Debian Tech Committee Falling Further Into Disarray

[pal666]

Another excellent piece of writing (an essay, perhaps?) courtesy of Russ Allbery:
https://lists.debian.org/debian-ctte.../msg00390.html

this his writing is not as excellent as usual
he basically says "this judge is a nice guy and i know him for a long time, so let's allow him to judge his relatives"
do you live in jurisdiction where such thing is allowed ? do you want to live in one ?
also, he tells us that people who work on upstart do not have bias. well, he must be using some special definition of bias, but he fails to explain it. in fact, having bias does not require having ill will.
here is what wikipedia says on subject
http://en.wikipedia.org/wiki/Cognitive_bias
http://en.wikipedia.org/wiki/List_of_cognitive_biases

[pal666]

And for the users actually defending the poisonous politics (!)

oh, i almost forgot. please share your deep thoughts on poisonous politics of openssh developers. do you refuse to use openssh due to its explicit non-portability to anything but openbsd ?

[curaga]

hmmmm, do we speak about same os? monolithic kernel and all that, you know... and since most people use unix moto: do 1 thing and do it well, which one thing would that be in linux kernel? do we see buildings crumbling?

this is same BS as people who proclaim wayland is a failure, coz X11 has network transparency. and they say that not even realizing X11 lost that ages ago and all that is there are barely working remnants which cripple while being network transparent

The Linux kernel is indeed too complex, but using it as an excuse to add more failure points is unacceptable. "Shit sucks" is not a reason to "make it suck more".

The same argument could be applied to Wayland incidentally :P X11 networking may suck currently, but the solution is not "let's rip networking out altogether" which would make it suck more.

[pal666]

The Linux kernel is indeed too complex, but using it as an excuse to add more failure points is unacceptable. "Shit sucks" is not a reason to "make it suck more".

computers are complex, do yourself a favor and stop using them

The same argument could be applied to Wayland incidentally :P X11 networking may suck currently, but the solution is not "let's rip networking out altogether" which would make it suck more.

wayland didn't "rip networking out altogether" but made it work at last - more uninformed bs from you

[curaga]

oh, i almost forgot. please share your deep thoughts on poisonous politics of openssh developers. do you refuse to use openssh due to its explicit non-portability to anything but openbsd ?

I do consider the OpenSSH politics bad, though not as bad as systemd ones. I also mainly use dropbear, not OpenSSH.

But I do not refuse to use OpenSSH, because a portable version exists already, and because of that I may ignore the politics there. If there was a systemd fork that fixed the two problems, pid 1 and hostile management, I would evaluate it.

[erendorn]

The same argument could be applied to Wayland incidentally :P X11 networking may suck currently, but the solution is not "let's rip networking out altogether" which would make it suck more.

It is, and has been discussed ad nauseam. Network rendering can only happen within a toolkit (you send the rendering commands. X11 was a toolkit, using other rendering ways broke X11 network rendering). Binding a toolkit to a display protocol suck. So you can't have network rendering in the display protocol. There it is.
You can have remote display, which wayland based compositors accommodate quite nicely. If you want remote rendering, use a remote rendering toolkit, like html5, or add that to an existing native one (gtk/qt).

[pal666]

But I do not refuse to use OpenSSH, because a portable version exists already, and because of that I may ignore the politics there.

well, you only need portable version because you don't use openbsd. if you were using openbsd, you wouldn't need portable version. so either you don't use linux and you are not target audience of systemd(read: nobody cares about your opinion), or you use linux and will happily use systemd just like you happily use openssh. you will post some uninformed bullshit on phoronix, but after that you will follow everyone else into bright systemd future.

btw, you failed to mention your hate towards freebsd's model of development

[justmy2cents]

The Linux kernel is indeed too complex, but using it as an excuse to add more failure points is unacceptable. "Shit sucks" is not a reason to "make it suck more".

The same argument could be applied to Wayland incidentally :P X11 networking may suck currently, but the solution is not "let's rip networking out altogether" which would make it suck more.

soldier, targets are to your left. the ones you keep shooting are scarred spectators aka. you couldn't miss more the point made

it's actually harder keeping current archaic hacks secure in order to keep up with current needs than writing something small and secure a new. not only you keep making hacks in init, you also need to provide hacky way from solutions. whole hw&sw got a lot more complex from 90s on and so are the needs. systemd is not some big unmaintainable code (it;s small compared to kernel) with long lost legacy of what is doing what. check how many contributors there are. not to mention, check mailing list and see how many things gets flat out refused with "we don't want to include it, it belongs elsewhere"

X11 HAS NO NETWORK TRANSPARENCY FOR A LONG TIME. unless, you maybe use Motif. for last few years, all toolkits paint clientside and beg X11 to not touch it. really good video on this http://www.youtube.com/watch?v=F6PFjoYuml0 (then watch this http://www.youtube.com/watch?v=Ux-WCpNvRFM for visual representation of previous video) and comment was not about ripping networking code out, but to rip xorg out and change it with wayland.

sometimes, you need to admit the cow you're betting on is long since it died and got rotten. might be time to get a new one. in translation, if you close your eyes, clap your shoes 3 times and keep repeating "it's rosy and secure....". guess what, init is still deficient and archaic and XOrg still sucks really bad.

[justmy2cents]

I have to disagree, people like this build up a bit of hatred from their fellow teammates until someone has had enough and reveals that discuss plainly. After that many others with a dies taste for an individual feel free to vent their frustration. This isn't the type of personality that you would want to let back in team wise.

I use the word team carefully here because that is what the TC is. If the team can't work together it will degenerate into a bunch of people that only want to argue. The fact is if someone can't follow the majority of the team members wishes expressed in this vote then he is more of a problem than a help.

if one could earn enough respect to be given TC position in as large community as debian... he has damn worked his ass off and removal on 1st point would be nonsense and immature.

note that i don't agree 1 word with Ian on how he did any of the steps in this process. but, sure as hell, he deserves the right to voice it again once he clears his mind.

in my POV... he needs to notice where he made mistake and backtrack with calm head. most notable mistakes being
- approaching vetting option with let's make it equally sucky is not how you create something better
- proposing SIGSTOP while proclaiming standards as reasoning. if any maintainer accepts that in proposed fashion. that is one sure sign project died, vision is not there and no one has a clue wtf they are doing. where do you stop? what other nonsense will be added later? well, if that project is to be taken as a car for example, when someones wife would be sitting in behind the wheel, 1st question would be "so... push break 2x, turn radio on/off, push cluch and break again to turn on the lights?"

[GreatEmerald]

And so systemd wins:

https://lists.debian.org/debian-ctte/2014/02/msg00405.html

Unfortunately, Andreas Barth started rolling the ball on the L voting:

https://lists.debian.org/debian-ctte/2014/02/msg00399.html

I expect Allbery to object, and given the current state of discussions, I don't think this has the chance of passing right now. In fact, I doubt Langasek would vote for it at this point, either, and most of the others disagree with L on principle to begin with.