Log in

activity wholist changelog info go back go back go forward go forward
burn like an animal - now if you've got a pair of headphones...
you'd better get 'em on and get 'em cranked up
burn like an animal
I really am stunned that anyone thinks this is a worthwhile use of time.

Stunned is one word for it. Dismayed is another. Take your pick.

Tags: ,
music: rob zombie - dragula

pong! (x20) || ping?
From: ext_174884 Date: April 6th, 2009 10:05 pm (UTC) (link)


And you thought _my_ project was useless? :P

Seriously, it's just a nice proof of concept, would be nice if it actually brought some improvement like buildsystems that discover the interfaces instead of assuming to know the environment once they find the uname output. But that's about it I guess.

-- Flameeyes
ajaxxx From: ajaxxx Date: April 6th, 2009 10:59 pm (UTC) (link)

Re: Heh...

I did, and still do.

And you realize that the only reason that's even a concern is because of projects like this that introduce variance where none previously existed. Right?
From: ext_174884 Date: April 6th, 2009 11:21 pm (UTC) (link)

Re: Heh...

Not really. Interfaces are added to various operating systems from time to time; for instance FreeBSD 7 added lots of interfaces that brought it much more on par with glibc than the previous versions.

By the way, the variance added by G/FBSD, to the contrary of Debian's GNU/kFBSD, is not especially cumbersome, it just allows to choose slighty different parts of software stacks which are already available on Linux.
From: (Anonymous) Date: April 6th, 2009 11:13 pm (UTC) (link)

Why not?

Yeah, but who says it has to be a worthwhile use of time? I agree running Debian userspace under a BSD kernel isn't necessarily a useful thing, but hey, it's pretty neat that it works...
ajaxxx From: ajaxxx Date: April 6th, 2009 11:23 pm (UTC) (link)

Re: Why not?

As a first order problem, sure. Do whatever makes you happy, it's free software after all. But the problem I'm often more concerned with is doing things that make the world a better place. Solving problems twice is directly contrary to this goal. The primary value of free software, in fact, is that it aggressively expands the set of problems that have already been solved so no one need ever solve them again.

So if you're going to solve a problem a second time, why do it badly? Why change the underlying kernel if it gains you nothing? Worse, if it creates more work for everyone else - for the installer, and for the X server, and for the C library, and so on - again for no measurable benefit?
zaitcev From: zaitcev Date: April 7th, 2009 01:58 am (UTC) (link)

Re: Why not?

So, down with KDE?
ajaxxx From: ajaxxx Date: April 7th, 2009 02:57 am (UTC) (link)

Re: Why not?

Pretty much, yeah. (Don't read this as fanboying Gnome, more an acknowledgement of a fait accompli.)
From: kevin_kofler Date: April 7th, 2009 11:54 pm (UTC) (link)

Re: Why not?

Where's the "fait accompli"? KDE has a sizable userbase, it may have lost some users due to the KDE 4 migration pain, but GNOME has their GNOME 3 migration still to come. And there are also users who switched from GNOME to KDE because of KDE 4 and its innovation.
neillparatzo From: neillparatzo Date: April 7th, 2009 12:33 am (UTC) (link)
Hedging their bets in case Linus Torvalds dies / is kidnapped for ransom?

Yeah I dunno.
From: kevin_kofler Date: April 8th, 2009 12:01 am (UTC) (link)
Or in case SCO wins their appeal. ;-)

(Or was that thrown out already?)
ivazquez From: ivazquez Date: April 7th, 2009 12:50 am (UTC) (link)
Stupified? Does stupified work?
grok_mctanys From: grok_mctanys Date: April 7th, 2009 02:35 pm (UTC) (link)
Monocultures are bad; diversity is good.

Also, "because it's there" (or "just for fun" if you prefer) is reason enough.
From: ex_cgwalter Date: April 7th, 2009 05:39 pm (UTC) (link)
The linked article is just arguing that one is bad, not that going from 2 (maybe 3) to N is good.

As for the "just for fun" argument, well as someone working just above the kernel stack, I don't find dealing with e.g. N different inotify type APIs "fun". There is a cost to this "fun" for others.
grok_mctanys From: grok_mctanys Date: April 7th, 2009 07:12 pm (UTC) (link)
So which 2 would you rather have? Linux and Windows? Or Linux and BSD? Or Linux and Mac OS X server? Or even BSD and Mac OS X server?

OTOH, I think that the article does imply that the more varied the software "ecosystem" is (to continue the "monoculture" biological analogy) the safer things are in general. If you have a monoculture, a pathogen (e.g. virus) can take out 100% of your infrastructure. If you have a bi-culture(?), a virus should only take out 50%. If you have 3, that drops to 33%, and 4 implies 25%. The more varied your systems, the more robust they become.

If you don't like dealing with multiple, e.g. inotify APIs, then I don't see how this suddenly forces you to. If you currently don't support BSD even though BSD clearly exists, why do you imagine that you'll suddenly have to support Debian GNU/BSD just because it exists?
From: ex_cgwalter Date: April 7th, 2009 07:54 pm (UTC) (link)
My personal preferences are irrelevant (but obvious if you know who employs me).

What is relevant is the article, or more precisely the lack thereof. We could argue about "implies", but the point is that's *not* what they're arguing.
grok_mctanys From: grok_mctanys Date: April 7th, 2009 08:30 pm (UTC) (link)
Not what they're saying? Really? From the Executive Summary:
Most of the world’s computers run Microsoft’s operating systems, thus most of the world’s computers are vulnerable to the same viruses and worms at the same time. The only way to stop this is to avoid monoculture in computer operating systems, and for reasons just as reasonable and obvious as avoiding monoculture in farming.

Because Microsoft's near-monopoly status itself magnifies security risk, it is essential that society become less dependent on a single operating system from a single vendor if our critical infrastructure is not to be disrupted in a single blow. The goal must be to break the monoculture.

and from Section 1 - The Problem in Principle
• A monoculture of networked computers is a convenient and susceptible reservoir of platforms from which to launch attacks; these attacks can and do cascade.
• This susceptibility cannot be mitigated without addressing the issue of that monoculture.
Risk diversification is a primary defense against aggregated risk when that risk cannot otherwise be addressed; monocultures create aggregated risk like nothing else.

(emphasis mine)

I read that as arguing that the more diverse our systems are, the safer and better off we all are.

I see that it could be read as saying that 2 different systems are sufficient for safety, but ... *shrug* I guess it's not that big a deal.

On the other hand, you complain about supporting 2 different flavours of inotify. Surely supporting two different implementations of roughly the same API (e.g. Linux and BSD via POSIX) with a few differences here and there (e.g. inotify) would be easier than supporting two completely different APIs (POSIX + WinAPI).
hub_ From: hub_ Date: April 7th, 2009 07:27 pm (UTC) (link)
At least it will bring a better package management to FreeBSD.

But then I agree it just make the combinational space of software interaction much larger.

zaitcev From: zaitcev Date: April 7th, 2009 09:49 pm (UTC) (link)
Only if you care. I'm actually thinking about dropping autotools and just ship three makefiles for Win, Solaris, and Linux.
From: (Anonymous) Date: April 10th, 2009 08:50 am (UTC) (link)

Unifiying communities

It has managed to unify *BSD and Linux communities - both appear to passionately hate this effort :P

From: evan Date: April 10th, 2009 10:54 pm (UTC) (link)

i think it's cool

I would've picked FreeBSD over Linux back in the day if the userspace could've been Debian. A old friend of mine is one of the core FreeBSD devs and would often rant about various Linux crazinesses.
pong! (x20) || ping?