|
Petulant Fig Nemesis (Mark 11:12-14)
@
mwotton
Ann Arbor, MI
|
|
lambda mechanic for @simspacecorp, occasional goblin technologist
|
|
|
38.846
Tweetovi
|
1.439
Pratim
|
2.532
Osobe koje vas prate
|
| Tweetovi |
|
Petulant Fig Nemesis (Mark 11:12-14)
@mwotton
|
3 h |
|
oh man, you've been missing out - it's an amazing track.
|
||
|
|
||
|
Petulant Fig Nemesis (Mark 11:12-14)
@mwotton
|
4 h |
|
Would never argue against actually learning a language, but immersion seems to work radically better than anything else
|
||
|
|
||
|
Petulant Fig Nemesis (Mark 11:12-14)
@mwotton
|
4 h |
|
Original statement is explicitly about what lambdaschool grads are missing by not having language education in the style of current universities, so I don't think that it is a separate question at all.
|
||
|
|
||
|
Petulant Fig Nemesis (Mark 11:12-14)
@mwotton
|
4 h |
|
I'd be more swayed by this line of argument if there were evidence that a reasonable chunk of Americans who pass language courses in school came out with anything even close to fluency. Hard to credit other claims when we can't detect a difference when we have decent metrics.
|
||
|
|
||
|
Petulant Fig Nemesis (Mark 11:12-14)
@mwotton
|
7 h |
|
I think a secure voting system is legitimately not that hard a problem, once you solve key distribution. But
1. That's a pretty big if.
2. Still going to require a fair chunk of education to teach people how to verify the transaction - would be a very juicy target.
|
||
|
|
||
|
Petulant Fig Nemesis (Mark 11:12-14)
@mwotton
|
9 h |
|
not that you don't know this already, just thinking aloud how you'd convey the difference to a layman.
|
||
|
|
||
|
Petulant Fig Nemesis (Mark 11:12-14)
@mwotton
|
9 h |
|
i think the really tricky bit is the proof of correctness - when you're beating the go master, the proof's in the performance. Proving voting schemes correct & verifiable starts getting into opaque cryptography pretty quickly.
|
||
|
|
||
|
Petulant Fig Nemesis (Mark 11:12-14)
@mwotton
|
9 h |
|
i accidentally opened two youtube windows a fraction of a second apart and it took me until 1:30 to notice. only slightly less disorienting with one.
|
||
|
|
||
|
Petulant Fig Nemesis (Mark 11:12-14)
@mwotton
|
14 h |
|
I think there's a real danger that nerds like us will retreat to these tools smugly, while ordinary users without technical chops are left high and dry.
|
||
|
|
||
|
Petulant Fig Nemesis (Mark 11:12-14)
@mwotton
|
15 h |
|
to be fair, this isn't something the tech team had any control over at all. Twitter would crash too if Jack cruised around the streets of SF shouting root passwords through a megaphone.
|
||
|
|
||
|
Petulant Fig Nemesis (Mark 11:12-14)
@mwotton
|
15 h |
|
lol. the security threats are always so much more mundane than we assume. Even XKCD was too optimistic:
xkcd.com/538/
|
||
|
|
||
|
Petulant Fig Nemesis (Mark 11:12-14)
@mwotton
|
16 h |
|
Is it clear that it's actually a load issue? Absent further info, I would have assumed "it works on my machine" style deployment screwups. (Agreed that you could write something that handled that load as a shell script in cgi-bin without difficulty)
|
||
|
|
||
|
Petulant Fig Nemesis (Mark 11:12-14)
@mwotton
|
16 h |
|
yeah, i wouldn't think it usually is (unless you're doing significant work in the TH itself - you do need to load up all that machinery, but would assume it's a one-time cost per compile?)
|
||
|
|
||
|
Petulant Fig Nemesis (Mark 11:12-14)
@mwotton
|
16 h |
|
it is but i don't get the impression that parsing is superlinear - it's the later passes that try to inline and optimise etc? I could be wrong, haven't rigorously benched it.
|
||
|
|
||
|
Petulant Fig Nemesis (Mark 11:12-14)
@mwotton
|
16 h |
|
plus the superlinear compilation time with GHC as haskell files get bigger.
|
||
|
|
||
|
Petulant Fig Nemesis (Mark 11:12-14)
@mwotton
|
16 h |
|
we sort of expect our massive hand-hewn files to be slow and make allowances, TH just strikes us as an injustice because it _feels_ compact.
|
||
|
|
||
|
Petulant Fig Nemesis (Mark 11:12-14)
@mwotton
|
16 h |
|
my understanding is mostly that it makes it possible to create a lot of code, both types and functions, so sometimes you end up with way more to compile than you'd think from looking at the source file.
|
||
|
|
||
|
Petulant Fig Nemesis (Mark 11:12-14)
@mwotton
|
16 h |
|
it's certainly easy to lose all the benefits of a good programming language by doing other things poorly. there's a maslow's hierarchy of sorts.
|
||
|
|
||
|
Petulant Fig Nemesis (Mark 11:12-14)
@mwotton
|
16 h |
|
every TH models file pulsing a dark satanic purple
|
||
|
|
||
|
Petulant Fig Nemesis (Mark 11:12-14)
@mwotton
|
4. velj |
|
At the very least, the reaction to noticing the gap between your estimate and the actions taken by an experienced practitioner should be to recheck your numbers.
|
||
|
|
||