I was requested to forward the following mail by Sven Luther: ----- Forwarded message from Sven Luther <s...@powerlinux.fr> -----
From: Sven Luther <s...@powerlinux.fr> To: Gunnar Wolf <gw...@gwolf.org>, listmas...@debian.org Cc: Romain Beauxis <to...@rastageeks.org>, debian-devel@lists.debian.org, debian-v...@lists.debian.org Date: Wed, 25 Mar 2009 07:01:17 +0100 Subject: Re: [dissenting]: Proposal: Enhance requirements for General resolutions Message-ID: <20090325060117.ga19...@powerlinux.fr> References: <87vdq3gcf6....@vorlon.ganneff.de> <20090322225302.ga24...@yellowpig> <200903240112.34470.to...@rastageeks.org> <20090325035739.gf8...@cajita.gateway.2wire.net> X-Spam-Level: X-Spam-Status: No, score=-2.6 required=5.0 tests=AWL,BAYES_00 autolearn=ham version=3.1.7-deb3 On Tue, Mar 24, 2009 at 09:57:39PM -0600, Gunnar Wolf wrote: > > Romain Beauxis dijo [Tue, Mar 24, 2009 at 01:12:34AM +0100]: > > Le Sunday 22 March 2009 23:53:02 Bill Allombert, vous avez écrit : > > > Furthermore I am a Debian since 2001 and I see no evidence than the GR > > > process was abused during that time. On the contrary, some GR were delayed > > > to the point where it was inconvenient for the release process. > > > > I agree. I fail to see where the GR process was abused. Since that seems > > the > > main argument in favour of this change, I fail to see the motivation for it. > > This proposal does not come from an abuse to the GR process, but to > generalized frustration about the way 2008_002 and specially 2008_003 > were handled. But the reason for this are in no way related with the number of seconds, but rather in the way the debian project considers discussion consensus and such. It seems to me that most people see it more as a blood sport where everything is fine as long as his ideas win, than a process where there is respect for the ideas and convictions of others. In general, we should revisit the way we handle GRs, go away from the current process, where the first step of the vote is to make sure only ideas which have your support get on the ballot, instead of searching a ballot whose many options may give a chance of the voters to represent every possible opinion. I strongly believe that the amendment process is the one who is responsible for this issue. The current proposal is only a stop-gap way of trying to limit votes, and doesn't consider the real issue. Voluntarily not considering darker motives which come to mind when reading this proposal and seeing the position of the proposers. The proposers should keep in mind that this proposal can be interpreted in such a darker way given a certain degree of resentment of the project toward their high-handness, but that is another issue. In general, the GRs who turned the more disastrous (such as vorlon's solo firmware GR, bypassing the kernel team's reflexion on the subject) are often perceived as a way to force an opinion because one moves first, and is more vocal about it. Many of the votes are of the "let's vote, and be done with it, we would much prefer to work on technical stuff" category. We should modify the GR process to be something like : 1) Some DDs (5, 15, whatever) decide to have a vote about a topic. There is no actual text yet at this stage, just a topic, and the DDs have to give a motivation about why they want to have this topic voted on. 2) The main proposer of the vote is then made responsible of drafting a ballot, which will have enough orthogonal options to represent all the current of opinions in the project. To do this, he helds a discussion on -vote, whose objective is not to defend ones idea, but to make sure every current of ideas in debian is represented on the ballot. This step should be non confrontational, and not lead to wild debates. options should be added liberally, without the need of seconds, and are of the responsability of the proposer. The ballot options each should get a rationale and description as part of this process 3) Once the ballot is ready, the proposed ballot is posted on d-d-a or some other list reaching every developer, and a period of time (1 week ?) is set for people who missed step 2 to object to the ballot. During step 2 and 3, if the responsible of the vote proves stubborn, or refuses to add options, an appeal to the secretary, DPL or technical committee should be possible to avoid problems and couter-balance the power of the responsible of drafting the ballot. 4) if after the ballot scrutinization period, no objections where made, the ballot is put to vote. 5) a heated discussion period can be had to defend the different ballot opinions, but this heated discussion is not weaved with using amendmens to confuse the issue, or tentatives to subvert existing proposals by subtle modifications of the text by seemingly innocent amendmens quickly accepted by the original proposer. => This would allow us to : 1) have a trully representative ballot 2) limit the heated discussion period of a vote (it is good to have it, because it is the essence of democracy, but a too long discussion who result in a ballot not representative of each expressed opinion is what leads to resentment and frustration). 3) have an easier to read ballot, which would make the result of the vote more meaningful. 4) attract more voters, for which the current electoral process is something to be shuned as a major flamewar where one gets hurt for participating in it. Gunnar, please forward this mail to the list, since i am being censored (altough it seems the DAM and others feel offended for me using the word censored, with its darker connotations. Well, if they dislike it, they should not do it). Listmasters, i ask you to lift the censorship for the purpose of participating in this discussion, or to name an official censor whom i can CC to make sure my emails fully go to the list, and not part of them being left in the void as happened last time i tried to participate in such a discussion. Sadly, Sven Luther ----- End forwarded message ----- -- To UNSUBSCRIBE, email to debian-devel-requ...@lists.debian.org with a subject of "unsubscribe". Trouble? Contact listmas...@lists.debian.org