Re: GR: Selecting the default init system for Debian
Le Thu, Jan 23, 2014 at 04:14:41AM +0100, Wouter Verhelst a écrit : On Thu, Jan 23, 2014 at 09:58:14AM +0900, Charles Plessy wrote: In that case, I think that the project should decide via using this or that system (“vote with the feet”). For the packages where init scripts are a limitation, just depend on systemd, upstart, openrc, or combinations of them, and if and only if it is not possible to install Debian because pairs of core packages depend on different single init systems, let's vote. So, let me get this straight. Hi Wouter, OK, let's be straight :) You're saying let's do nothing until the entire system breaks because of a component that nobody really cares about, so that we can _then_ try to start a procedure which will take weeks (if not months) to maybe unbreak it, leaving the system in an utterly broken state in the meantime? What I am saying is: Let's allow the Debian system to evolve freely: the result will not be breakage, but systemd as a de facto default. If some parts of the system become mutually exclusive, I do not see it as problematic. We do not support the co-installation of some mail or FTP servers packages even though in theory one can configure them to listen on different ports; if tomorrow one desktop manager depends on upstart and another on systemd, then the solution is to call this unsupported as well. I would also argue the same if it were web browsers. I would call a system broken if it would not be possible to do anything useful with any of the init systems. I do not see how this could happen. First, these init systems are developed and tested on computers that run them, such as Fedora, Ubuntu and Gentoo, which shows that there is no critical missing piece in one or the other system in the context where they are intended for. Second, at least systemd runs fine on Debian currently, and to my knowledge, there are no core components that are likely to drop systemd support in the near future. Then, there is the fear that because systemd or upstart is much easier to support than our current init system, the non-Linux architectures that can not run them will dissapear because init scripts will be dropped massively. To me it is a total overstatement. What is at stakes is whether these ports will benefit as much as before from the work on mainstream systems such as Linux on amd64. The answer with is “no”, unless we enforce a default with this goal in mind, that will cost to others what it gains to the non-Linux architectures. But that “no” does not make these projects impossible. At worse, it will force them to focus on their userbase instead of working on total coverage of the Debian package supermaket, and I think is would actually be a good change (please do not waste your time sending patches to leaf packages until you know that somebody is actually planning to use them on your port). Lastly, there is the political part. Should we boycott systemd or upstart ? Should we make choices that in practice mean to show the door to our GNOME team ? Should we push even more our contributors to participate to the porting on specialised architectures ? Let's releive the technical comittee from the pressure to step in that field. The best reaction to these questions is to ignore them. So definitely, thanks Steve and the sysinit maintainers for making transition between init systems easier; with this I do not think that we need a decision on a default system anymore. Cheers, -- Charles Plessy Tsurumi, Kanagawa, Japan -- To UNSUBSCRIBE, email to debian-vote-requ...@lists.debian.org with a subject of unsubscribe. Trouble? Contact listmas...@lists.debian.org Archive: http://lists.debian.org/20140125112035.gl24...@falafel.plessy.net
Proposed amendment (was: Re: GR: Selecting the default init system for Debian)
On Sat, Jan 25, 2014 at 08:20:35PM +0900, Charles Plessy wrote: Le Thu, Jan 23, 2014 at 04:14:41AM +0100, Wouter Verhelst a écrit : On Thu, Jan 23, 2014 at 09:58:14AM +0900, Charles Plessy wrote: In that case, I think that the project should decide via using this or that system (“vote with the feet”). For the packages where init scripts are a limitation, just depend on systemd, upstart, openrc, or combinations of them, and if and only if it is not possible to install Debian because pairs of core packages depend on different single init systems, let's vote. So, let me get this straight. Hi Wouter, OK, let's be straight :) You're saying let's do nothing until the entire system breaks because of a component that nobody really cares about, so that we can _then_ try to start a procedure which will take weeks (if not months) to maybe unbreak it, leaving the system in an utterly broken state in the meantime? What I am saying is: Let's allow the Debian system to evolve freely: the result will not be breakage, but systemd as a de facto default. This argument has been brought up before (indeed, even by me), and has been debunked by several people. There are several problems with that approach for the choice of init system: First, a change of MTA, FTP server, or browser produces a user-visible change in an area that most users will care about. An init system does not--and note the most in the previous sentence. Second, it is possible to define the interface which an MTA, FTP server, or web server should provide to the rest of the system (e.g., serve files in this directory by default, or send mails to remote users if passed to the sendmail binary) without going into too much technical detail on how exactly the MTA, FTP daemon, or web server needs to do so, and also without sacrificing features that we might want. The same is not true for init systems. For instance, while we could just declare that all packages need to provide initscripts (which then means that even sysv-rc could still be used), that really is just the status quo, and we might as well not bother. I am personally convinced that we *do* need a better init system. I don't actually care _which_ init system that is, and am contend to leave that decision to the people who do. But we should not retain the status quo. If you think systemd will become the de facto default, then why not just throw out the years of bickering and bikeshedding and just decide that _now_? We should have made a decision on this subject years ago. The debate is reducing the quality of our mailing lists, is holding the entire project hostage, and we're *still* no closer to an answer. Even the TC seems to be having difficulties reaching a decision. So, let me propose the following amendment, then: - If this option wins, the project secretary, in the presence of at least two other Debian Developers, will roll a dice. If the dice comes at rest with 1 or 2 facing up, systemd will become the default init system for Debian. If the dice comes at rest with 3 or 4 facing up, upstart will become the default init system for Debian. If the dice comes at rest with 5 or 6 facing up, openrc will become the default init system for Debian. - I am looking for seconds. And no, that's not a joke; at this stage the debate is essentially deadlocked, and I am doubtful that the debate will *ever* reach a conclusion which will be the best on a technical and/or political level. All available options feature some things that the others don't, all have downsides, and none of the available options will ever be a perfect solution. We could discuss this ad infinitum and end up with a non-solution, or we could just bite the bullet and make a decision. At this point, I think any decision is better than no decision, even if that decision is the throw of a dice. -- This end should point toward the ground if you want to go to space. If it starts pointing toward space you are having a bad problem and you will not go to space today. -- http://xkcd.com/1133/ signature.asc Description: Digital signature
Re: Proposed amendment (was: Re: GR: Selecting the default init system for Debian)
Hi, On Samstag, 25. Januar 2014, Wouter Verhelst wrote: So, let me propose the following amendment, then: - If this option wins, the project secretary, in the presence of at least two other Debian Developers, will roll a dice.[...] - I am looking for seconds. And no, that's not a joke; Well, my option this is hard, my brain hurts, lets go shopping was a joke and essentially the same as the above. If you cannot wrap your mind around a problem, please dont declare defeat for the whole project or propose silly solutions. Just because the problem is too hard for some, doesnt mean there aint sensible solutions. Rolling a dice aint one of them. cheers, Holger P.S.: Also, btw unrelated: reasonable dices have 20 sides. Or 4 or 12. signature.asc Description: This is a digitally signed message part.
Re: GR: Selecting the default init system for Debian
Before I forget, there's one thing I wanted to say about this: On Sun, Jan 19, 2014 at 01:01:44AM +0100, Guillem Jover wrote: [...] Option A [...] Option B [...] Option C [...] Option D [...] Option E [...] Option F [...] Option G [...] Please don't do that. If you want to propose a GR, please only propose those options that you actually want to see win. When seconding, please only second those options that you actually want to see win (or lose against NOTA, so nobody will ever bring it up again). A ballot with too many options is never a good ballot, and no matter how hard you try you'll always miss one or two possibilities. That would mean we'd get a ballot with 8 or 9 options, which is too many in my opinion. When drafting a GR text for an option that you think is not the best option, the result will be that you'll end up with a text that those people who *do* think is the best option don't want to support. They'll not be willing to vote for or second those options then, or (worse) will try to propose their own amendment which is different, but only in small details--resulting in yet *another* option on the ballot. If enough people actually think some option in your list deserves being put on the ballot, rest assured they'll propose amendments and get them seconded. I have enough faith in our developers that this will happen. -- This end should point toward the ground if you want to go to space. If it starts pointing toward space you are having a bad problem and you will not go to space today. -- http://xkcd.com/1133/ signature.asc Description: Digital signature
Re: Proposed amendment (was: Re: GR: Selecting the default init system for Debian)
On Sat, Jan 25, 2014 at 05:31:50PM +0100, Holger Levsen wrote: Hi, On Samstag, 25. Januar 2014, Wouter Verhelst wrote: So, let me propose the following amendment, then: - If this option wins, the project secretary, in the presence of at least two other Debian Developers, will roll a dice.[...] - I am looking for seconds. And no, that's not a joke; Well, my option this is hard, my brain hurts, lets go shopping was a joke and essentially the same as the above. If you cannot wrap your mind around a problem, please dont declare defeat for the whole project or propose silly solutions. Just because the problem is too hard for some, doesnt mean there aint sensible solutions. Rolling a dice aint one of them. I'm not saying that rolling a dice is the best option. But I *do* think it is a better option than 'further discussion', so if this ever gets to a vote I will most definitely rank this above NOTA. We need to make a decision on this subject. I'm still hoping the TC will be able to make that decision, but it remains possible that they don't. If that is the case, and this does come to a vote, I want to have this option on the ballot. Think of it as a last resort. I do want to go with the technically correct choice, if we as a project can make it. But if the technical committee fails to make a decision, and if a GR does the same, we'd end up with no decision. If given that option and rolling a dice, then I think rolling a dice is the lesser of the two evils. -- This end should point toward the ground if you want to go to space. If it starts pointing toward space you are having a bad problem and you will not go to space today. -- http://xkcd.com/1133/ signature.asc Description: Digital signature
Re: Proposed amendment (was: Re: GR: Selecting the default init system for Debian)
Hi! On Sat, 2014-01-25 at 17:06:45 +0100, Wouter Verhelst wrote: On Sat, Jan 25, 2014 at 08:20:35PM +0900, Charles Plessy wrote: What I am saying is: Let's allow the Debian system to evolve freely: the result will not be breakage, but systemd as a de facto default. This argument has been brought up before (indeed, even by me), and has been debunked by several people. I very much disagree this argument has been debunked, I do accept other people have a different opinion on it though, but there's no point in discussin this here. I am personally convinced that we *do* need a better init system. I don't actually care _which_ init system that is, and am contend to leave that decision to the people who do. But we should not retain the status quo. We should have made a decision on this subject years ago. The debate is reducing the quality of our mailing lists, is holding the entire project hostage, and we're *still* no closer to an answer. Even the TC seems to be having difficulties reaching a decision. I was letting at least one week pass, to possibly get input from other people, or from the secretary, and I was/am planning on looking for sponsors on a revised GR draft tomorrow (including the defer to TC option and reworded Option B). If there's any conflict with the running TC resolution, the secretary can point it out before the vote, if enough people sponsor the GR, of course. So, let me propose the following amendment, then: - If this option wins, the project secretary, in the presence of at least two other Debian Developers, will roll a dice. If the dice comes at rest with 1 or 2 facing up, systemd will become the default init system for Debian. If the dice comes at rest with 3 or 4 facing up, upstart will become the default init system for Debian. If the dice comes at rest with 5 or 6 facing up, openrc will become the default init system for Debian. - I am looking for seconds. And no, that's not a joke; at this stage the debate is essentially deadlocked, and I am doubtful that the debate will *ever* reach a conclusion which will be the best on a technical and/or political level. All available options feature some things that the others don't, all have downsides, and none of the available options will ever be a perfect solution. We could discuss this ad infinitum and end up with a non-solution, or we could just bite the bullet and make a decision. At this point, I think any decision is better than no decision, even if that decision is the throw of a dice. Ok, given what you mentioned above, your preference is not easily represented with the current GR draft, and I don't think this amendment makes much sense (at least to me). You want a change, but don't care which; in which case I think it would be more appropriate to let the people who care decide, as you pointed out. I could see a decision by dice, being questioned as non-transparent, etc. But could see an option that essentially says (with better wording and all that): * Switch the init system to something else than sysvinit + sysv-rc. - a decision for a new init system needs to be made now, letting this undecided will keep causing frustration and project tension. - the init system chosen will be the one the project at large has a preference on, by selecting the winning option among options C-G. If something along those lines satisfies you, I'm happy to include a polished version in the GR draft. Thanks, Guillem -- To UNSUBSCRIBE, email to debian-vote-requ...@lists.debian.org with a subject of unsubscribe. Trouble? Contact listmas...@lists.debian.org Archive: http://lists.debian.org/20140125165047.ga27...@gaara.hadrons.org
Re: Proposed amendment (was: Re: GR: Selecting the default init system for Debian)
Hi, On Samstag, 25. Januar 2014, Wouter Verhelst wrote: But if the technical committee fails to make a decision, and if a GR does the same, we'd end up with no decision. No. __If__ that happens, we'd end up with a decision keep the status quo, aka keep sysv as the default init system. That would be a valid decision. (And I don't understand why you didn't have it on your dice roll...) cheers, Holger signature.asc Description: This is a digitally signed message part.
Re: Proposed amendment (was: Re: GR: Selecting the default init system for Debian)
On Sat, Jan 25, 2014 at 05:50:47PM +0100, Guillem Jover wrote: On Sat, 2014-01-25 at 17:06:45 +0100, Wouter Verhelst wrote: [...] So, let me propose the following amendment, then: - If this option wins, the project secretary, in the presence of at least two other Debian Developers, will roll a dice. If the dice comes at rest with 1 or 2 facing up, systemd will become the default init system for Debian. If the dice comes at rest with 3 or 4 facing up, upstart will become the default init system for Debian. If the dice comes at rest with 5 or 6 facing up, openrc will become the default init system for Debian. - I am looking for seconds. And no, that's not a joke; at this stage the debate is essentially deadlocked, and I am doubtful that the debate will *ever* reach a conclusion which will be the best on a technical and/or political level. All available options feature some things that the others don't, all have downsides, and none of the available options will ever be a perfect solution. We could discuss this ad infinitum and end up with a non-solution, or we could just bite the bullet and make a decision. At this point, I think any decision is better than no decision, even if that decision is the throw of a dice. Ok, given what you mentioned above, your preference is not easily represented with the current GR draft, and I don't think this amendment makes much sense (at least to me). Well, that's of course your prerogative, but the fact that you came up with a long list of options doesn't negate my right to attempt to add another option, even if you think it doesn't belong there. You want a change, but don't care which; in which case I think it would be more appropriate to let the people who care decide, as you pointed out. That would of course be the best option, and I would be happy if we were to reach that. I could see a decision by dice, being questioned as non-transparent, etc. Hence the bit about two other DDs need to be present. I suppose we could possibly require a video recording. Alternatively, we could choose some factoid about the vote itself; like number of votes received modulo 3 decides the winner, or (all timestamps on all mails sent and received by devotee during the course of this vote represented in unix epoch, added together), modulo 3 decides the winner, or some other variation on that theme. The point is to essentially pull a decision out of thin air if all other attempts to make a decision failed. We need a decision. I don't care what that decision is, but we need one. Since a few months, this endless debate has reached the point where every thread on every mailinglist, given enough time, eventually turns into yet another instance of the init system debate. This is unhealthy for our community and needs to stop. But could see an option that essentially says (with better wording and all that): * Switch the init system to something else than sysvinit + sysv-rc. - a decision for a new init system needs to be made now, letting this undecided will keep causing frustration and project tension. - the init system chosen will be the one the project at large has a preference on, by selecting the winning option among options C-G. If something along those lines satisfies you, I'm happy to include a polished version in the GR draft. That would introduce interdependencies between votes, which has a serious risk of skewing the result (e.g., people would feel more compelled to rank one option lower, so that the chance of it winning indirectly through this option gets smaller; that would mean they wouldn't be expressing their actual opinion). I don't think that's a good idea. -- This end should point toward the ground if you want to go to space. If it starts pointing toward space you are having a bad problem and you will not go to space today. -- http://xkcd.com/1133/ signature.asc Description: Digital signature
Re: Proposed amendment (was: Re: GR: Selecting the default init system for Debian)
On Sat, Jan 25, 2014 at 05:58:57PM +0100, Holger Levsen wrote: Hi, On Samstag, 25. Januar 2014, Wouter Verhelst wrote: But if the technical committee fails to make a decision, and if a GR does the same, we'd end up with no decision. No. __If__ that happens, we'd end up with a decision keep the status quo, aka keep sysv as the default init system. That would be a valid decision. (And I don't understand why you didn't have it on your dice roll...) Because I _have_ been convinced that we should replace sysv-rc by something else. I don't care what that something else is, but we need to do it. I'm also doubtful that if 'further discussion' wins this vote, the endless debate will end. Quite the contrary. -- This end should point toward the ground if you want to go to space. If it starts pointing toward space you are having a bad problem and you will not go to space today. -- http://xkcd.com/1133/ -- To UNSUBSCRIBE, email to debian-vote-requ...@lists.debian.org with a subject of unsubscribe. Trouble? Contact listmas...@lists.debian.org Archive: http://lists.debian.org/20140125173001.ge22...@grep.be
Re: Proposed amendment (was: Re: GR: Selecting the default init system for Debian)
On Sat, 2014-01-25 at 18:15:46 +0100, Wouter Verhelst wrote: On Sat, Jan 25, 2014 at 05:50:47PM +0100, Guillem Jover wrote: Ok, given what you mentioned above, your preference is not easily represented with the current GR draft, and I don't think this amendment makes much sense (at least to me). Well, that's of course your prerogative, but the fact that you came up with a long list of options doesn't negate my right to attempt to add another option, even if you think it doesn't belong there. Oh, absolutely, it just seemed a bit at odds with what you had just mentioned before. And in any case, I was trying to find an option or a vote solution that might satisfy your preference (which I've also seen elsewhere), and to avoid ending up with a monster ballot, with many options with very small variations over mostly the same. You want a change, but don't care which; in which case I think it would be more appropriate to let the people who care decide, as you pointed out. That would of course be the best option, and I would be happy if we were to reach that. I could see a decision by dice, being questioned as non-transparent, etc. Hence the bit about two other DDs need to be present. I suppose we could possibly require a video recording. Alternatively, we could choose some factoid about the vote itself; like number of votes received modulo 3 decides the winner, or (all timestamps on all mails sent and received by devotee during the course of this vote represented in unix epoch, added together), modulo 3 decides the winner, or some other variation on that theme. The point is to essentially pull a decision out of thin air if all other attempts to make a decision failed. Yes also considered the video, but still, what about the dices themselves, etc; the other options you mention are a bit better. But in any case I don't think it's a good idea, really, it would probably piss off anyone who cares about a specific choice, and people would keep challenging the vote, becuse it was random. Obviously if you still think it is a good idea and others agree, then it will end up being included. We need a decision. I don't care what that decision is, but we need one. Since a few months, this endless debate has reached the point where every thread on every mailinglist, given enough time, eventually turns into yet another instance of the init system debate. This is unhealthy for our community and needs to stop. Well, I think a vote by the project, whatever the outcome (status quo, postpone, specific option), will be a clear message that people would stop going on about it (at least for a while :). But could see an option that essentially says (with better wording and all that): * Switch the init system to something else than sysvinit + sysv-rc. - a decision for a new init system needs to be made now, letting this undecided will keep causing frustration and project tension. - the init system chosen will be the one the project at large has a preference on, by selecting the winning option among options C-G. If something along those lines satisfies you, I'm happy to include a polished version in the GR draft. That would introduce interdependencies between votes, which has a serious risk of skewing the result (e.g., people would feel more compelled to rank one option lower, so that the chance of it winning indirectly through this option gets smaller; that would mean they wouldn't be expressing their actual opinion). I don't think that's a good idea. Actually I agree, because I just realized your preference (I think) can actually be represented with the current ballot. You could rank all options that specify a change from the current default above NOTA, and the reset below. If you (as in the generic voter), really don't care can rank all options above NOTA equally. Given this I'm not planning on adding such option. Thanks, Guillem -- To UNSUBSCRIBE, email to debian-vote-requ...@lists.debian.org with a subject of unsubscribe. Trouble? Contact listmas...@lists.debian.org Archive: http://lists.debian.org/20140125212413.ga31...@pulsar.hadrons.org
