From mboxrd@z Thu Jan 1 00:00:00 1970 From: Erik Faye-Lund Subject: Re: [PATCH] push: Use sideband channel for hook messages Date: Fri, 5 Feb 2010 17:14:12 +0100 Message-ID: <40aa078e1002050814m1edb90f8lf3d8d138fa65c332@mail.gmail.com> References: <20100205033748.GA19255@spearce.org> <4B6C07E3.5030705@viscovery.net> <20100205153252.GC19255@spearce.org> <4B6C3E9C.7040009@viscovery.net> Reply-To: kusmabite@gmail.com Mime-Version: 1.0 Content-Type: text/plain; charset=ISO-8859-1 Content-Transfer-Encoding: QUOTED-PRINTABLE Cc: "Shawn O. Pearce" , Junio C Hamano , git To: Johannes Sixt X-From: git-owner@vger.kernel.org Fri Feb 05 17:22:09 2010 Return-path: Envelope-to: gcvg-git-2@lo.gmane.org Received: from vger.kernel.org ([209.132.180.67]) by lo.gmane.org with esmtp (Exim 4.69) (envelope-from ) id 1NdQwb-00020L-GU for gcvg-git-2@lo.gmane.org; Fri, 05 Feb 2010 17:22:09 +0100 Received: (majordomo@vger.kernel.org) by vger.kernel.org via listexpand id S933440Ab0BEQWE convert rfc822-to-quoted-printable (ORCPT ); Fri, 5 Feb 2010 11:22:04 -0500 Received: from mail-fx0-f211.google.com ([209.85.220.211]:43938 "EHLO mail-fx0-f211.google.com" rhost-flags-OK-OK-OK-OK) by vger.kernel.org with ESMTP id S1754541Ab0BEQWB convert rfc822-to-8bit (ORCPT ); Fri, 5 Feb 2010 11:22:01 -0500 X-Greylist: delayed 466 seconds by postgrey-1.27 at vger.kernel.org; Fri, 05 Feb 2010 11:22:00 EST Received: by fxm3 with SMTP id 3so4178019fxm.39 for ; Fri, 05 Feb 2010 08:21:59 -0800 (PST) DKIM-Signature: v=1; a=rsa-sha256; c=relaxed/relaxed; d=googlemail.com; s=gamma; h=domainkey-signature:mime-version:received:reply-to:in-reply-to :references:date:message-id:subject:from:to:cc:content-type :content-transfer-encoding; bh=nDhjjDxbFYhnaZ10SZ6GZ6/lnfNt78cdQyXVE0E6fpA=; b=xvCNjgAXORSVkh7Kd9kk59khbPUdR/8CBO3/gGlBG+puuZGuEF1q65EQnwcHamd/Gv J0jWstwV21QDhCmAW6j2HSgY37VzWoGk+6oNcDGcGhilV+2BnXT916eg7/n0tFuvfpum MJvdmZ+iLqqr12jGkA4XY6UbeSx07TXn4sdCE= DomainKey-Signature: a=rsa-sha1; c=nofws; d=googlemail.com; s=gamma; h=mime-version:reply-to:in-reply-to:references:date:message-id :subject:from:to:cc:content-type:content-transfer-encoding; b=dvnVEiNR89KJXnsNZOqGJd30zjtp/NVQ0OA2zrcT9qnCQ36QPcgCY853VjbDywRqms HGjGGj5+qC4BjeUfjsJGhsqB2IjURH466miqvpwLPjeE6mMfTU7bfDXtMFunGnG6gfUq 2kSpEBgbKaI9xRUlJUFsLRExvxOwz3YX50+lI= Received: by 10.216.89.135 with SMTP id c7mr1578717wef.62.1265386452480; Fri, 05 Feb 2010 08:14:12 -0800 (PST) In-Reply-To: <4B6C3E9C.7040009@viscovery.net> Sender: git-owner@vger.kernel.org Precedence: bulk List-ID: X-Mailing-List: git@vger.kernel.org Archived-At: On Fri, Feb 5, 2010 at 4:51 PM, Johannes Sixt wr= ote: > Shawn O. Pearce schrieb: >> Johannes Sixt wrote: >>> What would you think about passing both channels to the async callb= ack, >>> and the communicating parties must agree on which channel they comm= unicate >>> by closing the unused one? It would require slight changes to all c= urrent >>> async users, though. (It also requires in the threaded case that we= pass >>> dup()s of the pipe channels.) >> >> Yup, I could do that. =A0I feel like it might be over-engineering th= e >> solution a bit. =A0But I'll respin the patch by splitting it apart, >> and doing a bidirectional async here, since you asked nicely. > > I do agree about the over-engineering aspect. I mentioned it because = in > one patch in the past Erik Faye-Lund also extended the async > infrastructure for bidirectional communication to use it in git-daemo= n > (Windows port). Just for reference, here's the latest version I wrote of that patch, in case it's useful to have a peak at or something: http://repo.or.cz/w/git/kusma.git/commit/682d90a174fc128910c1c8a4f81edb= 3cf9f0d9e2 --=20 Erik "kusma" Faye-Lund