.c, which I never needed to look at for more than 5 sec). But this is not about you or me. Do you really believe that I'm afraid of the dumper code, having tweaked the display engine to speak RTL? However, this is about the next generation of Emacs maintainers who will inherit this code and will have to maintain it. Granted, the proposed dumper is not very complicated. But it isn't trivial either. So if we can achieve a similar effect by using the "normal" loadup code, which is much simpler and doesn't really require understanding anything new, I think it's more beneficial for the project's future. The key issue is performance, as was stated many times. If the performance of loadup after all the speedups is not good enough, then we will have to conclude that this idea cannot fly, at least not yet. (That conclusion might be revisited later, if/when disks, filesystems, and machines become faster.) It will also not fly if no one will pick up the gauntlet and work on that idea to get it to completion. But IMO we should at least try implementing that idea, because its benefits are so clear. We shouldn't give up without trying, and we have no special reason to decide the idea isn't viable until we've tried. Once again, if those ideas seem strange, let alone incorrect, to you-all out there, just say the word, and I will step down. Then I can stop worrying about the portable dumper, and you can stop worrying about my strange ideas. Emacs is not my private project; I'm only entitled to promote my ideas if the others either support them or trust me to DTRT. Please decide which one is it, and let's move on.