computers suck.

"Note that this will work incorrectly if there are any filenames containing newlines or spaces."

i love how search sucks so hard in so many mail systems. gmail doesn't do stemming. what the heck is up with that? how is that even remotely acceptable? thunderbird finds results for "foo bar" (w/out the quotes typed in as part of the search) but then gives no results for "foo bar " (extra space at the end, still no quotes). this is the future? this is productivity boosting? this is quality? this is allowed? this goes un fixed? nobody notices? nobody cares?

goddammit, i am pretty sick and tired of not having dependent typing.

dammit! i need time to learn Typed Scheme, already.

nice how the finder fails to show a newly created (from unzipping a file) directory. and there's no refresh command like in windows. even after browsing other directories and coming back, to see if that would wake it up. even though it shows up in terminal. deleting the zip file of the same name finally let the directory show up!?

this is better than windows?

i love how mac os x (10.5) fails to eject my flash-stick-drive. here i thought only windows was an utter piece of excrement in that fashion, but nooooo.

'"All architecture is design but not all design is architecture. Architecture represents the significant design decisions that shape a system, where significant is measured by cost of change." What follows from this is that an effective architecture is one that generally reduces the significance of design decisions. An ineffective architecture will amplify significance.'

> One of the things I now focus on is "dead ends". We can easily chose a path if the cost of changing to the other path is low but we would try to avoid a path where the cost of change is high. Rather than focus on getting the "Right" path ( which is impossible without a crystal ball v3.75 ), I focus on the cost of changing if we get the wrong choice.

> FYI. This is what Dr. Gregory House in "House M.D." does pretty much every episode.

wow!

well.

i. hate. slide. share.

so, like, things can't get in through the shields, but they can shoot phasers and torpedoes out through them? i guess presumably they briefly blip the shields off-then-on.

"what did we do wrong?"

uh, yeah, well, maybe you used a shitty language. if you know what i mean.

so, yeah, in some sense it seems that alan kay is right.

maybe it is just me. i guess it is just me, otherwise why would so many things be done the way they are, the way that i hate? to wit: i hate sudden jumps in ui, like the way document (word, reader) viewers sometimes do depending on their mode; like the way google calendar does from month to month. i prefer things to be continuous and to not have a sudden jump in the view. it drives me bonkers.

[blogger: oh, sorry, we utterly failed to post that. i mean, it isn't as if we're run by an org with more computing power than god. and it isn't as if the whole bloody point of a blogging system is to let bloggers post. well, i guess some are more equal than others, or something.]

[blogger: failed again!]

programming sucks. programming languages suck. i mean, if it is so important to have a virtual destructor in certain cases, if that is such a common bug, then why the heck is it even possible to get into that situation without big alarms set off by the toolchain?! (or any other "top bugs in language X".)

it is amazing how java succeeded despite the best attempts to torpedo it by sun itself.

by jove, spot on, eh wot!?

"Who is ParkingCarma? We are a high-tech service company that makes your parking experience more convenient."

no, actually, ParkingCarma is a crappy web site that can't even successfully charge things to a visa card. "high-tech"?! "service"?! my tax dollars at work?! augh!

i hate that the jvm doesn't have tail call optimization. having to use loop/recur in Clojure is really making me cry. i mean, here i am getting a chance to mostly think in nice functional ways, and then slap wham bam a lama ding dong, i have to switch mental gears to be able to insert some looping crap.

h8.

ui is hard. ui toolkits/frameworks are hard. they are hard to create w/out turning out poo. they are hard to use. i think a lot of the problem is that ui is usually a pretty stateful thing, that the state space is large. so even knowing what is required to e.g. get a widget to be able to receive key-press events can be a little daunting. how sad is that? one thing i can think of to help reduce the suck would be some sanity checker utility functions a developer could call which would tell them if something could do X (like "could get a keypress event") and if not then what is causing that to not work (like "has not had the setFocusable state set to true"). maybe.

Basically, I'm just wondering what one would use if one wanted to make concurrent commercial software that sucked less wrt correctness. (One issue is that everybody has a different sense of what is good enough.)

What are the current best model checking systems that are industry-ready? In particular, ones which check for concurrency issues (races, deadlock, livelock) with distributedness a bonus.

It just seems like whenever I go trawling google I mostly find research projects, or apparently dead, or expensive commercial checkers. (McErlang, Spin, Opis, JPF seem to be the most relevant thing I've found; FDR2 appears to be deadish and/or expensive.)

Alternatively, are there ready-for-industry languages which obviate some such needs? Cf. deterministic concurrency (Oz, FRP, discretely deterministic); skeletons for parallelism.