Mmm. Probably a good idea, but not enough to stop Embrace and Extend (Engulf and Devour?)
I'm not an expert student of Microsoft's activities and offenses, but as I understand it, the problem isn't so much that MS variants/extensions don't conform to the established standard, as that they add to the established standard (hence the term "extensions") in proprietary ways. Then, because of Microsoft's well-known dominance of the market, the world at large starts using the larger, extended variant that Microsoft controls, under the assumption that everyone has the software for it ("everyone has Windows, right?") Suddenly, the de facto standard for the net is Microsoft's standard, which nobody else can wiggle into, since they control it.
Actually, Netscape did that a while too (and I remember how much it annoyed me!). They added all those extensions to HTML, and then everyone used them in their web pages, and suddenly you had to use Netscape or you'd miss all that formatting. There are several features now in HTML that started out as Netscape "extensions". It was a naughty thing when Netscape did it, and it's as naughty when Microsoft does it (if not naughtier).
Sorry if the above was a rehash of the obvious for anyone following Microsoft's history, but the point is that regression tests won't help. A regression test generally specifies that the product must conform to the minimum specification, without mandating that it necessarily fail for things that break the specification. That's good design; it allows for graceful degradation in the face of later versions. But it means that a version with "extensions," à la Microsoft's Kerberos or Netscape's HTML, will pass such a test, with flying colors. Of course they support the standard... they just support some more stuff too. So we're back in the same boat.
There may be other ways to accomplish what you're after... maybe the license specifies that you can't add extensions? Sounds like a bad (if not disastrous) idea, and very much Not the Free Software Way. Maybe there's something. But as I understand it, the tests aren't it.
Yeah; I have a copy of it. He still... pauses... And his pronunciation is execrable. Then again, so is that of most of the other actors. And he plays a character that reminds me of so many Star Trek episodes! His character is a war-hero, saintly and pure of soul... falls in love oh-so-quickly, and ready to turn worlds over for his newfound love... This isn't sounding right. Just that somehow some of the situations seem awfully familiar from his later work.
Various things like this have been tried. You might try looking at Lojban. Even if you don't think anything will work as an international language but English, or that anything will work as an international language at all, learn Lojban. It's a fascinating language and can really make you think and help you get a grip on some of the trickier aspects of language in general.
Lojban's grammar (not its semantics!) is unambiguous and computer-parseable. We (I'm also on the Lojban board, as well as being Assistant Director of the Klingon Language Institute; I get around) have a YACC-based parser that really will parse Lojban sentences, if they conform to the baselined grammar. Lojban's not strictly LALR(1), but is with a little pre-processing. Anyway, so its grammar is computer-understandable, and even the ambiguities in its semantics are at least well-understood. By which I mean that you (or a computer) can know where the ambiguities lie, and what's more you have ways of asking clearly for further clarification of them. Lojban even has a set of exclamations that just express emotion, so something like "Ouch!" translates without relying on someone else knowing how English speakers express pain.
There are some less well-known (to me and probably also to others, since I do try to keep up on these things) attempts in this vein. There are languages that were based on cataloguing all the various concepts to be expressed in a sort of Dewey Decimal System on steroids, with the hope that you could compartmentalize thought into neat nesting categories, and join them up with some mathematical glue. This goes all the way back to Francis Lodowyck's "Common Writing", published in 1647. There was one called... Lincos I think? I can't find my copy. I think that was it. It's more recent, very big on numbers and sets and mathematical notation and such.
In short, your idea isn't new... not necessarily bad, but not untried either.
I recall that it has been proven (fairly irrefutably and uncontroversially) that so-called "Fully-Automated High Quality Machine Translation" is impossible (the proof by a man named Yehoshua Bar-Hillel; I fear I don't have references, but you can probably look them up). During the Cold War, it was thought that language translation could be tackled using cryptography techniques, which is why it was the government decrypters that got the copy of the book in the To Serve Man episode of Twilight Zone. Bar-Hillel proved you couldn't do such a thing: languages simply rely too much on human experience and pragmatics for a machine to do it without being as intelligent and sapient (and as error-prone and demanding of a decent salary) as a person.
That's why Babel Fish were considered such a proof of (non-)existence of God...
I'm the Assistant Director of the Klingon Language Institute, and also a constructed language enthusiast, so I know more or less whereof I speak.
Klingon is not "based" on any particular language or languagegroup. It is certainly true that the inventor of the language (Dr. Marc Okrand) has his field of expertise in Amerindian languages, and therefore Klingon probably shows a stronger influence from those than anything else, but saying it's "based on Navajo" is grossly inaccurate. It has conjugations for object/subject combinations like many Amerindian languages (a feature not entirely peculiar to them, but not common); it has a word-order found in very very few languages (I think the one usually quoted is from South America). But he deliberately didn't "base" it on anything in particular. Notably, many students, on learning Klingon, confide that it's "just like" their favorite or native language. We've heard Basques say it was very Basque, Poles say it seemed Polish, etc. It's certainly simpler than the lot (which is interesting when you consider that unlike Esperanto, Klingon doesn't have simplicity or learnability as a stated goal).
Has Klingon got what it takes to be "the" International language? I tend to doubt it. But that's not why I study it.
Basic English is a headache and a half. It's about as tough as English, for a non-English-speaker: it has just about all the irregularities and plenty opaque idioms from English. And it's a pain for English-speakers too, who have to remember that 90% of their vocabulary is missing. So it's unnatural for them as well, when they have to remember they can't say "selfish" but must instead say "having no thought for others."
Basically, it suffers from just about all the deficiencies of English (as perceived by non-native speakers), including English's famously arcane spelling system, and also is a pain for the native-speakers!
Maybe that's good, at least it's fair... but if that's your logic, you might as well be fair and pick a language that's not native to anyone (or hardly anyone), like Lojban, Esperanto, Klingon, etc...
Re:DeCSS & CPHack aren't illegal until October!
on
'Battling Censorware'
·
· Score: 1
I wonder if one could counter this in court by saying "Hey, I broke it; it was cake. It wasn't effectively controlling access!" After all, if your security is "don't press the Z key!" I doubt you could stop people from using that hack.
More importantly, they could revoke Apex's CSS player key. All future DVDs would be mastered without an Apex key, so they could not be played on an Apex player.
...at which point I, as a (very recent) Apex purchaser would have good cause to be good and mad at Apex or the DVD folks or both. After all, I bought the thing with the express understanding that it would be able to play the same DVDs that other contemporary players can. Why should I get punished because of some feature Apex put in that I knew nothing about? (OK, I did know about it, but that's neither here nor there. And maybe someone else didn't). If they did that, someone would get sued, and probably lose.
Slashdot Mirror
Mmm. Probably a good idea, but not enough to stop Embrace and Extend (Engulf and Devour?)
I'm not an expert student of Microsoft's activities and offenses, but as I understand it, the problem isn't so much that MS variants/extensions don't conform to the established standard, as that they add to the established standard (hence the term "extensions") in proprietary ways. Then, because of Microsoft's well-known dominance of the market, the world at large starts using the larger, extended variant that Microsoft controls, under the assumption that everyone has the software for it ("everyone has Windows, right?") Suddenly, the de facto standard for the net is Microsoft's standard, which nobody else can wiggle into, since they control it.
Actually, Netscape did that a while too (and I remember how much it annoyed me!). They added all those extensions to HTML, and then everyone used them in their web pages, and suddenly you had to use Netscape or you'd miss all that formatting. There are several features now in HTML that started out as Netscape "extensions". It was a naughty thing when Netscape did it, and it's as naughty when Microsoft does it (if not naughtier).
Sorry if the above was a rehash of the obvious for anyone following Microsoft's history, but the point is that regression tests won't help. A regression test generally specifies that the product must conform to the minimum specification, without mandating that it necessarily fail for things that break the specification. That's good design; it allows for graceful degradation in the face of later versions. But it means that a version with "extensions," à la Microsoft's Kerberos or Netscape's HTML, will pass such a test, with flying colors. Of course they support the standard... they just support some more stuff too. So we're back in the same boat.
There may be other ways to accomplish what you're after... maybe the license specifies that you can't add extensions? Sounds like a bad (if not disastrous) idea, and very much Not the Free Software Way. Maybe there's something. But as I understand it, the tests aren't it.
Yeah; I have a copy of it. He still... pauses... And his pronunciation is execrable. Then again, so is that of most of the other actors. And he plays a character that reminds me of so many Star Trek episodes! His character is a war-hero, saintly and pure of soul... falls in love oh-so-quickly, and ready to turn worlds over for his newfound love... This isn't sounding right. Just that somehow some of the situations seem awfully familiar from his later work.
It's at http://www.incubusthefilm.com/.
Various things like this have been tried. You might try looking at Lojban. Even if you don't think anything will work as an international language but English, or that anything will work as an international language at all, learn Lojban. It's a fascinating language and can really make you think and help you get a grip on some of the trickier aspects of language in general.
Lojban's grammar (not its semantics!) is unambiguous and computer-parseable. We (I'm also on the Lojban board, as well as being Assistant Director of the Klingon Language Institute; I get around) have a YACC-based parser that really will parse Lojban sentences, if they conform to the baselined grammar. Lojban's not strictly LALR(1), but is with a little pre-processing. Anyway, so its grammar is computer-understandable, and even the ambiguities in its semantics are at least well-understood. By which I mean that you (or a computer) can know where the ambiguities lie, and what's more you have ways of asking clearly for further clarification of them. Lojban even has a set of exclamations that just express emotion, so something like "Ouch!" translates without relying on someone else knowing how English speakers express pain.
There are some less well-known (to me and probably also to others, since I do try to keep up on these things) attempts in this vein. There are languages that were based on cataloguing all the various concepts to be expressed in a sort of Dewey Decimal System on steroids, with the hope that you could compartmentalize thought into neat nesting categories, and join them up with some mathematical glue. This goes all the way back to Francis Lodowyck's "Common Writing", published in 1647. There was one called... Lincos I think? I can't find my copy. I think that was it. It's more recent, very big on numbers and sets and mathematical notation and such.
In short, your idea isn't new... not necessarily bad, but not untried either.
I recall that it has been proven (fairly irrefutably and uncontroversially) that so-called "Fully-Automated High Quality Machine Translation" is impossible (the proof by a man named Yehoshua Bar-Hillel; I fear I don't have references, but you can probably look them up). During the Cold War, it was thought that language translation could be tackled using cryptography techniques, which is why it was the government decrypters that got the copy of the book in the To Serve Man episode of Twilight Zone. Bar-Hillel proved you couldn't do such a thing: languages simply rely too much on human experience and pragmatics for a machine to do it without being as intelligent and sapient (and as error-prone and demanding of a decent salary) as a person.
That's why Babel Fish were considered such a proof of (non-)existence of God...
I'm the Assistant Director of the Klingon Language Institute, and also a constructed language enthusiast, so I know more or less whereof I speak.
Klingon is not "based" on any particular language or languagegroup. It is certainly true that the inventor of the language (Dr. Marc Okrand) has his field of expertise in Amerindian languages, and therefore Klingon probably shows a stronger influence from those than anything else, but saying it's "based on Navajo" is grossly inaccurate. It has conjugations for object/subject combinations like many Amerindian languages (a feature not entirely peculiar to them, but not common); it has a word-order found in very very few languages (I think the one usually quoted is from South America). But he deliberately didn't "base" it on anything in particular. Notably, many students, on learning Klingon, confide that it's "just like" their favorite or native language. We've heard Basques say it was very Basque, Poles say it seemed Polish, etc. It's certainly simpler than the lot (which is interesting when you consider that unlike Esperanto, Klingon doesn't have simplicity or learnability as a stated goal).
Has Klingon got what it takes to be "the" International language? I tend to doubt it. But that's not why I study it.
~mark
Basic English is a headache and a half. It's about as tough as English, for a non-English-speaker: it has just about all the irregularities and plenty opaque idioms from English. And it's a pain for English-speakers too, who have to remember that 90% of their vocabulary is missing. So it's unnatural for them as well, when they have to remember they can't say "selfish" but must instead say "having no thought for others."
Basically, it suffers from just about all the deficiencies of English (as perceived by non-native speakers), including English's famously arcane spelling system, and also is a pain for the native-speakers!
Maybe that's good, at least it's fair... but if that's your logic, you might as well be fair and pick a language that's not native to anyone (or hardly anyone), like Lojban, Esperanto, Klingon, etc...
I wonder if one could counter this in court by saying "Hey, I broke it; it was cake. It wasn't effectively controlling access!" After all, if your security is "don't press the Z key!" I doubt you could stop people from using that hack.