The Goldfish Chronicles

“[7] The way to make programs easy to read is not to stuff them with comments. I would take Abelson and Sussman’s quote a step further. Programming languages should be designed to express algorithms, and only incidentally to tell computers how to execute them. A good programming language ought to better for explaining software than English. You should only need comments when there is some kind of kludge you need to warn readers about, just as on a road there are only arrows on parts with unexpectedly sharp curves. “

Hallelujah. We’ve all known deep down that this is true.

Various people talking about Paul Graham’s latest article http://www.paulgraham.com/hp.html. His basic point being that painters and hackers are akin. I’m utterly unconvinced, but possibly because I don’t understand his definition of a painter. If he just means the renaissance style painting shop, then maybe. It was a point at which art was at its most craftsmenlike, before it became pretentious nonsense.

I do completely agree with him that ‘computer science’ and ’software engineering’ are utterly wrong as terms. I am not an engineer in any way.

The way hackers tackle a piece of code, by sitting and playing and tweaking until they coomprehend it utterly, is exactly the way I believe mathematicians deal with an item of mathematics. It is full of intuition and ‘leaps of faith’ to reuse a term that does not fit with mathematics. This could possibly just be how I handle mathematics, but it fits other mathematicians I’ve computed with.

Computer scientists on the other hand are often far less ‘hacky’. University has imbued them with beliefs that they are engineers and should have lots of black and white answers to questions. Odd that, as mathematics is the only study in which black and white answers can exist, and even then only in certain areas.

I do like how Paul Graham talks about the day job choice. It definitely fits with the way I separate open-source [ie Me] code and day-job [ie Work] code. My latest job is nice in that I can have some cross-over, which is nice, though I’m glad I’m no longer paged at 3am.

Generally my problem with his article is that mathematicians are not scientists. They are makers. All Paul Graham is really noticing is that maths is an art, not a science, and that programmers are a novel new brand of mathematician.

We are mathematical craftsmen.

With my minor differences from the article in my defence of maths, I agree with lots. Especially what I call code responsibility. Code ownership is bad, but every code should be under a single person’s responsibility, even if 8 people hack on it from time to time.

I believe I suffer from the same problem as Paul Graham, I believe in my own education, but I have long thought that mathematicians make the best ‘Pragmatic Programmers’.

flamefew.net’s priamry server ‘umbongo’ has been migrated from umbongo-snr to umbongo-jnr. Mainly to save me renting space S’been a slow moving project over the last month and while it all generally ended up working in the end I managed to screw a few things up.

SuSE were overriding the ip address for postfix, so for an hour I only had SMTP on localhost.
I forgot to migrate the http access logs, and I forgot that ikonboard uses local files and not mysql so the data was out of date.

The only bad part, no spam-assassin at the moment. So I’m back to the world of spam

Still, job done. Pretty much. Time to move on to other things

I did a tiny bit of happy code yesterday for Commons Lang, and I brought two
oreilly books [OS X Hacks and Linux Hacks]. I almost feel as though I’ve rediscovered the love, or if not rediscovered, received a ‘wish you were here’ card.

Amusing when DNS doesn’t point to where it used to. This should have pointed to the party for Cornish independence, instead it points to a Scottish university:

http://www.kernow.eu.org/

Well, I started to work on something [http://dev.osjava.org] but this site has long since beaten me to it: http://javaindex.org/

Looks good.

Because it’s so shameful, I have to share that I installed Windows 3.11 this weekend Grabbed the IBM ADK, a 90 day version of JDK 1.0.1, then realised that IE3 for WfW 3.11 comes with a 1.0.1 JDK. Hadn’t realised this was the case.

Just realised that you can’t do this on consumer routers. I’d assumed you can.

You can redirect an IP, so port 25 on my router can be redirected to port 25 on some machine inside my network, but I cannot redirect port 80 on my router to port 8080 on a machine inside my network.

This is with a netgear wireless router, however it has seemed to be a downgrade in quality over the older netgear wireless router I owned, so maybe the older one had this capability. The only improvement the new space-age router seems to have is a slightly stronger signal.

Well, one month of spam-assasin fighting, and 1300 emails are automatically dealt with. Taking the 10 minutes to go through them now, I find only 2 false-positives, both from my new install of JIRA on osjava.org. I’ll have to white-list that. They both scored exactly 5.00, the amount needed to be classed as spam. So if I delete 8.0 and above, it should be pretty safe.

So not a bad first month. 1300 emails I never had to deal with.