|
Sniep posted:i find a bunch for sale on ebay and surplus sites but nobody says what it /does/ other than "analyzer' it can show you the packets on a bus with a standard defined in the early 70s, i dunno how specific i can be without you being able to work out where i work
|
#
?
May 21, 2014 14:16
|
|
|
|
| # ? Jan 18, 2025 15:17 |
|
|
Awia posted:it can show you the packets on a bus with a standard defined in the early 70s, i dunno how specific i can be without you being able to work out where i work im not saying that it was specifically built for this, but thats what we use it for
|
|
#
?
May 21, 2014 14:18
|
|
|
Sniep posted:i find a bunch for sale on ebay and surplus sites but nobody says what it /does/ other than "analyzer' logic analyzer i.e. an o-scope that exists specifically for dealing w/ digital stuff on PCBs
|
|
#
?
May 21, 2014 14:25
|
|
|
itt awia showing us his o-scope
|
|
#
?
May 21, 2014 15:10
|
|
|
Progressive JPEG posted:itt awia showing us his o-scope everyday
|
|
#
?
May 22, 2014 05:20
|
|
|
$3000 for a used one. that lil box was how much new, back when?
|
|
#
?
May 22, 2014 06:24
|
|
|
you can probably get a new one with 1000x the power and features for a 10th of the price but you have to use this one because a document somewhere says so.
|
|
#
?
May 22, 2014 06:26
|
|
|
i think we had actual pci cards that did the same thing at one point, but then another department wanted them and so we gave them all away? i have no idea how much they cost new, an absurd amount to be sure
|
|
#
?
May 22, 2014 07:50
|
|
|
i'm curious. are there any unfixable design flaws or things that could be better in the linux kernel? i'm not talking vague stuff like 3rd party drivers or whatev i am talking about specific things e.g. if the design of threads prevents better memory sharing or whatev basically. if you went back in time and also became linus what are the changes you would make? e: asssuming u have todays tech and languages
|
|
#
?
May 25, 2014 21:18
|
|
|
the internals can be fixed, it's all the POSIX API stuff that we're stuck with and is terrible
|
|
#
?
May 26, 2014 00:20
|
|
|
yeah the posix api is utter poo poo, but sadly, it will never change :/
|
|
#
?
May 26, 2014 00:29
|
|
|
I'd fix the driver APIso that we can more easily suspend and resume them. The current hibernate / suspend tricks we play are terrible, and they break literally every kernel release. The issue is that the internal API for drivers is bad and assumes constant-power, and nobody wants to fix it. Too much work.
|
|
#
?
May 26, 2014 00:32
|
|
|
thing is, the probem seems to me to be more that higher-level apis don't seem to congeal on linux like they do elsewhere. all the primitive apis on windows are *awful*, but there are a bunch of stacks above the cleans up and straightens things out. the ecosystem on linux does for some reason not really push in that direction as naturally. java plasters over all of it nicely, and has a pretty big role in server software as an immediate result
|
|
#
?
May 26, 2014 00:37
|
|
|
Cybernetic Vermin posted:thing is, the probem seems to me to be more that higher-level apis don't seem to congeal on linux like they do elsewhere. all the primitive apis on windows are *awful*, but there are a bunch of stacks above the cleans up and straightens things out. the ecosystem on linux does for some reason not really push in that direction as naturally. java plasters over all of it nicely, and has a pretty big role in server software as an immediate result by this line of thinking, higher-level APIs don't exist on windows either. they get implemented in userspace. new kernel interfaces are rare and often unstable (you are supposed to go through the cool userspace library) there are shitloads of higher-level userspace apis on linux. you already mentioned java. qt is also worth mentioning (it abstracts everything including poo poo like the threading model). the mono ecosystem is cool. perl/python/ruby have worlds unto themselves. being userspace all this wild poo poo is usually ported to other unix systems
|
|
#
?
May 26, 2014 02:16
|
|
|
pseudorandom name posted:the internals can be fixed, it's all the POSIX API stuff that we're stuck with and is terrible not just "can," they actually do the internals get "fixed" all the fuckin time. sweeping changes happen between releases
|
|
#
?
May 26, 2014 02:17
|
|
|
on the weekend i blew away my ubuntu-server NAS and replaced it with wheezy, all went well because of the similarities and i'm happy that i finally have it all documented i wanted to use a more recent sabnzbd though so i added the jessie repos to my sources.list, set "stable" as the default-release in apt.conf, and ran apt-get for sabnzbdplus with -t testing two questions: is that the best way to do it since it's the only package i want from testing? the pinning stuff seemed overkill for this. when apt gets the dependencies for that package, does it install them from testing too or does it get them from stable if the required version is available?
|
|
#
?
May 26, 2014 03:22
|
|
|
you've wasted your weekend
|
|
#
?
May 26, 2014 03:48
|
|
|
gabensraum posted:is that the best way to do it since it's the only package i want from testing? the pinning stuff seemed overkill for this. http://jaqque.sbih.org/kplug/apt-pinning.html p.s. mixing stable and testing is kinda icky/dangerous. mixing testing and unstable is n.b.d. gabensraum posted:when apt gets the dependencies for that package, does it install them from testing too or does it get them from stable if the required version is available? try to satisfy from stable: code:code:
|
|
#
?
May 26, 2014 03:53
|
|
|
Notorious b.s.d. posted:stuff Last Chance posted:you've wasted your weekend all good advice thanks. would prefer not to mix testing but it's for this one app that is otherwise way behind https://packages.debian.org/search?keywords=sabnzbdplus
|
|
#
?
May 26, 2014 03:58
|
|
|
gabensraum posted:all good advice thanks. yeah. just upgrade the whole thing to testing
|
|
#
?
May 26, 2014 04:02
|
|
|
hey is this the point in the conversation where someone is wishing debian had more regular releases so problems like this wouldn't happen so often
|
|
#
?
May 26, 2014 04:05
|
|
|
ShadowHawk posted:hey is this the point in the conversation where someone is wishing debian had more regular releases so problems like this wouldn't happen so often debian has perfectly good regular releases. i am recommending one of the rolling releases. is this the part where you pitch for a year old snapshot of debian with chunks replaced by idiots?
|
|
#
?
May 26, 2014 04:07
|
|
|
Notorious b.s.d. posted:debian has perfectly good regular releases. i am recommending one of the rolling releases.
|
|
#
?
May 26, 2014 04:10
|
|
|
ShadowHawk posted:no not really rolling releases of testing are a reasonable thing for debian to do and I'm not really sure why they bother with stable anymore lol
|
|
#
?
May 26, 2014 04:23
|
|
|
To be clear Debian doesn't really have the infrastructure or organization to do or support stable releases properly. They do, however, have the ability to incorporate growing amounts of automated testing and phased migrations which you absolutely need if you're going to even attempt rolling releases. I'll note that a lot of the testing infrastructure is big in Ubuntu (this is the sort of stuff that benefits both projects when Ubuntu devs put on their Debian hats).
|
|
#
?
May 26, 2014 04:29
|
|
|
you really don't sense the irony in your shitposts do you
|
|
#
?
May 26, 2014 04:32
|
|
|
Notorious b.s.d. posted:you really don't sense the irony in your shitposts do you Debian's ok, Ubuntu is better.
|
|
#
?
May 26, 2014 05:16
|
|
|
a linux is garbage hth
|
|
#
?
May 26, 2014 06:18
|
|
|
this thread is good and informative thanks all  colbylol posted:a linux is garbage hth as opposed to....?
|
|
#
?
May 26, 2014 07:49
|
|
|
osx given by steves (pbuh) own hand
|
|
#
?
May 26, 2014 07:52
|
|
|
theadder posted:osx given by steves (pbuh) own hand *said in geese from fatal fury voice* PREDICTABO
|
|
#
?
May 26, 2014 07:59
|
|
|
maybe canvass the thoughts of gbs
|
|
#
?
May 26, 2014 08:04
|
|
|
theadder posted:maybe canvass the thoughts of gbs
|
|
#
?
May 26, 2014 09:11
|
|
|
Notorious b.s.d. posted:yeah. just upgrade the whole thing to testing cool thanks. had considered that on installation but didn't care how old everything else was and didn't know mixing was fraught
|
|
#
?
May 26, 2014 11:33
|
|
|
how long do you have to use lunix for before you can do the thing you actually wanted to do on the computer instead of just loving with the operating system to cajole it into doing basic poo poo
|
|
#
?
May 26, 2014 11:38
|
|
|
penus de milo posted:how long do you have to use lunix for before you can do the thing you actually wanted to do on the computer instead of just loving with the operating system to cajole it into doing basic poo poo install Ubuntu/mint and you'll be up and running from the get go
|
|
#
?
May 26, 2014 11:54
|
|
|
the time to actual functional use in any linux is infinite, as there will always be one niggling thing stopping you from doing actual work
|
|
#
?
May 26, 2014 11:55
|
|
|
penus de milo posted:how long do you have to use lunix for before you can do the thing you actually wanted to do on the computer instead of just loving with the operating system to cajole it into doing basic poo poo i have no interest in using a desktop linux again any time soon but it's the easiest way to get this servery bullshit done
|
|
#
?
May 26, 2014 11:57
|
|
|
penus de milo posted:how long do you have to use lunix for before you can do the thing you actually wanted to do on the computer instead of just loving with the operating system to cajole it into doing basic poo poo ten years
|
|
#
?
May 26, 2014 19:52
|
|
|
|
| # ? Jan 18, 2025 15:17 |
|
|
saucepanman posted:install Ubuntu/mint and you'll be up and running from the get go yeah don't do this the great thing about ubuntu is i can immediately tell who is an idiot at conferences, tech events etc
|
|
#
?
May 26, 2014 19:52
|
|