I must say that I agree with this. Too many interviews, and rarely with someone makes me go look at either the potential questions or the answers. The one with Stroustrup was an exception, as far as I'm concerned, but that's about it so far for (roughly) the whole of 2000.
Another problem is that if I do look at the list of potential questions, I find that far more questions already got moderated up to 5 than any interviewee will ever have time to answer. So we get tens of questions at level 5, only about 7 or so make it through the final selection (based on what criterion actually?). All this again makes me wonder why I should bother spending my moderation points on helping to select questions.
Seems like the/. interviewing process is a bit flawed, IMHO. At the very least it suffers from fixed the upper limit on moderation points per post that serves us rather well in the normal discussions.
One of the advantages of ksh over bash is that ksh more or less comes standard with any commercial Unix. This means that one can rely on it being available wherever one's scripts need to run. Except, until now at least, on the free Unices but fortunately pdksh addresses this. That's why I use ksh for all my shell scripts.
The same cannot be said about bash. Yes, it can most likely be compiled on all (serious) kinds of Unix, but that doesn't mean it will actually be available on any particular machine. In fact, unless your target "customers" use Linux, bash will most likely not be installed.
Now that AT&T have opened up the real Korn shell, it really is (or can become) universal. It won't kill off bash, however, if only because it's not GPLed.
They may well be actual engineers. I work for an 800+ IT company, and if we were to attend Cebit, you can rest assured that some of the actual product engineers would be there.
See you in March at DATE-2000. And yes, I am an actual product engineer for the stuff that I will be exhibiting there.
The Inverted Pyramid really should be used on the net a lot more then, considering the amount of times that my browser has gone down when I was half done reading some article.
PS: I'm sort of serious, and this definitely *is* on topic. Crash resistance is one of the major advantages that newspapers have over the average PC.
Funny that I should say that, since I never liked that push frenzy on the WWW some time ago... Maybe it's because in the case of newspapers, the push is a lot more gentle.
My newspaper informs me about lots of things that I would not know how to find on the web (and search engines are not a solution for news). It's the kind of stuff that I wouldn't actively go looking for, but will read if my eye catches it in the paper. Some of that can be very interesting and make me go look for more on the net, but it wouldn't happen this often without a newspaper.
In a way, this is similar to Usenet. Usenet will also feed me with stuff that I did not actively go looking for but that might still be interesting. That's also why I still use it. With newspapers the "effective push scope" is even broader that on a Usenet group dedicated to some (set of) topic(s), so I expect I will stop reading Usenet sooner that my newspaper.
An other reason to have newspapers is that computers still are too heavy and clumsy to cary about wherever a newspaper can go. Also, connecting to the net costs money. Newspapers also cost, but comparing the amount of verified information per buck that I get from my paper with the the amount of unverified stuff that I get from the web, the newspaper is cheaper. A lot cheaper actually, if I consider that on the web I'd need to spend a lot of time and money searching for the same amount of news first.
And last be not least: newspapers do not crash while I'm in the middle of an article. Netscape does that all the time on me.
Simply by posting this message you have undone the faulty moderation. As well as any other moderation you may have performed in this discussion, unfortunately
Robbing only half the money actually will in general lead to lesser penalties. Especially if you can convince the judge that did it for "non selfish" reasons.
Anyway, the all important differences that invalidate your analogy are:
robbing is illegal and "always" has been;
if you rob someone, you take something away from that person which was legally his or hers, whereas in this case the domain name had no prior owner.
Finally, not only did the domain name not have a prior owner, also was there no existing company that "obviously" would have a right to it.
I sure have written crappy code in the past, when deadline pressure became too big and the boss was breathing down my neck. This works "fine" in a closed environment: if the program works, things are OK (hmmm, actually, most of the time things already are OK if it merely exists, but my professional pride demands that it also works). But tell me to open source that code, and that same professional pride will make me insist on cleaning it up first.
You seem not to be all that knowlegeable about the past status of things. Which really makes me wonder whether you are actually interested in using it, or only in talking about how bad it is that it (or rather: the current version) was not available.
Anyway: it is deemed now ready for release, which is not the same as being ready. The point of not releasing before it is ready for release is to make sure that people can start using it before they fully understand it. This is good.
I agree that "insightful" is not the correct classification, but otherwise I strongly disagree the rest of your reply and even more with its inflammatory nature.
Worst, however, is that you didn't check that "No Score +1 Bonus" box.
Sometimes I think the +1 bonus is a good thing. But then I see posts like this and say "thank god for moderation!"
I have not seen the webcast and any Transmeta related web site is incredibly slow at the moment, so I've not been able to look for more detailed accounts, but I would guess that it all depends on the definitions of "made a mini Linux distribution for its licensees" and "distribute".
IMHO this part of the GPL can be worked around temporarily by being very careful in what the contracts between (in this case) Transmeta and the licensees say. One way might be to state "we'll make it for you now and you can come and inspect it and play with it at our premises, but we will only deliver the product to you on January 19, 2000 etc."
Have you used it for so-called "serious" programming?
I've seen some amazing software engineering support systems done in Python already. Stuff that is quite a bit more complex to design and implement than make et al. Stuff of the kind that previously made a long time C++ user like me say "this screams for C++". Until, that is, I saw it done and maintained in Python with much less (and cleaner) code and time than would ever be possible in C++.
In the world of business real "unathorized" leaks are in general quickly identified as such. That is to say: they end up being attributed to "an unnamed source at Xyz Inc. stated blahblahblah, from which we can deduce with a high degree of certainty The Company In Question must have organised the leak because blahblahblah".
In the case of Transmeta, all we appear to have seen is (old) patents and massive speculation by people who don't actually know anything besides what others have speculated already.
Note: All the speculation may wel be correct, but that is another topic.
Strictly speaking, 10 years is impossible though. Linus started working on The Thing in the spring of 1991, and first publicly announced something on July 5, 1991. See http://li.org/li/linuxhistory.shtml for more details.
I'm not at all sure forking would be a good idea. In part because some of the better known Linux vendors are backing XFree86 quite a bit. For a fork to be successful in the end, you'd need to convince them to switch. Possible, but definitely not trivial, I'd assume.
In any case I'd wait for 4.0 before forking if I were you. Unless it really takes forever to turn up, that is. Forking right before the XFree86 team comes out with a major new update seems like a recipe for needles integration trouble.
And finally: only the current copyright holders can change a license (and only if they all agree about it). Others need to stick to whatever they got in the first place.
Sorry about getting the comparison wrong. Does the time of year count as an excuse?:-)
Having said that, in 1999, it displayed 99. This is generally correct in a human-computer interface context, since leaving out the century is a well accepted convention (and has been for much longer then we've had computers). It may be a silly or in some contexts even dangerous convention and all that, but that in itself doesn't make it incorrect.
In 2000 it displays 100, which is incorrect except if the spec says something like "display tm_year". So, the question then becomes: does anyone have the spec for this thing written down and approved by management?:-)
Any object not traceable from the ground up to the head is not human end of story. The computer itself could do the check.
Except that there are many such objects that one may be carying about without being a potential terrorist. Think of metal pens, metal belt buckles, a bunch of keys, coins,... Also, how about things like dental fillings and all that? Consider also that the objects are likely to be moving about during the whole process, especially when there is no human operator around. It's very easy for a human to identify the kind of objects I menstioned for what they are, but making a computer do it reliably is an other matter.
The reliability is a major issue. A single false negative can virtually kill the company or deploying agency. Yet, too many false positives and the whole automation isn't worth the investment either.
I'm not saying that it cannot be done and in fact I'm quite sure that it will be done at some point, but...
Speaking of being lame, in JavaScript, things are even lamer. In versions prior to 1.2, if the year is less than 2000, you get the number of years since 1900, but after that you get the full year. To make matters worse, this has been changed in later versions. See http://ourworld.compuserve.com/homepages/Ken_North /y2k_web.htm for more details.
I really wonder who came up with all these wonderful ideas and what stuff they have been smoking at the time.
Well, it doesn't matter who is at fault, it's a Y2K bug anyway. Having said that: it's actually always a coder. If not the one of the application, it's the one of the library, the compiler, the OS, the cpu (yes),...
Not neccessarily. 2Km = 2 Kilo meter = 2000 meter. Only in computerland 2K equals 2048, and while The Bug mainly is a computer problem, counting years is not a computer concept. Besides, what to do with paper forms on which the 19 has been preprinted? It's the basically same problem even with no computer in sight.
Today I read in a local paper (sorry, no link (besides it would be in Dutch)) that someone from the organisation "owning" the problem with the 20000 credit card terminals in the UK said: "This was not a Y2K related issue. It was more like a century problem, but now try to explain that to people out there." Amazing. As if relabeling the error makes it more excusable to have a rollover bug in a system that was supposed to have been fixed for Y2K.
I do agree that the name millenium bug is silly (for more reason than one), but somehow I seriously doubt that that's the idea was what that spokesperson wanted to express.:-)
Slashdot Mirror
Another problem is that if I do look at the list of potential questions, I find that far more questions already got moderated up to 5 than any interviewee will ever have time to answer. So we get tens of questions at level 5, only about 7 or so make it through the final selection (based on what criterion actually?). All this again makes me wonder why I should bother spending my moderation points on helping to select questions.
Seems like the /. interviewing process is a bit flawed, IMHO. At the very least it suffers from fixed the upper limit on moderation points per post that serves us rather well in the normal discussions.
--
The same cannot be said about bash. Yes, it can most likely be compiled on all (serious) kinds of Unix, but that doesn't mean it will actually be available on any particular machine. In fact, unless your target "customers" use Linux, bash will most likely not be installed.
Now that AT&T have opened up the real Korn shell, it really is (or can become) universal. It won't kill off bash, however, if only because it's not GPLed.
--
See you in March at DATE-2000. And yes, I am an actual product engineer for the stuff that I will be exhibiting there.
--
PS: I'm sort of serious, and this definitely *is* on topic. Crash resistance is one of the major advantages that newspapers have over the average PC.
--
My newspaper informs me about lots of things that I would not know how to find on the web (and search engines are not a solution for news). It's the kind of stuff that I wouldn't actively go looking for, but will read if my eye catches it in the paper. Some of that can be very interesting and make me go look for more on the net, but it wouldn't happen this often without a newspaper.
In a way, this is similar to Usenet. Usenet will also feed me with stuff that I did not actively go looking for but that might still be interesting. That's also why I still use it. With newspapers the "effective push scope" is even broader that on a Usenet group dedicated to some (set of) topic(s), so I expect I will stop reading Usenet sooner that my newspaper.
An other reason to have newspapers is that computers still are too heavy and clumsy to cary about wherever a newspaper can go. Also, connecting to the net costs money. Newspapers also cost, but comparing the amount of verified information per buck that I get from my paper with the the amount of unverified stuff that I get from the web, the newspaper is cheaper. A lot cheaper actually, if I consider that on the web I'd need to spend a lot of time and money searching for the same amount of news first.
And last be not least: newspapers do not crash while I'm in the middle of an article. Netscape does that all the time on me.
--
--
It doesn't, and I don't see why it should.
--
Anyway, the all important differences that invalidate your analogy are:
Finally, not only did the domain name not have a prior owner, also was there no existing company that "obviously" would have a right to it.
--
I sure have written crappy code in the past, when deadline pressure became too big and the boss was breathing down my neck. This works "fine" in a closed environment: if the program works, things are OK (hmmm, actually, most of the time things already are OK if it merely exists, but my professional pride demands that it also works). But tell me to open source that code, and that same professional pride will make me insist on cleaning it up first.
--
You seem not to be all that knowlegeable about the past status of things. Which really makes me wonder whether you are actually interested in using it, or only in talking about how bad it is that it (or rather: the current version) was not available.
Anyway: it is deemed now ready for release, which is not the same as being ready. The point of not releasing before it is ready for release is to make sure that people can start using it before they fully understand it. This is good.
--
Worst, however, is that you didn't check that "No Score +1 Bonus" box.
Sometimes I think the +1 bonus is a good thing. But then I see posts like this and say "thank god for moderation!"
--
IMHO this part of the GPL can be worked around temporarily by being very careful in what the contracts between (in this case) Transmeta and the licensees say. One way might be to state "we'll make it for you now and you can come and inspect it and play with it at our premises, but we will only deliver the product to you on January 19, 2000 etc."
Beware: IANAL.
--
--
I've seen some amazing software engineering support systems done in Python already. Stuff that is quite a bit more complex to design and implement than make et al. Stuff of the kind that previously made a long time C++ user like me say "this screams for C++". Until, that is, I saw it done and maintained in Python with much less (and cleaner) code and time than would ever be possible in C++.
--
In the case of Transmeta, all we appear to have seen is (old) patents and massive speculation by people who don't actually know anything besides what others have speculated already.
Note: All the speculation may wel be correct, but that is another topic.
--
Strictly speaking, 10 years is impossible though. Linus started working on The Thing in the spring of 1991, and first publicly announced something on July 5, 1991. See http://li.org/li/linuxhistory.shtml for more details.
--
In any case I'd wait for 4.0 before forking if I were you. Unless it really takes forever to turn up, that is. Forking right before the XFree86 team comes out with a major new update seems like a recipe for needles integration trouble.
And finally: only the current copyright holders can change a license (and only if they all agree about it). Others need to stick to whatever they got in the first place.
--
Having said that, in 1999, it displayed 99. This is generally correct in a human-computer interface context, since leaving out the century is a well accepted convention (and has been for much longer then we've had computers). It may be a silly or in some contexts even dangerous convention and all that, but that in itself doesn't make it incorrect.
In 2000 it displays 100, which is incorrect except if the spec says something like "display tm_year". So, the question then becomes: does anyone have the spec for this thing written down and approved by management? :-)
--
The computer itself could do the check.
Except that there are many such objects that one may be carying about without being a potential terrorist. Think of metal pens, metal belt buckles, a bunch of keys, coins, ... Also, how about things like dental fillings and all that? Consider also that the objects are likely to be moving about during the whole process, especially when there is no human operator around. It's very easy for a human to identify the kind of objects I menstioned for what they are, but making a computer do it reliably is an other matter.
The reliability is a major issue. A single false negative can virtually kill the company or deploying agency. Yet, too many false positives and the whole automation isn't worth the investment either.
I'm not saying that it cannot be done and in fact I'm quite sure that it will be done at some point, but...
--
I really wonder who came up with all these wonderful ideas and what stuff they have been smoking at the time.
--
--
--
Not neccessarily. 2Km = 2 Kilo meter = 2000 meter. Only in computerland 2K equals 2048, and while The Bug mainly is a computer problem, counting years is not a computer concept. Besides, what to do with paper forms on which the 19 has been preprinted? It's the basically same problem even with no computer in sight.
--
I do agree that the name millenium bug is silly (for more reason than one), but somehow I seriously doubt that that's the idea was what that spokesperson wanted to express. :-)
--
--