Gentoo Forums
Gentoo Forums
Gentoo Forums
Quick Search: in
The Politics of systemd Part 3
View unanswered posts
View posts from last 24 hours

Goto page Previous  1, 2, 3, 4, 5 ... 22, 23, 24  Next  
This topic is locked: you cannot edit posts or make replies.    Gentoo Forums Forum Index Gentoo Chat
View previous topic :: View next topic  
Author Message
Tony0945
Advocate
Advocate


Joined: 25 Jul 2006
Posts: 2544
Location: Illinois, USA

PostPosted: Mon Jul 10, 2017 1:34 pm    Post subject: Reply with quote

NeddySeagoon wrote:
How many Microsoft engineers does it take to change a lightbulb ...
One to hold the chair, one to select the bulb,one to turn the bulb, 50 to write the Hot Fixes for the instruction manual.
Back to top
View user's profile Send private message
NeddySeagoon
Administrator
Administrator


Joined: 05 Jul 2003
Posts: 41108
Location: 56N 3W

PostPosted: Mon Jul 10, 2017 1:39 pm    Post subject: Reply with quote

Tony0945,

That's not the one I was thinking of.
_________________
Regards,

NeddySeagoon

Computer users fall into two groups:-
those that do backups
those that have never had a hard drive fail.
Back to top
View user's profile Send private message
Tony0945
Advocate
Advocate


Joined: 25 Jul 2006
Posts: 2544
Location: Illinois, USA

PostPosted: Mon Jul 10, 2017 3:19 pm    Post subject: Reply with quote

Ah! It must be this one:
Quote:
Q: How many Microsoft hardware engineers does it take to change a light bulb?
A: None, they redefine darkness as an industry standard...

Very apropos. The other was for software engineers.


Last edited by Tony0945 on Mon Jul 10, 2017 3:20 pm; edited 1 time in total
Back to top
View user's profile Send private message
depontius
Advocate
Advocate


Joined: 05 May 2004
Posts: 3349

PostPosted: Mon Jul 10, 2017 3:20 pm    Post subject: Reply with quote

NeddySeagoon wrote:
Tony0945,

That's not the one I was thinking of.


None - they declare darkness to be the new standard.

(I presume that's what you were looking for?}
_________________
.sigs waste space and bandwidth
Back to top
View user's profile Send private message
roki942
Apprentice
Apprentice


Joined: 18 Apr 2005
Posts: 267
Location: Seattle

PostPosted: Mon Jul 10, 2017 7:12 pm    Post subject: Reply with quote

NeddySeagoon wrote:
How many Microsoft engineers does it take to change a lightbulb ...

Its much the same with systemd
None --- It's a hardware problem!
*-Not a Bug ---- Won't Fix -*
Back to top
View user's profile Send private message
Naib
Watchman
Watchman


Joined: 21 May 2004
Posts: 5401
Location: Removed by Neddy

PostPosted: Mon Jul 10, 2017 7:16 pm    Post subject: Reply with quote

Tony0945 wrote:
Ah! It must be this one:
Quote:
Q: How many Microsoft hardware engineers does it take to change a light bulb?
A: None, they redefine darkness as an industry standard...

Very apropos. The other was for software engineers.

AND do you know how apt that is for systemd...

6th July reply...
https://lists.freedesktop.org/archives/systemd-devel/2017-July/039221.html

Quote:
Now, because this is so weakly defined, we hence do not follow POSIX
rules, but filter out more that might be dangerous. Specifically:

1. We do not permit empty usernames
2. We don't permit the first character to be numeric
(This also filters out fully numeric user names)
3. We do not permit dots in usernames, neither at the beginning nor in
the middle.
4. We do not permit "-" at the beginning of usernames (something which
POSIX explicitly suggests, btw)
5. We require that the user name fits in the utmp user name field, so
that we can always log properly about it.


This isn't even documented in Systemd
10th july
https://lists.freedesktop.org/archives/systemd-devel/2017-July/039238.html
Quote:
> Is this documented somewhere? If not, it would be great to have it
> documented. I'm pretty sure that this exact paragraph would be ok.

There's a longer (and not entirely complete) comment about this in the
sources, but other than that it's not explicitly documented.


A few utils do restrict this much, but a few don't. what is even more besar is the We do not permit dots in usernames, neither at the beginning nor in
the middle.
YET fedora's adduser regex is: [a-zA-Z0-9._][a-zA-Z0-9._-]{0,30}[a-zA-Z0-9._-$]? which includes dot and dash...

So because systemd implement their own definition of what a valid user is RATHER than just parsing the passwd file (or equiv), quite a few packages are going to be updated.
restricting username starting with a number might be be preference but restricting dot and dash is bad really bad... windows usernames accept that quite happily so now a systemd machine cannot be part of an AD domain and a unit file executed by said user..
_________________
The best argument against democracy is a five-minute conversation with the average voter
Great Britain is a republic, with a hereditary president, while the United States is a monarchy with an elective king
Back to top
View user's profile Send private message
asturm
Developer
Developer


Joined: 05 Apr 2007
Posts: 6376
Location: Austria

PostPosted: Mon Jul 10, 2017 7:35 pm    Post subject: Reply with quote

wow...
_________________
backend.cpp:92:2: warning: #warning TODO - this error message is about as useful as a cooling unit in the arctic
Back to top
View user's profile Send private message
ct85711
Veteran
Veteran


Joined: 27 Sep 2005
Posts: 1603

PostPosted: Mon Jul 10, 2017 8:38 pm    Post subject: Reply with quote

Quote:
So because systemd implement their own definition of what a valid user is RATHER than just parsing the passwd file (or equiv), quite a few packages are going to be updated.
restricting username starting with a number might be be preference but restricting dot and dash is bad really bad... windows usernames accept that quite happily so now a systemd machine cannot be part of an AD domain and a unit file executed by said user..


I think this is really going to end up being a "do as I say, not what I do" kind of thing... After they realize what all this breaks, i.e. breaking samba and AD domains, this will probably end up being silently changed to magically work.
Back to top
View user's profile Send private message
saellaven
Guru
Guru


Joined: 23 Jul 2006
Posts: 478

PostPosted: Mon Jul 10, 2017 8:43 pm    Post subject: Reply with quote

And yet, one of the arguments made by the systemd proponents was that it was better documented than openrc...

Everything that is happening right now is precisely the stuff that, those of us that they labeled "systemd haters" predicted would come true... and we still haven't seen the worst of it yet.

These people have absolutely no clue what they are doing. There is no design that they are implementing. It's just a matter of continuously throwing things into the blob and then encouraging other projects to become dependent upon them.

There's a much, much larger security nightmare waiting to happen and the increasing monoculture (every systemd distribution will eventually be a clone of RedHat) is going to make it worse... this is still just little stuff.
Back to top
View user's profile Send private message
Naib
Watchman
Watchman


Joined: 21 May 2004
Posts: 5401
Location: Removed by Neddy

PostPosted: Mon Jul 10, 2017 8:59 pm    Post subject: Reply with quote

ct85711 wrote:
Quote:
So because systemd implement their own definition of what a valid user is RATHER than just parsing the passwd file (or equiv), quite a few packages are going to be updated.
restricting username starting with a number might be be preference but restricting dot and dash is bad really bad... windows usernames accept that quite happily so now a systemd machine cannot be part of an AD domain and a unit file executed by said user..


I think this is really going to end up being a "do as I say, not what I do" kind of thing... After they realize what all this breaks, i.e. breaking samba and AD domains, this will probably end up being silently changed to magically work.
yup.. one of the great things of linux is its organic evolution... a problem is seen and something comes along and improves things. This is either further improved upon or something else comes along ... the best solution prevails.

This design by committee or design by a fool is wrong... Take this username BS... there is actually only a few really limiting things on what a username can be, a whitespace (delimiter), a colon (again delimiter) etc... Coreutils explicitly states "try as a username then fallback to UID" for this very reason & this is also why a username as all numbers is bad practice.
_________________
The best argument against democracy is a five-minute conversation with the average voter
Great Britain is a republic, with a hereditary president, while the United States is a monarchy with an elective king
Back to top
View user's profile Send private message
Ant P.
Watchman
Watchman


Joined: 18 Apr 2009
Posts: 5135

PostPosted: Mon Jul 10, 2017 10:06 pm    Post subject: Reply with quote

NeddySeagoon wrote:
How many Microsoft engineers does it take to change a lightbulb ...

None - they'll bounce you around a loop of 10 different premium rate numbers until you give up asking.
Back to top
View user's profile Send private message
tld
Veteran
Veteran


Joined: 09 Dec 2003
Posts: 1287

PostPosted: Tue Jul 11, 2017 3:52 pm    Post subject: Reply with quote

Quote:
Now, because this is so weakly defined, we hence do not follow POSIX
rules, but filter out more that might be dangerous.
OMG..."filter out" as in "replace with root"??!!
saellaven wrote:
There's a much, much larger security nightmare waiting to happen and the increasing monoculture (every systemd distribution will eventually be a clone of RedHat) is going to make it worse... this is still just little stuff.
Absolutely. I've been saying all along that the real black hats out there probably already have an arsenal of exploits, and are saving them for that very monoculture...especially once really important servers move to RHEL7, assuming everyone actually falls for that one. Scary stuff.
Tom
Back to top
View user's profile Send private message
gwr
Apprentice
Apprentice


Joined: 19 Nov 2014
Posts: 194

PostPosted: Wed Jul 12, 2017 11:20 am    Post subject: Reply with quote

Quote:
"We require that the user name fits in the utmp user name field, so that we can always log properly about it."


I can't handle this level of stupidity this early in the morning.
Back to top
View user's profile Send private message
Zucca
Veteran
Veteran


Joined: 14 Jun 2007
Posts: 1312
Location: KUUSANKOSKI, Finland

PostPosted: Wed Jul 12, 2017 3:08 pm    Post subject: Reply with quote

gwr wrote:
Quote:
"We require that the user name fits in the utmp user name field, so that we can always log properly about it."


I can't handle this level of stupidity this early in the morning.
I'm rather impressed about this. The level of stupidity is just about to reach 11.
_________________
..: Zucca :..

Code:
ERROR: '--failure' is not an option. Aborting...
Back to top
View user's profile Send private message
Dorsai!
Apprentice
Apprentice


Joined: 27 Jul 2008
Posts: 275
Location: Bavaria

PostPosted: Wed Jul 12, 2017 5:44 pm    Post subject: Reply with quote

Glibc has a performance "regression" with systemd.

Basically they are calling getpid() in a while(1) loop instead of using callbacks.
Of course Lennart is demanding this to be a bug and it being fixed.

Be sure to check out the proposed workaround by Iliana Petrova for some extra gloating.
Back to top
View user's profile Send private message
gwr
Apprentice
Apprentice


Joined: 19 Nov 2014
Posts: 194

PostPosted: Wed Jul 12, 2017 6:03 pm    Post subject: Reply with quote

I can't even word right now.

Dorsai! wrote:
Glibc has a performance "regression" with systemd.

Basically they are calling getpid() in a while(1) loop instead of using callbacks.
Of course Lennart is demanding this to be a bug and it being fixed.

Be sure to check out the proposed workaround by Iliana Petrova for some extra gloating.
Back to top
View user's profile Send private message
Zucca
Veteran
Veteran


Joined: 14 Jun 2007
Posts: 1312
Location: KUUSANKOSKI, Finland

PostPosted: Wed Jul 12, 2017 6:34 pm    Post subject: Reply with quote

If glibc is going to bend under systemd, then I guess I'll go musl libc. I've already been thinking of that, but I need to make sure I can use all the programs I want.
_________________
..: Zucca :..

Code:
ERROR: '--failure' is not an option. Aborting...
Back to top
View user's profile Send private message
asturm
Developer
Developer


Joined: 05 Apr 2007
Posts: 6376
Location: Austria

PostPosted: Wed Jul 12, 2017 6:48 pm    Post subject: Reply with quote

Zucca wrote:
If glibc is going to bend under systemd, then I guess I'll go musl libc.

Don't you think that is *a bit* premature? You have no idea if a resulting patch has any real-world implication for you.
_________________
backend.cpp:92:2: warning: #warning TODO - this error message is about as useful as a cooling unit in the arctic
Back to top
View user's profile Send private message
NeddySeagoon
Administrator
Administrator


Joined: 05 Jul 2003
Posts: 41108
Location: 56N 3W

PostPosted: Wed Jul 12, 2017 6:55 pm    Post subject: Reply with quote

Zucca wrote:
gwr wrote:
Quote:
"We require that the user name fits in the utmp user name field, so that we can always log properly about it."


I can't handle this level of stupidity this early in the morning.
I'm rather impressed about this. The level of stupidity is just about to reach 11.


Just wait a while https://xkcd.com/670/
_________________
Regards,

NeddySeagoon

Computer users fall into two groups:-
those that do backups
those that have never had a hard drive fail.
Back to top
View user's profile Send private message
Ant P.
Watchman
Watchman


Joined: 18 Apr 2009
Posts: 5135

PostPosted: Wed Jul 12, 2017 7:09 pm    Post subject: Reply with quote

asturm wrote:
Zucca wrote:
If glibc is going to bend under systemd, then I guess I'll go musl libc.

Don't you think that is *a bit* premature? You have no idea if a resulting patch has any real-world implication for you.

glibc is already crap.
Back to top
View user's profile Send private message
Zucca
Veteran
Veteran


Joined: 14 Jun 2007
Posts: 1312
Location: KUUSANKOSKI, Finland

PostPosted: Wed Jul 12, 2017 7:09 pm    Post subject: Reply with quote

asturm wrote:
Zucca wrote:
If glibc is going to bend under systemd, then I guess I'll go musl libc.

Don't you think that is *a bit* premature? You have no idea if a resulting patch has any real-world implication for you.
I was rather talking in general. glibc full of tweaks and fetures only useful to systemd... Then I'll change at least.
But I've been thinking of musl in few places. I just need to find out if it's worth it.
_________________
..: Zucca :..

Code:
ERROR: '--failure' is not an option. Aborting...
Back to top
View user's profile Send private message
R0b0t1
Apprentice
Apprentice


Joined: 05 Jun 2008
Posts: 255

PostPosted: Wed Jul 12, 2017 7:44 pm    Post subject: Reply with quote

Zucca wrote:
asturm wrote:
Zucca wrote:
If glibc is going to bend under systemd, then I guess I'll go musl libc.

Don't you think that is *a bit* premature? You have no idea if a resulting patch has any real-world implication for you.
I was rather talking in general. glibc full of tweaks and fetures only useful to systemd... Then I'll change at least.
But I've been thinking of musl in few places. I just need to find out if it's worth it.
Support is noticeably better than it has been so I would invite you to try it. Reportedly you can launch X11 and use most software, but as you start pulling in what you specifically use you will likely find failures that need to be addressed.
Back to top
View user's profile Send private message
Hu
Moderator
Moderator


Joined: 06 Mar 2007
Posts: 12613

PostPosted: Thu Jul 13, 2017 1:00 am    Post subject: Reply with quote

To me, the strange part of this pid cache argument is that using setns to move to a pid namespace only changes the namespace that will be used for new children, but does not move the caller into that namespace. I seem to recall that the cited reason for this was specifically because of concerns that it would confuse userspace to have getpid change its result as a side effect of calling setns. The Red Hat explanation cites concern about the interaction of the cache with pid namespaces, but pid namespaces were intentionally made slightly weird (relative to other namespaces) to avoid those interactions.
Back to top
View user's profile Send private message
Naib
Watchman
Watchman


Joined: 21 May 2004
Posts: 5401
Location: Removed by Neddy

PostPosted: Sat Jul 15, 2017 4:14 pm    Post subject: Reply with quote

https://lkml.org/lkml/2017/7/6/577
Linus wrote:
So I see many different approaches (that could be combined: I like
combining (a) and (c), for example), and absolutely none of them
involve the random "take some values from init".

And yes, a large part of this may be that I no longer feel like I can
trust "init" to do the sane thing. You all presumably know why.

_________________
The best argument against democracy is a five-minute conversation with the average voter
Great Britain is a republic, with a hereditary president, while the United States is a monarchy with an elective king
Back to top
View user's profile Send private message
steveL
Watchman
Watchman


Joined: 13 Sep 2006
Posts: 5099
Location: The Peanut Gallery

PostPosted: Sun Jul 16, 2017 3:27 pm    Post subject: Reply with quote

Dorsai! wrote:
Glibc has a performance "regression" with systemd.

Basically they are calling getpid() in a while(1) loop instead of using callbacks.
Of course Lennart is demanding this to be a bug and it being fixed.
Thanks for the links.
Wow, that was a weird read; the original reporter made a lot of sense, explaining how debian had already patched out the caching in glibc.
The somewhat obtuse response to this (from a regular) was
Quote:
Optimization is never in a vacuum. If glibc does something cheaply, it seems reasonable to take advantage of it.
which is simply bulshytt: it sounds like it's saying something meaningful, but it's just hot air.

The upstream bug clearly shows the behaviour is being removed, and Linus' historical commentary (from 2004, no less) makes it obvious that this has always been a contentious approach; not something to "take advantage of", but something to patch out.

Then you get to Poeterring's reply which is simply bizarre, showing zero insight into the problem, and ending with:
Quote:
we could certainly cache that value in our code too, but given
that glibc does that already in the normal case I think this is better
left to be fixed in glibc rather than our code.
At this point, I simply "cannot word", either. How dumb is this guy?
Naturally, someone picks him up on it, as in: "You do know you have the history backwards, right?" and the response is more bulshytt:
Quote:
Debian undid the PID caching to fix some issue that has been fix
properly now, and hence the PID caching should be turned on again.
showing he still thinks the pid caching is kewl, and hasn't even bothered to fill in the gaps in his knowledge with the historical commentary provided on his own list.
Blathering on about how Fedora gets it right (more 'reason' why "the PID caching should be turned on again") is shot down, and confirmed by one of the regulars, and only then does Poeterring finally come out and acknowledge maybe there's an issue; in someone else's code of course.
Quote:
Seems Fedora regressed on
this too recently. Meh.

Somebody should probably file a bug about this regression and get
clarification if this is going to remain slow, or if they are going to
fix that again... If this is going to remain slow, we shoud probably
find a different way to detect forks...
Note how he completely ignores the detailed information given about glibc moving forward, and indeed the bug report which makes it obvious that the changes are upstream.
No, this dumbass wants more "clarification" from a division of the company he works for.

The bit that really made me laugh was
Quote:
we could install a pthread_atfork() handler and set some flag...
as if he's the guy with the knowledge.
This was suggested to him on the bug report he filed; after admitting he's "Not sure I follow," he goes on to whine about how he "would have preferred if glibc wouldn't regress on this without more consideration."
That just takes the biscuit; there's been a great deal of consideration on this issue, and Poeterring is clearly the guy who moves ahead and shoves out code without thinking, never mind undetaking a bit of basic research.
OFC the guy has to hit him over the head with it again (in comment 7), before he gets it.
I love that he's having to use POSIX interfaces for everything. ;-)

I am sorry to have gone into such detail; I simply could not accommodate the disparity between how everyone else involved takes time to think, and ponder what is said to them, then does some basic research to find out more, before coming back to the list, or bug-report; and how cavalier the guy they are all trusting with their systems is: about the basic approach of learning your tools, and learning the domain, before you even approach implementation.
Reading it, I simply could not believe that this is how programming will be done in the future; not with clarity, focus and humility, but with bluster, hype and bulshytt in place of reason.

I have a feeling we're just watching some of the dinosaurs that won't be around in 20 years, go through their death-throes.
We don't hear about the ones from 30 or 40 years ago, because no-one still uses them, and everyone involved is either dead, or too embarrassed to admit it.

They had all the associated hype, all the brouhaha and marketing campaigns, to cover for the lack of basic competence, too.
Quote:
Be sure to check out the proposed workaround by Iliana Petrova for some extra gloating.
LOL, thanks for the light relief.
Certainly is in line with systemdbust norms; "invalid user? -- run as root." ;) NOT A BUG.

Not a true coder, either.
Back to top
View user's profile Send private message
Display posts from previous:   
This topic is locked: you cannot edit posts or make replies.    Gentoo Forums Forum Index Gentoo Chat All times are GMT
Goto page Previous  1, 2, 3, 4, 5 ... 22, 23, 24  Next
Page 4 of 24

 
Jump to:  
You cannot post new topics in this forum
You cannot reply to topics in this forum
You cannot edit your posts in this forum
You cannot delete your posts in this forum
You cannot vote in polls in this forum