on beating dead horses (Re: Debian With Alternate Init Systems)
On Mon, Feb 10, 2020 at 06:36:08PM +0100, Samuel Thibault wrote: > (leaving the other parts of threads to later, when I'll get time to > think what useful answer I can give, beyond just apologizing) please dont reply later, everyone is sick of init system discussions on debian-devel@l.d.o. - discussions also don't change anything here. (instead please file bugs, write patches, have a walk outside, read a book, start some other more interesting discussion, work on some uploads, whatever.) -- cheers, Holger --- holger@(debian|reproducible-builds|layer-acht).org PGP fingerprint: B8BF 5413 7B09 D35C F026 FE9D 091A B856 069A AA1C There are no jobs on a dead planet. (Also many other things but people mostly seem to care about jobs.) signature.asc Description: PGP signature
Re: Debian With Alternate Init Systems
Hi! Am 10.02.2020 um 12:18 schrieb Svante Signell: > I would really like to know what the second part of the sentence > "Systemd but we > support exploring alternatives" means. in debian terms, if someone likes to send patches and they don't break anything, they are welcome, but patches which enable systemd specific behavior are not rejected for not supporting other systems. if the patches are nice enough to challenge systemd they will be discussed. greetings
Re: Debian With Alternate Init Systems
Hello, Svante Signell, le lun. 10 févr. 2020 12:18:37 +0100, a ecrit: > On Sat, 2020-02-08 at 16:40 +0100, Samuel Thibault wrote: > > > Nice, the first thing I'll do is to shut down the Debian/GNU Hurd > > > buildd mahler. > > > Can't you see you are here exactly very precisely here *again* putting > > pressure to get something through? > > It is not about pressure, definitely not! That is however how, I believe, everybody understood it. > It's frustration on you trying to punish me for having an opinion. I am not trying to "punish" (I don't even see how that would happen). I am telling Sam that we have never managed to make you change the *way* you are expressing your opinions through pressure. > What do you think would be obtained with the above words? That you perhaps at least understand that what you were writing above *is* pressure, even if you don't realize it when writing it. > BTW: How many euros do you think having a buildd running 24/7 cost me > personally > every month? I'm spending a large amount of my time supporting free software, > including hosting several buildds (without a single (euro)cent in funding). I really appreciate the effort you spend here, I am grateful for that. > > Really, please grow up. You have already worded such a threat in the > > past. That's at best childish. > > I'm asking you to apologize for what you just wrote. The above is an outright > insult to me. Right. I apologize for this sentence, I didn't mean to make it an insult. The thing is: at some point I don't know how to explain that the way you are engaging discussions is most often immediately with big words and everything, and that it's not Okay. On the d-i init question, there hasn't even been any discussion on debian-boot (and thus no worded opposition), and then you already said that some developers "out to leave the project". That's really not a good way to start a discussion... And this is a recurrent scheme I have seen on IRC etc., your starting topics is most often already "isn't it about time to do [...]?". Honestly, when I see such a start, I have no other will than working on something else than what you are talking about, and thus your way of suggesting improvements is actually completely anti-productive. If you simply worded it "could we perhaps do [...]?" things would flow way better. And similarly on d-i init, discussion can lead to compromises. I guess you'll again find this coward, like you already said previously, but really, long-term wise, I have seen it work way better than immediate battles that rather antagonize people. Samuel
Re: Debian With Alternate Init Systems
On Mon, Feb 10, 2020 at 06:27:55PM +0100, Svante Signell wrote: Not much space for other init systems than systemd within Debian. I was hoping for too much. Let's move on with our lives. I think we'd all appreciate if you would do that and stop sending messages about systemd!
Re: Debian With Alternate Init Systems
(leaving the other parts of threads to later, when I'll get time to think what useful answer I can give, beyond just apologizing) Svante Signell, le lun. 10 févr. 2020 18:27:55 +0100, a ecrit: > Regarding your opinion about the GR becomes very clear by reading > https://hartmans.livejournal.com/99395.html again. > Not much more to comment about that. This part is especially interesting: > "Proposal B is a systemd focused proposal. It's very similar to Proposal F. > The > text is different, but the implications of both proposals are similar." > > Not much space for other init systems than systemd within Debian. Please also quote: « My experience is that key maintainers and teams maintaining central infrastructure or packages often need to work with people who are trying to integrate new features. The difference between Proposal B and F is that under Proposal B, we commit to making that happen for technologies that are important in exploring alternatives to systemd. » That *does* leave space for other init systems. Much more than F would have. Samuel
Re: Debian With Alternate Init Systems
Back on debian-devel. On Mon, 2020-02-10 at 14:22 +, Sam Hartman wrote: >From a private reply to me from Sam: > I hear your frustration at Samuel's message. I want Samuel to apologize on _this_ email list for insulting me. Regarding your opinion about the GR becomes very clear by reading https://hartmans.livejournal.com/99395.html again. Not much more to comment about that. This part is especially interesting: "Proposal B is a systemd focused proposal. It's very similar to Proposal F. The text is different, but the implications of both proposals are similar." Not much space for other init systems than systemd within Debian. I was hoping for too much. Let's move on with our lives. Thanks!
Re: Debian With Alternate Init Systems
Sam: This reply is addressed to Samuel and another to you follows, then I'll shut up. On Sat, 2020-02-08 at 16:40 +0100, Samuel Thibault wrote: > > Nice, the first thing I'll do is to shut down the Debian/GNU Hurd > > buildd mahler. > Can't you see you are here exactly very precisely here *again* putting > pressure to get something through? It is not about pressure, definitely not! It's frustration on you trying to punish me for having an opinion. What do you think would be obtained with the above words? BTW: How many euros do you think having a buildd running 24/7 cost me personally every month? I'm spending a large amount of my time supporting free software, including hosting several buildds (without a single (euro)cent in funding). > Really, please grow up. You have already worded such a threat in the > past. That's at best childish. I'm asking you to apologize for what you just wrote. The above is an outright insult to me. And you are talking about pressure! Which is the most polite way to communicate?
Re: Debian With Alternate Init Systems
Sam: This reply is addressed to you, and another to Samuel, then I'll shut up. On Sat, 2020-02-08 at 08:27 -0500, Sam Hartman wrote: > I can't speak for anyone else, but yeah I've come to my own conclusions. > I was fairly open about them: > https://hartmans.livejournal.com/99395.html > > For myself I've concluded that systemd (at least for Linux) is far > better than anything else out there today. My question is merely: You believe that systemd is the best solution at the moment for Linux. Of course your opinion is fully respected. However, there are a number of Debian Developers, Maintainers and Users that would like to use something "inferior" like sysvinit, openrc, runit, s6 or Shepherd. These users are blocked from using these init systems on Debian (they can be used with a lot of manual intervention but that requires non-trivial knowledge). I would really like to know what the second part of the sentence "Systemd but we support exploring alternatives" means. > Something else might come along tomorrow even better than systemd. Thank you for your time!
Re: Debian With Alternate Init Systems
On 08.02.20 22:39, Steve McIntyre wrote: > +1. A reasonable patch to allow this kind of thing would of course be > considered. It could even be pre-seeded to allow for easy control of > it. > I *think* we said something like that ages ago, but no sign of > patches. The installer patches are trivial once debootstrap is able to do that. Patches to support this have already been rejected once, with the rationale that installing with systemd and then running "apt-get -y install sysvinit-core" afterwards should work (ironically, the patch is in a bug report about installing systemd instead of sysvinit, #668001). I wouldn't feel comfortable submitting a debian-installer patch that uses this method, because it lacks the "reasonable" property, but at the same time I don't see much of an alternative here, given that the GR result does not specify a course of action. Simon signature.asc Description: OpenPGP digital signature
Re: Debian With Alternate Init Systems
In article <20200208121622.lu4vla6sasagven2@function> you write: >Svante Signell, le sam. 08 févr. 2020 13:03:20 +0100, a ecrit: >> On Thu, 2020-02-06 at 21:12 +, Sam Hartman wrote: >> > Unfortunately, I think that Debian has decided that's not directly our >> > focus. >> >> Just to make sure: If I submit patches for the Debian installer for support >> of >> more than one init system, they would be outright rejected. > >Nope. Nothing said so. Probably debian-boot wouldn't like to make it a >question asked by default, but in expert mode, probably yes. > >It's not a question for debian-devel or the DPL anyway, please bring it >to debian-boot. +1. A reasonable patch to allow this kind of thing would of course be considered. It could even be pre-seeded to allow for easy control of it. I *think* we said something like that ages ago, but no sign of patches. -- Steve McIntyre, Cambridge, UK.st...@einval.com Armed with "Valor": "Centurion" represents quality of Discipline, Honor, Integrity and Loyalty. Now you don't have to be a Caesar to concord the digital world while feeling safe and proud.
Re: Debian With Alternate Init Systems
Hello, Alf Gaida, le sam. 08 févr. 2020 17:18:09 +0100, a ecrit: > Am 08.02.20 um 16:15 schrieb Svante Signell: > > Anything else? > > > Yes, you forget something - as long the discussion culture in Debian > stays this way i'm not motivated to start my NM process again. Please really do not take Svante as a representative for Debian. He is not a Debian Developer. His way of discussing is *not* representative of the general discussion levels in Debian. Really not. Samuel
Re: Debian With Alternate Init Systems
Am 08.02.20 um 16:15 schrieb Svante Signell: > > Anything else? > Yes, you forget something - as long the discussion culture in Debian stays this way i'm not motivated to start my NM process again. Waste of time to join a bunch of unripe kids. Sorry, but i find no other words with my limited english. Thank you for remind me, will wait maybe another year. Cheers Alf
Re: Debian With Alternate Init Systems
This thread has served its purpose. Continuing this thread is neither consistent with our Code of Conduct nor with the idea of facilitating a productive on-topic discussion within the scope of the debian-devel mailing list. Among other reasons, the discussion is starting to repeat itself and to fail to respect the participants. > "Didier" == Didier 'OdyX' Raboud writes: Didier> Le samedi, 8 février 2020, 16.15:13 h CET Svante Signell a Didier> écrit : >> On Sat, 2020-02-08 at 14:51 +0100, Samuel Thibault wrote: >> > Sam Hartman, le sam. 08 févr. 2020 08:27:24 -0500, a ecrit: > > >> It sounds like you were suggesting that people should leave as a >> way of > > pressuring or punishing the project. >> >> Why not, if the project is heading in the wrong way according to >> their conviction. Some people have already left Debian, as a >> result of the GR. Didier> The Debian project members voted for a project's position Didier> statement of the day, that resulted in option B winning Didier> [0]. You might dislike this result, but option B is the Didier> Debian project's current opinion on Init systems, multiple Didier> init systems, and the use of systemd facilities. Didier> Now with this Debian's position statement in mind, you can Didier> choose to help making (this) Debian better, or you can Didier> choose to put your energy in other projects that pursue Didier> goals more aligned with yours. That's not pushing you out, Didier> that's Debian saying "this is where we're going; are you Didier> going along?". >> > I'd say do not try to fix it, we've been trying unsuccessfully >> within the > Hurd community. >> >> Samuel, so you want me to stop working on Hurd? Nice, the first >> thing I'll do is to shut down the Debian/GNU Hurd buildd mahler. Didier> Threats and emotional blackmail are not welcome in Didier> Debian. Please stop. Didier> (If you don't feel like maintaining an piece of Didier> infrastructure for the Debian project, by all means organize Didier> its replacement to free it from your hands. It's clearly Didier> your right. But don't use work you do for the project as Didier> argumentative lever; that's really not acceptable.) Didier> OdyX Didier> [0] https://www.debian.org/vote/2019/vote_002#textb
Re: Debian With Alternate Init Systems
Le samedi, 8 février 2020, 16.15:13 h CET Svante Signell a écrit : > On Sat, 2020-02-08 at 14:51 +0100, Samuel Thibault wrote: > > Sam Hartman, le sam. 08 févr. 2020 08:27:24 -0500, a ecrit: > > > It sounds like you were suggesting that people should leave as a way of > > > pressuring or punishing the project. > > Why not, if the project is heading in the wrong way according to their > conviction. Some people have already left Debian, as a result of the GR. The Debian project members voted for a project's position statement of the day, that resulted in option B winning [0]. You might dislike this result, but option B is the Debian project's current opinion on Init systems, multiple init systems, and the use of systemd facilities. Now with this Debian's position statement in mind, you can choose to help making (this) Debian better, or you can choose to put your energy in other projects that pursue goals more aligned with yours. That's not pushing you out, that's Debian saying "this is where we're going; are you going along?". > > I'd say do not try to fix it, we've been trying unsuccessfully within the > > Hurd community. > > Samuel, so you want me to stop working on Hurd? Nice, the first thing I'll > do is to shut down the Debian/GNU Hurd buildd mahler. Threats and emotional blackmail are not welcome in Debian. Please stop. (If you don't feel like maintaining an piece of infrastructure for the Debian project, by all means organize its replacement to free it from your hands. It's clearly your right. But don't use work you do for the project as argumentative lever; that's really not acceptable.) OdyX [0] https://www.debian.org/vote/2019/vote_002#textb signature.asc Description: This is a digitally signed message part.
Re: Debian With Alternate Init Systems
Svante Signell, le sam. 08 févr. 2020 16:15:13 +0100, a ecrit: > On Sat, 2020-02-08 at 14:51 +0100, Samuel Thibault wrote: > > Sam Hartman, le sam. 08 févr. 2020 08:27:24 -0500, a ecrit: > > > Svante> Perhaps all [...] ought to leave the project, including > > > Svante> those having package not being dependent on systemd. > > > > > > I am frustrated reading this. > > > It sounds like you were suggesting that people should leave as a way of > > > pressuring or punishing the project. > > Why not, if the project is heading in the wrong way according to their > conviction. Some people have already left Debian, as a result of the GR. Re-read what he wrote. He wrote about pressure and punishment. Not about people just leaving the boat that is not heading the direction that they would like. > > That's the way Svante is always behaving, through diverse forms of > > pressure. > > It is not pressure, it is free speech. No, it really is about pressure. It really seems you do not even realize it. I guess you have spent your whole life doing things this way, and thus don't even realize the mechanism. > > He doesn't seem to know or even be able to learn another way > > of having his goals received. > > Ok, you teach me how to achieve my goals. Doing like you, keep silent and > suffer > from not voicing your opinion? No. By actually explaining what the goals are, and discuss about ways to do it. Yes, that takes time, but that's the only way that will be not hurtful for everybody. That also explains why I'm not doing it for everything I'd like to do (which is fine), and even less for everything *you* would like me to do. Can't you even realize that the very sentence above "keep silent and suffer from not voicing your opinion" is again your putting pressure by trying to make me think I'd be suffering and that I'd have an opinion worth taking the time to express? > > I'd say do not try to fix it, we've been trying unsuccessfully within the > > Hurd > > community. > > Samuel, so you want me to stop working on Hurd? I never wrote anything like that. And I don't mean it either. > Nice, the first thing I'll do is to shut down the Debian/GNU Hurd > buildd mahler. Can't you see you are here exactly very precisely here *again* putting pressure to get something through? Really, please grow up. You have already worded such a threat in the past. That's at best childish. Samuel
Re: Debian With Alternate Init Systems
On Sat, 2020-02-08 at 14:51 +0100, Samuel Thibault wrote: > Hello Sam, > > Sam Hartman, le sam. 08 févr. 2020 08:27:24 -0500, a ecrit: > > Svante> Perhaps all [...] ought to leave the project, including > > Svante> those having package not being dependent on systemd. > > > > I am frustrated reading this. > > It sounds like you were suggesting that people should leave as a way of > > pressuring or punishing the project. Why not, if the project is heading in the wrong way according to their conviction. Some people have already left Debian, as a result of the GR. > That's the way Svante is always behaving, through diverse forms of > pressure. It is not pressure, it is free speech. > He doesn't seem to know or even be able to learn another way > of having his goals received. Ok, you teach me how to achieve my goals. Doing like you, keep silent and suffer from not voicing your opinion? > I'd say do not try to fix it, we've been trying unsuccessfully within the Hurd > community. Samuel, so you want me to stop working on Hurd? Nice, the first thing I'll do is to shut down the Debian/GNU Hurd buildd mahler. Anything else?
Re: Debian With Alternate Init Systems
Hello Sam, Sam Hartman, le sam. 08 févr. 2020 08:27:24 -0500, a ecrit: > Svante> Perhaps all [...] ought to leave the project, including > Svante> those having package not being dependent on systemd. > > I am frustrated reading this. > It sounds like you were suggesting that people should leave as a way of > pressuring or punishing the project. That's the way Svante is always behaving, through diverse forms of pressure. He doesn't seem to know or even be able to learn another way of having his goals received. I'd say do not try to fix it, we've been trying unsuccessfully within the Hurd community. Samuel
Re: Debian With Alternate Init Systems
> "Svante" == Svante Signell writes: >> Yes, this may mean that the support for a particular alternate >> init system may be better in a downstream that focuses on that >> system than in Debian itself. Svante> Perhaps all Debian Developers and Maintainers opposing Svante> systemd ought to leave the project, including those having Svante> package not being dependent on systemd. I am frustrated reading this. It sounds like you were suggesting that people should leave as a way of pressuring or punishing the project. I'm disappointed; I hope for a community that respects all its members better. You and your happiness matter. Debian getting to make decisions and choices for its distributions matters. If you or anyone else would be happier focusing on a downstream, then do that. I hope you do that to make you happier rather than to punish other people. And yeah, if a bunch of people leave Debian because they would be happier elsewhere, we'll have to find people to do their work (or get less work done). However, it is possible to focus on a downstream without leaving. There are great folks in Devuan who are doing the parts of their work they can do upstream in Debian, while focusing on the parts Debian won't accept for Devuan. Debian actually making a decision is good for these people as well as systemd proponents. They know which issues Debian is likely to help with and which issues they are going to get pushback on. Already I've seen issues getting quick responses that might previously have been stalled for years to avoid conflict. Sometimes the responses aren't what people wish to hear, but at least we're respecting them enough to tell them what Debian will accept and what it will not. Many of the conversations are more productive than conversations before the GR. (Some, sadly, are not.) But yeah, some people are bitter, frustrated and angry. If you'd be happier not being involved in Debian, please take care of yourself and leave. If you don't try to burn the house down on the way out, we'll be happy to see you again later if your needs or Debian change. We'd love to hear the cool things you're working on meanwhile. You know what? It doesn't matter to me at least. It's more important to me that people do what makes them happy (while respecting the needs of others including Debian) than that any particular work gets done in Debian. And if not enough people want to work on the things Debian wants to do? Well, Debian will change or it will have served its purpose and move into history. I ask you to respect yourself and us. Let Debian make its decisions. If you find those decisions aren't right for you, I hope you choose to do what is best for you. But I insist that you respect the Debian community enough to let it make its choices and not work to sabotage that. Svante> Just to make sure: If I submit patches for the Debian Svante> installer for support of more than one init system, they Svante> would be outright rejected. I don't know. Samuel has given you an answer. At the DPL level, the GR says that you are welcome to propose patches and that the debian-boot (DI) folks get to use their normal processes to review them. DI is reasonably conservative in what they accept: * I don't think we ended up with even a expert mode switch to turn off merged /usr even though it was proposed. If I'm wrong on this I apologize. * It's been years and we haven't gotten support for blends in tasksel. (Blends is a choice related issue that I think has wider practical support than alternative init systems at the moment.) At one level DI has committed to accepting blends support, but has set the bar high enough that the initial implementation did not meet this bar and we've been blocked on people with very little free time. So, I don't know about alternative init systems, but for a variety of factors, things I think are similar have not yet made it into DI. >> My take on the GR--and certainly my intent in writing proposal >> B--is that would take a new decision of the project. Svante> Maybe another GR? >> I understand that you and many in the project believe that >> existing alternative init systems meet that bar today. My take >> on the GR is that the project as a whole does not support that >> conclusion. Svante> Seems like you (and all other systemd-proponents) already Svante> have made up your mind. So non-systemd for Linux is a dead Svante> horse in Debian (the Universal Operating System!). I can't speak for anyone else, but yeah I've come to my own conclusions. I was fairly open about them: https://hartmans.livejournal.com/99395.html For myself I've concluded that systemd (at least for Linux) is far better than anything else out there today. Something else might come along tomorrow even better than systemd. Svante> My comments above are enough as a answer to this. signature.asc Description: PGP
Re: Debian With Alternate Init Systems
Svante Signell, le sam. 08 févr. 2020 13:03:20 +0100, a ecrit: > On Thu, 2020-02-06 at 21:12 +, Sam Hartman wrote: > > Unfortunately, I think that Debian has decided that's not directly our > > focus. > > Just to make sure: If I submit patches for the Debian installer for support of > more than one init system, they would be outright rejected. Nope. Nothing said so. Probably debian-boot wouldn't like to make it a question asked by default, but in expert mode, probably yes. It's not a question for debian-devel or the DPL anyway, please bring it to debian-boot. > Perhaps [...] ought to leave the project, Please stop bringing in negativity like this: it is useless, whatever the actual intention was. Bringing fight into discussion is *never* the best way to move forward, whatever your goals. Samuel
Re: Debian With Alternate Init Systems
On Thu, 2020-02-06 at 21:12 +, Sam Hartman wrote: > > > > > > "Svante" == Svante Signell writes: > > Svante> When is > Svante> Debian ever to offer a non-systemd alternative for > Svante> installation? > > I hear your frustration that you'd like to see Debian committed to > supporting alternate init systems. > I do understand you value that and have spent a lot of time on it. > > Unfortunately, I think that Debian has decided that's not directly our > focus. Just to make sure: If I submit patches for the Debian installer for support of more than one init system, they would be outright rejected. > In the most recent GR, we committed to providing people the tools they > need to explore and develop alternate init systems. It would be nice to grasp all consequences of the GR. > So, what would it take for us to offer a non-systemd varient for > installation? > Someone would have to develop something that the project looked at and > said, yeah, that's worth supporting as a first-class alternative to > systemd. See above. > My take on the GR--and certainly my intent in writing proposal B--is > that would take a new decision of the project. Maybe another GR? > I understand that you and many in the project believe that existing > alternative init systems meet that bar today. My take on the GR is that > the project as a whole does not support that conclusion. Seems like you (and all other systemd-proponents) already have made up your mind. So non-systemd for Linux is a dead horse in Debian (the Universal Operating System!). > You're welcome to work on any of the existing alternate init systems. > You are welcome to propose patches to make them work better with other > things. People may or may not take those patches. > It's okay even if the reason they don't take those patches is they don't > want to support the alternate init system you are working on. My comments above are enough as a answer to this. > Yes, this may mean that the support for a particular alternate init > system may be better in a downstream that focuses on that system than in > Debian itself. Perhaps all Debian Developers and Maintainers opposing systemd ought to leave the project, including those having package not being dependent on systemd. Looking at the GR vote results that would be a large number of people, effectively reducing the projects ability to create releases. All you non- systemd voters: Consider this option seriously! Thanks!
