Computers suck

[pmb]

Computers suck, and that's my contention. I can prove it on scratch paper and a pen. Give me a fucking Etch-A-Sketch, I'll do it in three minutes--the proof, the fact, I'll show my work, case closed.

--Bill Hicks (kinda')

Grr. So I graduated from a prestigious undergraduate university with a degree in Computer Science and Mathematics, I worked as a network software developer for two years at company filled with really damn smart people, and I'm now a graduate student in my chosen discipline of, you guessed it, computing.

And yet Tracy's OS X iBook still refuses to respect read-only permissions when she mounts her Linux home directory. Which causes RCS to barf. Which means that Emacs source control won't work. Which means that she can't use emacs to edit files remotely. Which sucks. And I can't fucking fix it. I've read man pages, searched the internet, downloaded source code for the "latest and greatest", recompiled, and all sorts of other things. ARGH. The whole point of learning about this and doing it for so long is so that I'll never have to have stupid errors like this eat away at me.

In less geeky news, however, we signed up for a CSA today! CSA (Community Supported Agriculture) gives us good food for cheap every week. We get to play Iron Chef with funky vegetables!

"This week's ingredient is ... DAIKON RADISH!!!!"

"What's that?"

"I don't know either"

"here goes nothing!"

Also, I have rediscovered my love of raytracing, which I have not felt since my junior year of hich school. Especially as typified by the IRTC and these two images. Right now my computer is using POV-RAY to slowly draw the first one at a large enough resolution for it to serve as my background.

So now I've displayed both the high and low points of my nerdy, nerdy day.

Comments

[flamingweasel.livejournal.com]

What I've found is that my (admittedly less complete and less impressive) stint as a Computer Science major has made stupid little errors like this, where it's (probably) the fault of some (perhaps) lazy (or maybe drunk or perhaps even dumb) developer somewhere along the line, so much more annoying. "Normal"* users would look at a problem like that, say, "gee, that's kinda dumb," then ssh into the linux machine and edit the files with whatever editor was installed on that machine. Whereas I look at it and scream and bang my head against it trying to fix it.

It's exactly this that makes my job extra annoying. We have to use software that is just chock full of kludges and sheer lazy programming. Most of my coworkers say it sucks, and it does, but I think I'm one of the only people there that knows exactly how to make the damn thing work better. But I'm only an input drone, not a developer, so...

*: In quotes because I have met very few "normal" users who'd do something like use emacs to edit remote files.

[pmb.livejournal.com]

Perhaps that is true. But more likely it is an essentially emergent property of the system, created by Apple's engineers making something new and shiny with new protocols and free software lagging a little bit, and both teams having a bug here and there. I'm not too upset that it happened. I'm upset that I wasn't able to find the problem when I did an (admittedly cursory) look through the netatalk source code to fix the problem.

When I needed something different (i.e. a bug fixed) from gnomecal, it wasn't a problem - I just edited the source (and became a contributor to the Gnome project - Woo!). Similarly for a buncha' other things. Besides the whole ethics of free software, which I heartily endorse, there's the fact that having the source is really convenient when you've been a coder - especially when you've had to decipher CS122 code.

So I'm not pissed that the permissions thing happened. I know quite well that interoperability is a pain in the butt, and completely implementing an RFC or whatever is almost completely impossible without more investment than almost anyone is willing to put in, at least initially. I'm pissed that my code-fu was too weak to find the problem and fix it.

But I am of the opinion that Microsoft and Netscape set software development back a decade by teaching users that buggy code is something to tolerate, rather than something to complain vociferously about. And I completely feel for you in about the hack-disgruntlement thing. I've used systems like that (all of them non-free) and using them pained my inner (or not-so-inner) nerd. I always found excuses to either use them as little as possible or to make do with other tools that, while not designed for the task at hand, actually worked as specified.

The solution I'm implementing is "Teach Tracy to use CVS". Which falls squarely in the "use another tool" camp. If you use Emacs for editing anything, then you should know about this. Because it will save your ass, and it works with RCS and CVS - which is the main reason that my alternate plan might succeed.

Anyhow, enough ranting, and I hope that you'll be able to complain to the right people loudly enough to get them to make a different tool. Particularly, when you complain you should point out that if a better tool allows for a 10% increase in efficiency, then they've saved yoursalary*numberofpeoplelikeyou*.1 dollars, which is potentially a very large number.

[chocolatesmudge.livejournal.com]

Geektacular!

p.s. I have a nose ring. wheee!!!

[cabanasloth.livejournal.com]

brother, I feel your pain.

I assume this is an NFS mount? I've been dealing with NFS issues quite a bit lately, and so there's a remote chance I could help you out, if you want to talk about it over IM or something. Are you convinced this is indeed an implementation bug, and not some odd feature of NFS? (Specifically, is Tracy's UID identical on the two boxes?)

Raytracing is very cool, though I've been more of an OpenGL kinda guy lately myself. And radishes frighten me.

Thank you for being the other guy out there who makes livejournal posts about how much computers suck...

[pmb.livejournal.com]

Thanks for the expression of solidarity.

It's not NFS, it's AFP - essentially the new appletalk, which now has the ability to be tunneled over IP instead of just Apple's funky "ethernet address only" scheme.

And raytracing rocks openGL like a diesel powered rocking machine. At least for still images. Also, I have lots of good POVRay memories that are all flooding back.

And if you think radishes are frightening, imagine getting a radish the size and shape of a cucumber, and that's a daikon. It was terrifying. We had no idea what to do.

[cabanasloth.livejournal.com]

It's not NFS, it's AFP - essentially the new appletalk, which now has the ability to be tunneled over IP instead of just Apple's funky "ethernet address only" scheme.

oh dear lord *why*? did the world really need a new appletalk?

And raytracing rocks openGL like a diesel powered rocking machine. At least for still images.

The "at least for still images" part is key. You let me know when that one image is finished rendering, okay? I'll deal with my 30fps of texturemapped glory. Who needs reflections, anyway?

[pmb.livejournal.com]

You let me know when that one image is finished rendering, okay?

It's more than half done now!

[goteam.livejournal.com]

And if you think radishes are frightening, imagine getting a radish the size and shape of a cucumber, and that's a daikon. It was terrifying. We had no idea what to do.

You forgot to mention that the daikon are big and white, so they look kind of like something out of the Bunnicula books. And what's really scary is that the ones we got were relatively little; I've seen daikon at the supermarket that were as big as my forearm (and often thicker!) Eeek!

[kweerious.livejournal.com]

there was much discussion about this and it was remedied somewhat with 10.2.3. essentially, the umask needs to be set earlier on in the boot process in order for anything else but a terminal session to respect it.

check out lists.apple.com and search the archives (i'd reccommend the mac-osx-server list, it covers it in detail...iirc it was back in fall/winter)

[pmb.livejournal.com]

Thank you dd! That was exactly the pointer I needed! Now, on to fixing the problem...

it's all about the radish

[kuddliphish.livejournal.com]

so what did you do with the radish? Did you like it? Stony has been tempted to pick up a daikon radish at the grocery store and try it out. He has this fascination with strange ugly vegetables, particularly tubers and roots (you haven't eaten an ugly vegetable until you've eaten celery root) I think he might also have a couple of recipes that call for daikon radish that he is interested in testing out.

Re: it's all about the radish

[pmb.livejournal.com]

Well, first we made a coleslaw-like-thing with lots of vineagar and sugar and grated daikon and carrot and cayenne pepper. Sounds weird. Tasted good. The Moosewood New Classics Cookbook contained instructions, and we followed them.

The next time we got one (next time??!?? yes, we got it in our weekly distribution a few times) we ended up grating it onto salads, a little at a time. We might also have roasted it like a turnip, but that could be me just making things up, too. You can also slice it into matchsticks and put it on stirfry.

Daikon radishes turn out to be quite tasty. Like a cross between a jicama (spelling?) and a radish. I'll have to try the celery root thing...

Re: it's all about the radish

[farmchick.livejournal.com]

Daikon were one of my least favorite vegetables when I was in Japan, but that was probably due to eating too many of them. The only way I really liked them was grated finely into a mushy paste with a little soy sauce dripped on top. They you just ate a little bit as a side or on top of some fried tofu.

So daikons are big and white and sometimes fuzzy, but for strange alien shaped vegetables I don't think you can beat kohlrabi. When the brasicas start coming in your share (broccoli, etc), look for a strange green bulb with knobby radial projections. It has a great taste like broccoli stems, but it looks really weird.