Comments for [insert title here]

Comment on College, part 1 by Kevin

Kevin — Mon, 30 Aug 2010 04:53:56 +0000

In addition to being interested in hearing how you ended up at MIT after Harvard, I’m also somewhat curious as to what effect MIT CPW had on you.

Comment on College, part 1 by Piper

Piper — Mon, 23 Aug 2010 06:28:59 +0000

== both of them

Comment on College, part 1 by Leonid

Leonid — Mon, 23 Aug 2010 02:23:15 +0000

==Josh. You should go on.

Comment on College, part 1 by Josh

Josh — Mon, 23 Aug 2010 01:39:12 +0000

This was very interesting. You should write more!

Comment on Why is software so poorly written? by kasittig

kasittig — Wed, 28 Jul 2010 06:31:08 +0000

“In my experience, code produced by companies does not tend to be especially awesome.”

true – also I’ve found that documentation tends to be awful, mostly because it’s heavily internal and it’s often easier to just add contact people than to actually document things. the fact that, in the real world, documentation is often an afterthought (and is usually in design docs and specifications rather than in actual code) was mindblowing.

“[...] the fact that it’s easy to ship new versions of the product if you mess it up the first time, as well as the fact that failures really don’t cause too much damage.”

I don’t think this is necessarily true – I think it’s a product of a ship cycle. the goal of shipping software appears to be to deliver on a set of features that your customers work – and if the code is behind the scenes, who cares if it’s messy? in school, we’re graded not only on our end product but on the style of our code, and this is the core difference – your code being sloppy in the Real World only seems to matter to the poor human who has to deal with it after you’re done writing it.

Comment on Source diving by Andrew Farrell

Andrew Farrell — Tue, 27 Jul 2010 14:53:51 +0000

I’ve recently noticed this too. This may possibly be because I’ve gotten fed up with looking through bunches of forum posts, my eyes and attention flitting from one pge to the next.
Source diving, I can just pop into pdb and follow the code.
Also, the code must be correct. The documentation could be wrong.

Frivolously, it feels so much more hardcore

Comment on Why is software so poorly written? by Andrew Farrell

Andrew Farrell — Tue, 27 Jul 2010 14:47:59 +0000

It may be that its alot harder to build up an intuitive sense for when something is built wrong in software.
As a result, this intuitive sense is alot weaker, and so it is easier to skip over something that feels a bit wrong out of laziness.

Comment on If argparse is the future, I want to build a time machine so I don’t have to deal with it by Andrew Farrell

Andrew Farrell — Tue, 27 Jul 2010 14:44:06 +0000

You forgot to mention:
The reason it does this is to support having options that can take one or no arguments.
Its not clear that you would want to maintain this property.

Comment on Why is software so poorly written? by joe

joe — Tue, 13 Jul 2010 04:40:14 +0000

(Clearly what we need is just more dependent types. I don’t see what’s so hard about it.)

I sympathize, though. I’ve recently had some fun times wrestling with bcfg2, e.g., and it’s far from the worst software I’ve ever dived into…

Comment on You know you’re no longer a Harvard student when… by ismith

ismith — Tue, 01 Jun 2010 01:05:14 +0000

Are you sure they don’t just disallow users coming from 18.*.*.*? You know you can’t trust anything from that IP range.