Sunday, May 26, 2002
Dang. Matthew Thomas has beaten me to it, and posted a blog saying exactly what I was going to say about Mozillazine’s comments on CNET’s review of Netscape’s newly pre-released browser. He even links to the same article by Joel God-comes-to-me-for-advice-on-software Spolsky that I was going to link to.
So I will have to either shut up (fat chance), or think of something else to say on the subject. How about this: if the browser crashed whenever it loaded a web-page containing the word “goat”, and a few reviewers suggested that this was less than ideal, would you still shrug it off by saying “It's in the release notes”?
Tuesday, May 14, 2002
Tonight we experienced our first California earthquake. It was magnitude 5.2 on the Richter scale, which is classified as “moderate”; the amount of energy released is equivalent to a 32 kiloton explosion, which seems like a lot to me, but is only about one thousandth as powerful as the Loma Prieta earthquake of 1989.
It's interesting how one reacts to something totally unexpected like an earthquake. At first I heard the cups rattling in the cupboards, and then noticed that the house was shaking. At this stage my mind was still working hard to fit what was happening into some category that it was used to, like “Ahinoam is knocking over the glasses” (she was getting us both a drink of water at the time) or “Aviad is jumping around in his room”. Then, as the shaking gets stronger and the rattling gets louder, there is a sudden leap to a new interpretation: “It’s an earthquake! Everybody get outside!” (This turns out not to be the right thing to do.) In this case, by the time we started moving outside everything was over, and I was left to deal with the after-effects of the adrenalin rush and calm down the children, who were dealing with the same effects, very nervous and full of questions.
My usual policy in such cases is that it’s more important to be reassuring than strictly accurate:
Is there going to be another earthquake?
No, there’s no more likely to be an earthquake now than any other time. A big lie, because I know all about aftershocks.
Aren’t seismographs a dreadful waste of paper?
They were in the old days, but nowadays all the data is stored on computer. A shot in the dark, but apparently accurate.
Friday, May 10, 2002
I’m sorry that I returned Antoine de Saint-Exupery’sThe Wisdom of the Sands to the library without having read much of it. I found it difficult to get into, I don’t know whether because of something in the book itself or because of defects in the translation.
Chiefly though, it isn’t really a book to take out of the library, read through and return after three weeks. I would like to have had it around for a few months and dip into it from time to time when feeling thoughtful, not race through it on the bus to and from work where I do most of my reading, which is fine for unwinding with a thriller after a long day’s hard concentration, but an injustice to a book full of ideas like this one.
I took the book out in the first place because Matthew Thomas pointed me towards it on IRC as the source of something I have been misquoting for years:
La perfection est atteinte non quand il ne reste rien � ajouter, mais quand il ne reste rien � enlever. [No translation will be provided this time, chiefly because every translation I’ve seen is lame. If you care, find out!]
The trouble with the book is that it has no particular structure. It is full of fascinating gems like that, which you could remove from the frame story and turn into an excellent common-place book, but the frame story itself never seems to get going nor to lead anywhere. There is no doubt in my mind that this is a totally superficial judgement on my part based on a very sketchy reading, but I (arrogantly) blame the book itself for not drawing me in to a less sketchy reading.
The parts that made the most impression on me were those where he talks about giving and receiving. Not so much the frequently-quoted “when you give yourself, you receive more than you give,” as the concept that in order for you to give, there must be someone to receive. This is not so frequently quoted &mdash I can't find any page on the internet with the exact text, and the point is a subtle one.
There is a difference between giving and meeting demands. The worst thing that one partner can do in a relationship is to cross that thin line between receiving a gift and demanding a service. Saint Exupery compares this kind of relationship to prostitution — you can’t give a prostitute anything, you can only supply her with the payment that she has earned.
If there is a doublet of this post below, it’s because I haven’t worked out how to cancel a blogger post with errors in it. Grrr.
Monday, May 06, 2002
Reading Netscape Time by Jim Clark, I was struck by the following passage
After all, it’s monkish work to write line after line of instructions in arcane computer languages day after day, month after month. The task must be something like translating the Bible from Hebrew and Greek to English (or rather, the other way around).
I’m not sure what “the other way around” is in this context, but never mind that. What this irresistably reminds me of is the statement in the Jerusalem Talmud: וכל האוכל מצה בערב פסח כבא על ארוסתו בבית חמיו [eating unleavened bread the day before Passover is like sleeping with your fiancée in your father-in-law’s house], and my response to Jim Clark’s statement is the same as Smolenskin’s to the Talmud’s: ניסיתי את שניהם ולא מצאתי דומים [I’ve done both and I didn’t notice the resemblance].
Thursday, May 02, 2002
I am having an ambivalent reaction to fixing a bug that has been resisting me for the last nine months.
Of course it’s immensely satisfying to have a result from the many hours that I’ve spent in the debugger chasing it down. Debugging is like a scientific research product (not that I’ve ever done one of those). As you acquire more data, a picture slowly emerges. Hypotheses suggest themselves, and are tested and rejected, or accepted and become the basis of another layer of hypotheses. The process that your bug is a small part of, which may be completely obscure and unfamiliar when you begin working, becomes intimately familiar to you. In extreme cases, you may even dream about it.
By the way, I have never had a Kekulé-like experience and fixed a bug in a dream, but more than once I have found a fix by taking a smoke break. During the debugging process, with my face deep in the details, I often fail to think things through and make deeper connections, but when I stop debugging, get out into the fresh air and think about something else for a few minutes, a solution will often just pop into my mind, seemingly out of nowhere.
You may be in doubt about the value of the possible solutions that you experiment with early in this process, but once you are totally “in the zone” and living and breathing the code, you have no doubts. If your fix is hacky, you feel low, even unclean (though you may accept it out of frustration), but if you discover the One True Fix ™, it will shine out of the screen at you like the Holy Grail. My wife says that she can tell when I have fixed a bug just from the tone of my voice when I pick up the phone, or by seeing the way I walk out of the office.
However, the fact is that last night I felt a bit let-down. Maybe it's the lack of anything substantial produced by all this effort. In the amount of time invested in fixing this bug, I could probably have written a book, designed a cathedral, or produced a play. Instead of that, I have produced one line of code, and no special new or cool functionality that I can point to in the program. It just does what it should have done all along.
And maybe on some level I’m sorry to say goodbye to the bug that I've come to know so well. It’s like splitting up with a partner — even if you know that the relationship was wrong for you and you are better off apart, there is always some sensation of regret.
Or maybe there's some significance to the nine months that this bug has existed. Could I be suffering from post-natal depression?
![[Powered by Blogger]](http://buttons.blogger.com/bloggerbutton1.gif){kind=small}
