* EthanG gets to the end of the 2nd sentence & automatically checks the date on the mail
* uriel goes to sleep too
* uriel has never quite understood dwm, the defaults are were too broken every time that I tried it
Thursday, March 11, 2010
why is netbsd more harmful than openbsd?
* happy (~g@w005.z209031033.sjc-ca.dsl.cnc.net) has joined #cat-v
ok
got netbsd working and it has gnarley nice SMP
but I noticed its on the harmful list!
DOH!
my db admin pal is asking how do web apps without a relational db?
do things get overwritten?
or?
it's on harmfulls? must be because of the PAM
I don't see it there
I see it, but there is no explanation
why is it there?
hahah head(1) has also made it to the list
I'm surprised tail hasn't, considering you can probably emulate it with an ed script
http://mail-index.netbsd.org/tech-userlevel/2009/10/24/msg002827.html
well, they've proposed nonsense ideas before and they never got on anywhere
* EthanG gets to the end of the 2nd sentence & automatically checks the date on the mail
netbsd also has an xml parser in the fucking kernel
yappa
or some other similar expression
mabe I should /part #netbsd if it sucks that much now eh
zcram: netbsd is harmful for many reasons, PAM is just one of them, 'magic symlinks on crack' is another, and there are many more
never heard of the magic symlink problem
22:12 < EthanG> I'm surprised tail hasn't, considering you can probably emulate it with an ed script
EthanG: way to totally miss the point
xml lol
could those be outlined on the table there? I don't think many other people will know this either
zcram: the table is supposed to make people think, but it seems to be failing miserably
uriel: but the point of head being considered harmful was you can emulate it with sed 10q
lol
because apparently everyone needs every idea and point of view completely pre-masticated and-regurgitated before they can swallow it
mmhmm
EthanG: wrong, that is only *part* of the point
oh
what about smp on netbsd and memory supprt in amd64 vs openbsd limitations
uriel: just a slight hint for netbsd... could make it a link to a small mockup page or whatever
magic symlinks in NetBSD are not default, you need to activate them with a sysctl
well, NetBSD certainly deserves its own page, but I got better things to do than to list everything that sucks on every thing in the world
I think they only implemented them for per user /tmp
I can clearly figure out what is wrong with head as opposed to the other approach but it's kind of difficult to assess the circumstances with reasonable presicion concerning something as big and universal as nbsd
DaCa: but what do they do exactly?
DaCa: uhu? they implemented a totally completely insanely fucked up retarded 'feature' to have 'per user /tmp'? are you kidding?
uriel, I like to think at least as much as the next guy, but head is so meaningless ot me that I don't care to wonder why it is harmful.
they take something fucked up as symlinks, and make it infinitely more fucked up
that is quite an accomplishment
erm, open was at one point almost having a per-user visible process tree
aka split init or whatever
to*
SmoothPorcupine: exactly, head(1) is completely meaningless and superflous, which is *the reason* why it is harmful
zcram: http://netbsd.gw.com/cgi-bin/man-cgi?symlink+7+NetBSD-5.0
Wait... I didn't say superfluous. Get out of my head!
But doesn't the same then apply to tail?
SmoothPorcupine: no it doesn't
aww that magic link stuff could be rather nifty actuly
note how Unix had tail for many years, but it never had head(1)
(until the fucking retards at Berkely added it, because fucking retarded students high on LSD have nothing better to do than to add totally retarded and useless features)
Marginally useful. :P
but it sucks obivously and there are alternative and perhaps better ways for doing it (still on about the symlinks)
zcram: no, it is totall, pure, and complete braindamage
symlinks are pure evil shit, magic symlinks take it to a completely different dimension of evil
well, tail... tail is good for tail -f!
the proper solution to the 'problem' is private namespaces
(and getting rid of suid and root)
yes, but it's impossible with adhereing to the holy stds they have now
16:30 < uriel> zcram: the table is supposed to make people think, but it seems to be failing miserably
STDs
no device designed to make people think has ever succeeded
rubic's cube?
cf rubiks cubes, math homework, valentine's day, etc
zcram: I don't buy it, APE implements posix
PoSix sucks, but it can be implemented on top of minimally sane private namespace/non-root semantics
linux has done private namespaces, and getting rid of suid has supposedly been a goal of linux and openbsd for many years now (of course they have not made much progress because all the retarded user space out there)
math lessons didn't teach me to think. not directly, and not well.
what about SuS
isn't SuS basically PoSix? this days it is all one huge ball of hair with various names AFAIK
it's a tad more afaik
but I dont follow standards, and really few if any people actually care about them
a superset of parameters, if you will
because they are irrelevant, it is all the crap code that depends on all kinds of crap non-features that is the problem, not the standards
which again, are irrelevant and often ignored by everyone
eh, crap stds, crap code
that's my idea
not the other way around
look at mycroft shop or the fruit shop
sure, there is a degree of mutual reinforcement between the two, but as much as I hate stds, the real blame belongs to crap code and people that continues to write and use crap code
people that justify their retarded code by saying they are following a standard would write just as evil code without the standard
if you adhere to a crappy std no good code is ever going to come out of it
and let's subst. 'crappy' with 'limiting' now
i disagree, look at how for example p9p uses pthreads
Limiting how?
(or ape implements PoSix)
Because I'm beginning to think the biggest problems we face in CS is systems that are too flexible.
hahahaha
HTTP, HTML, to name the biggest ones.
"With HyperText, you can do anything!"
you think html's biggest problem is that it's 'too flexible'?
I couldn't say anything with UNIX code history woudl be overy flexible
the frames are somewhat stiff
Yes. You can do almost anything in HT technologies.
That's why the web is so prolific.
html sucks. weren't we on about this last week?
I don't know if too flexible is quite the right term
SmoothPorcupine: I can't even process that line of 'logic'
also, I like rounded corners on shit. Not.
if that were the case then someone would have ported the linux kernel to perl by now
*snrot* yeah
cgi/perl?
Would people have explosively used and developed HTML if it wasn't flexible?
html *wasn't* flexible when it started exploding
it wasn't very flexible at first
it wasn't very much in the beginning, but now the bigCorps like to make it more shiney and so forth to sell you their crap
it was *made* flexible because it was *easy* and then there were thousands of assholes who couldn't do anything but html and javascript
afaik it took Netscape charged on ahead adding images to start things rolling
so lots of effort went into enabling stupid assholes to inappropriate things without having to exapnd their skills very much
I could easily argue the browser and format were incredibly flexible for that time.
and that evolution led to horrible shit like php
was*
Programmers need strict limits, or they do stupid stuff.
Take GUI for instance. The concept is much too open for tanybody to make a proper GUI anything.
y should see some PDP C code
and reconcider your statement
well, maybe not
anybody*
:)
ok
but openbsd seems less harmful because it is simpler, yet it supposedly has crappy SMP and wont suppor over 4g ram
hmm
I am enjoying ed though
it makes me edit using substitutions rather than moving my cursor all over
kinda cool
also the indictment of sql databases got me quie an earful from my dba at lunch
www.paulgraham.com mentioned success without a sql db
kinda cool
not sure how that would work or if its easy
file locks?
OBSD is a little strange :)
happy: what did you expect when you suggest he should be without a job
one more reason why deference to 'experts' is stupid
of course experts in anything will tell you that their field is essential and not-full-of-shit
who wants to admit that their job is useless and that they wasted their lives learning totally worthless idiotic shit?
I really needed to read Kernighan's paper, 'Program Design in the UNIX Environment' (http://harmful.cat-v.org/cat-v/unix_prog_design.pdf) in full. I'll be reading that a few times more, I think
reading the book is also a must
ahh
this is a great joke
http://research.swtch.com/2010/03/formal-logic-club.html
:}
lol it gets funnier the more I think about it
I can't seem to find it funny.
:)
Im sure its logically funny
SmoothPorcupine: hold your mouse over James Grimmelman's name & read the tooltip
so uriel I should dump netbsd in lieu of openbsd?
due to less harm
hmm
and fuck learning postgresql
I'm not sure that helps the joke but it's kinda mindblowing
Yes I read that.
happy: it all depends on what you are trying to do
I stopped caring about anything relating to the foundatiion crisis when I read it was related to something that was totally false.
oh?
(still, I have trouble thinking of any task for which NetBSD would be a good tool, other than teaching how to fuck up an OS)
I am trying to learn programming, make my own business, make custoemrs happy, get rich, and have sex with hot women
roughly
ok, then the answer is: switch to another field
netbsd desktop with pidgin, iceWM, and firefox now
aw shit
I believe you misunderstand. I totally stopped caring. I wont even bother to show why I am correct.
and if you use icewm willingly, you need to have your head examined
Until I prove it, that is.
meh, bed-time for me
I love iceWM
I hate tiling wm
I love to split window top and bottom and always forget in dwm the keys to do it esp when I have 8 windows open
icewm drag with mouse
bam done
:)
simplicity!!!
yeah the mouse is a great tool many WMs ignore
EthanG: because of retards that are incapable to fucking learn to use a fucking mouse
really sad
in any case, pointing out how retarded some other wm is is *NOT* an argument
uriel: yeah! I have motor control problems that I'm starting to think are pretty serious, and I'm willing to work around them for mouse-based interfaces
uriel you wrote dwm eh?
uhu?
I like ion3 for tiling, but I'm not sure it's not dumb under the surface. It's certainly not too easy to hack on.
the source for ion is abominable
full of really hideous macro hacks
ah, well now I feel better about having failed to get to gips with it ^^;;
anyway, sleep for me. night all.
s/gips/grips/
i like fluxbox, it works for me
i dont have a large enough screen to make tiling worthwhile
good night EthanG
* uriel goes to sleep too
night
tiling can be nice, but it is not the be-all-end-all that some make it to be
the rio+acme combination is powerful precisely because you are not stuck with only tiling
tagging is quite useful
(larswm provides a similar environment, and to some degree wmii does too)
'tagging'?
There is no be-all-end-all in UI.
uriel, its like virtual desktops, but you tag windows with identifiers
* uriel has never quite understood dwm, the defaults are were too broken every time that I tried it
and you can say, show a b c
or just b
or whatever
martian67: hah, funny, I think that was my idea ;P
tagging as in grouping windows
to be cycled through
uriel, dont blame yourself, hard to be orginal :p
martian67: I mean, that I suggested it to garbeam ages ago for wmii I think
I'm still not convinced it is a good idea really
it could be, if you find it useful
the whole tags fad on the web seems to have thankfully passed for the most part
i mean what other meaningful UI metric is there
what is tagging on the web?
its not like its any kind of scientific process that goes into UI design
Pretty.
martian67: there can be, there have been studies regarding how long it takes to do certain tasks, etc, but it is hard to construct such studies
also that may or may not apply to you
somebody that is used to a UI and can use it efficently
as opposed to someone who was introduced to it yesterday
somebody that is used to anything can use it in any kind of way
even php and java can be used 'efficiently' that doesn't mean they are any good
Why do you even need experimentation? It's pure logic.
SmoothPorcupine, ok there plato
plato can go fuck himself
pure logic is never enough when it comes to demonstrating a meaningful point in the real world
Don't mind me not wanting to waste my time seeing if it is actually faster to press two keys than it is to press one key.
although it can be useful, its not rigorous
Maybe those attempts to demonstrate were by terrible logicians.
SmoothPorcupine, experiment is a fundamental pillar in the scientific method
Actually, I think I'm going to have to move that up to a certainty for lack of anyone being decent at logic.
if you cannot show your theory in the real world, its not useful
scientificly that is
I'll let scientists figure out which idea that is already out there works best.
SmoothPorcupine, and since when does UI rely on any kind of "logic" besides maybee the logic of what you expect
which is not logic, its just expectation
Maybe they'll come up with something that wasn't logical at first glance.
GUI designs seem to be basicly based on an extended notion of cards or papers
I don't mean to say that any current UI has been designed logically.
:)
i think arguing about how logical a UI is almost entirely useless, if its useful for you thats what counts and matter
s
humans are not really logical
uriel, also, [18:10]
* EthanG gets to the end of the 2nd sentence & automatically checks the date on the mail
* uriel goes to sleep too
* uriel has never quite understood dwm, the defaults are were too broken every time that I tried it