On Thu, 19 Jan 2006 11:21:45 +0100, Carlo Traverso <traverso@dm.unipi.it> wrote:
I have times ago done some work to build a multilingual form of gutcheck, (and I still think that it is a very reasonable aim) but I stopped when Jim refused the very idea that this should be done.
"Opinions about Tolstoy and his work differ, but on one point there surely might be unanimity. A writer of world-wide reputation should be at least allowed to know how to spell his own name. Why should any one insist on spelling it "Tolstoi" (with one, two or three dots over the "i"), when he himself writes it "Tolstoy"? The only reason I have ever heard suggested is, that in England and America such outlandish views are attributed to him, that an outlandish spelling is desirable to match those views." Love that quote. From Louise Maude's Translator's Preface to "Resurrection". I really must re-scan that, if only to capture the image of Tolstoy's signature -- with a "y" -- above those words. I'm not a writer of world-wide reputation, of course, but I've recently heard such outlandish views attributed to me that I'm beginning to think of signing myself "Jim Tinslei", with, possibly, three decadent dots over the "i". So, did I ever "refuse the very idea that this should be done"? The society that is one of the referents of "Project Gutenberg", as I understand it -- and I'm not at all sure that I do -- is a pretty good model of a Libertarian society. It's even better than a Real Life Libertarian society, since anyone can opt out; try doing that next time your local tax-collector sends you a letter. People do (a) what they want to do and (b) what they think should be done strongly enough that they're willing to spend the hours of their lives doing it. Some people also do (c) what Other People want them to do. Occasionally, or usually. PPVs and WWs spend a lot of time on projects in which they, personally, have no interest. Toolmakers try to accommodate the people who use their tools. DP admins solve problems. Gravediggers move difficult projects along. Like that. The complex society that exists today in and around PG would all fall apart if some people didn't offer themselves as something of a "public utility" in some limited sphere. So I'm used to the idea that people write to me out of the blue asking for help or advice, just as I ask other "public utilities" within the project for help and advice. But in such cases I, or they, are free to refuse, or do something different. Project Gutenberg, however you define it, doesn't sign anybody's paycheck, or make anybody do anything at all. Neither do Michael or Greg as individuals. In fact, those two worthies would walk a country mile in tight shoes to assure you -- gesticulatingly -- that they have about as much influence over what I do as Uri Geller's daily horoscope has over the shape of Reese Witherspoon's toenails. What's more, having the experience of being a public utility in PG yourself, you know this better than most, which is why, when I dug your original email on the subject in July 2002 out of the dumpster of my archives, I was a little annoyed all over again that you had copied Michael and Greg on it. 'Sfunny: I didn't remember the thread, but when I saw the e-mail, I did remember that little sting of annoyance at the assumption that either of them had anything at all to do with my decisions about gutcheck. Which is, of course, as nothing to the annoyance gutcheck has inflicted over the years on various producers, so I guess, karmically, I have it comin'. People who "grew up" in DP were "born" with others looking over their shoulders, and so expect their homework to be corrected unmercifully. Everybody there has fully internalized the knowledge that they, and everyone else, makes mistakes, or, as Juliet more correctly and insightfully remarked, _overlooks_ mistakes. Your own recent excellent work on quantifying that will be invaluable in several ways. Producers who had been making certain kinds of mistakes for years without being aware of it, or having anyone correct them, though, fully appreciated the pun in the name. It is no fun at all having these things pointed out to you for the first time by someone else. Dave's comments are really quite temperate compared to many of the love-notes I received back in 2000-2002. My favorite was "DON'T YOU DARE RUN THIS THING OVER MY OLD TEXTS!!" I was more than somewhat sick myself when I first exposed some of my old work to jeebies, and saw the full extent of my own heebieness, but at least in that case, nobody else saw my shame, and I had no-one to be annoyed at but myself. I mention my annoyance because on re-reading what I wrote in response to your proposal, it does jump off the screen at me, and I apologize belatedly for that. It doesn't, however, have any bearing on my decisions then or now. What you actually proposed was that you should carve up gutcheck into separate files, dealing with separate languages. If there had ever been a day when I decided to sit down and write gutcheck, that's what I might well have done from Day One, but there never was such a day. To me, it's just a handy platform into which I can plug checks that I find useful. As I said at the time, and so often before and since, I don't actually think that the language-specific typo-checking functions should really be in there at all; every text needs a spellcheck, and for texts that have been spellchecked, these functions are only a source of false positives. That's why I added the -t switch when I started sending it out to other people. For me, they were handy as a quick way of getting a hint whether an incoming file had been spellchecked or not. Unfortunately, some producers lulled into a false sense of security by not seeing typos flagged in gutcheck didn't do the spellcheck, which was a problem I had to address around that time, but it seems to be resolved now. That's what I said, and that's what I still believe. I do think that punctuation checks for LOTE are an appropriate add-in, but as a devout monoglot, I'm in no position to define them. I don't have the experience of finding certain error-patterns by hand in LOTE texts. People have suggested specific changes like this from time to time, and I have usually incorporated them, where they don't cause problems somewhere else. A few days ago, I asked any PPVs who want certain punctuation type checks (or removal of existing checks) for LOTE to define some for me. We'll see what comes out of that. Until I see what the requested checks are, I'm not going to decide how to make the changes. I'm certainly not going to refactor the code, or commit myself to working with somebody else's refactored code, in advance of knowing in what way it needs to be changed. Reading over old emails is weird; it brings back context. You wrote to me when I was just setting up the SF site, to get it installed before I released the FAQ and to give the Software Site a permanent link. Up until then, people had got gutcheck directly from me, and often asked for individualized versions, which I mostly made for them. If the checks seemed good by my usual tests, I added them to "my" gutcheck as well. That was the way it worked in that era. I looked forward, at that time, to getting the damthing OUT, so that people could do their own customizing, and I would be free. Free!! Bwahahahah!! Heh. I wished you well in your own customization, and I still do. The volume of LOTE is much greater than it was then, and maybe somebody working in that area (those areas?) will do their own thing. Great! Maybe they'll ask me to customize some specific checks. Very occasionally, people do that still. Maybe some PPVs in specific languages will get together and suggest a coherent agenda to make gutcheck (or some variant thereof) friendly to those languages. I hope they do. They haven't yet. Until that happens, I have more than enough things I want to do, and think should be done, not to spend my limited PG time chasing Other People to tell me things they want me to do . . . or, for that matter, self-indulging in writing long posts to the vandalized wasteland that was once a productive resource for people making etexts. I really have been very lazy since the Christmas break. Back to the grindstone. jim