Now if they could only supply some good programmings...
A couple of years back, my wife got herself a present of a nice big screen to go with the DVD player, so she can watch all the old movies that she likes.
Then, about a month ago, she got herself a new Powerook G4, the one with the "superdrive" that does CD and DVD. She hasn't turned on the TV since then. She just takes the powerbook into the bedroom lies down, pops in the DVD, and she's happy. She joined Netflix so she can "rent" lots of old movies on DVD cheap, and has already bought a couple of them.
That's a large part of what cost AT&T in the ATT/USL vs BSD case -- AT&T had incorporated BSD code without the BSD copyright notices,...
In fact, anyone who worked with Sys/V much back in the 80's saw a lot of evidence of this. It included a lot of shell scripts with known origins, that included AT&T copyright notice. One of the most blatant examples was their version of the/bin/true command. I saw several "releases" of this, and several times I had the fun of posting the entire program to newsgroups, openly challenging AT&T's lawyers to sue me for violating their claimed copyright. I never heard from them.
Yes, they really did put their copyright notice in a file that was otherwise just blank lines. This makes it obvious that they had just run a program that added their copyright notice to everything. This example alone would probably have been enough to shoot down all their other claimed copyrights in any courtroom.
I wonder what SCO's version of/bin/true looks like?
On the off chance that someone wants to follow through with this idea, here's a traceroute that identifies their ISP:... 12 p4-1-0-0.a00.dnvrco02.us.ra.verio.net (129.250.16.52) 127.103 ms 127.464 ms 131.445 ms 13 p1-0-2-0.a00.dnvrco02.us.ce.verio.net (198.173.159.254) 120.930 ms 117.026 ms 118.175 ms 14 *
That should be sufficient information to contact verio.net.
To put this in some sort of historical perspective, I recently read an interesting observation about the coverage of the European exploration of the North American coast. The writer started with the observation that history books pay a lot of attention to the various expeditions from northern Europe during the 1500's. The Spanish and Portuguese are described as exploring South and Central America.
But there are records in Portuguese and Spanish archives showing that by around 1520, they had a regular (and highly profitable) fishing operation established on the Grand Banks, off the coast of New England and Nova Scotia. They had discovered the North Atlantic Gyre by then, and used it to get to the Grand Banks and back home fairly quickly. In English histories, the discovery of this circulation is generally attributed to Ben Franklin, 250 years later.
There's a lot of history of the English-speaking world simply ignoring exploration and advances in the rest of the world.
Here in the US, we almost never hear of ESA launches, unless there's some sort of disaster. Most Americans couldn't tell you what "ESA" stands for, and have no idea that any European countries other than Russia may have launched anything into space. Those few have heard of such things would tell you that the few launch attempts have mostly been failures.
Well, how I look at it is as a programmer. It seems to me that it would be fairly straightforward to write a routine that scans the disk for plain-text and html files, reads each, extracts anything formatted like an email address, and sends them all back to headquarters. There are a lot of spammers that would pay us for such lists, y'know.
Of course, you'd want the routine to record when it did this, in a binary format in some hidden file, so that it wouldn't do it more than once every few months. This would complicate the job just a little.
why does Linus Torvalds have to have 'S' at the end of his first and last name? I can't figure out where the apostrophy goes.;)
At this point you take advantage of the tradition in English of using the plural and/or possessive form of the language that the word came from. Since Linus, despite his Norse-sounding last name, is a native of Finland and speaks Finnish, you can use the Finnish forms.
So the possessive form would be "Torvaldsin" or "Linusin".
See, there's actually a practical use for all those silly-sounding foreign inflections that they tried to teach you in school.;-)
Yeah; it's not a good idea to tell people that they have weak security. For a really good example, ask google about "Randal Schwarz". His story is going onto a decade now, and still isn't over.
Basically, he had done a lot of consulting work for Intel, and they gave him permanent free accounts on some machines to use as he wished when not on a contract. He saw a new company doc about how to deal with poor passwords. So he thought he'd help them out by nabbing a few password crackers off the Net and applying them to nearby machines. He found that some company VPs had easily-guessed passwords. While he was writing up a report, the sheriff showed up at his door with an arrest warrant. He is now a conviced felon.
Reading between the lines, it seems pretty clear that the people in the legal system think this is ridiculous, and it's really Intel who should be convicted and punished. But there seems to be little that can be done about it. As the judges read the laws, following the company's published guidelines and testing security is a felony, no matter how stupid that sounds. Telling people in the company that their VPs are violating the company's own security rules is also a crime.
So if you find problems, the best practice is to keep quiet about it.
We were thinking "proof of concept", and "Hey, we keep hearing all these wonderful things about OSX; let's give it a try".
If you were developing tools that ran thru a web interface, why on earth were you going for Apple hardware?
That's not all that it does. The web interface is in fact an add-on, because the clients for this software are just now being dragged kicking and screaming into the 1990's, and their management doesn't have a clue about this newfangled web stuff.;-)
, or else your team are less than competent.
Oh, I think we'd all agree to that, with big grins. We're a gang of mostly unix/linux hackers. A few have user-level experience with Macs, but nobody has ever done any Mac software development. So we're approaching OSX as newbies, and understand quite well that we're not (yet) competent with it. One major question is whether we want to be.
So far, the concensus seems to be that OSX was an interesting diversion, but we should concentrate on making it run on linux. We do also have a couple of FreeBSD and Solaris boxes, and the software runs on them with just a "make install". But linux has the edge, because we can get things running there so quickly, and it seems to install on whatever hardware our managers decide to order this week.
I think I'll work on a personal project to try to port a lot of it to OSX. But it'll have to be an unofficial project, for my own edification, because the rest of the team has pushed the idea to a fairly low position on the stack. Not that anyone dislikes Apple, of course. We're all rooting for them. But we have to get things running, and the OSX wall seems just a bit too high for us at the moment.
Would be interesting to know how the system and software works, but then again, that information could be dangerous in the wrong hands."
Well, maybe, but if it can be kept secret by the authorities, they'll just "explain" it with reassuring PR, while not bothering to do any real fixes to the problems.
A lot of us have had far too much experience with big organizations to believe that secrecy will lead to solving the problems. The right way to prevent such things is to make the entire system public information. Then independent engineers can study it, point out the weakness, and suggest solutions, without worrying about losing their jobs if they go public with the bug reports.
(Hmmm... This sounds a lot like the explanations of why Open Source software is so much more secure and reliable than proprietary software.;-)
Creating a UFS partition would solve this problem.
No, it wouldn't. We can't ask customers to do that. They won't; they'll just go buy something else. (Actually, they're installing RedHat systems. Or rather, they're ordering the hardware, and we're doing the installation.;-)
Apple provides documentation about porting Unix apps for those who RTFM
Two problems here: One is that it took us several weeks to stumble across the actual problem. There was nowhere that we could find any clues about the bizarre file-naming problem, just insane behavior from the code that got "The code can't do that!" responses from the people trying to debug it. When we finally noticed the file-name problem, it was easy to find documentation. But until you know what keywords to look for, you don't find anything.
Second, even with that link, and knowing what I'm looking for, several hours of study didn't turn up even the slightest clue that there was a file-naming problem lurking. I'd seen a few of those pages already, actually. If there's a warning anywhere about this problem, I still haven't found it.
Apple's documentation for porting software should start off with a very prominent list of all the software gotchas. Case insensitivity is a major problem, and should be one of the first things. If you're writing a single, monolithic executable, it's not a biggie. But if you're writing a flock of tools tied together by scripts and web pages, it's a real killer.
Further down the list should be the usual library incompatibilities. But we expect those, and know how to add code to our Configure script, and how to put #ifdef lines in the C code.
If only we could get find(1) to work sanely. Sometimes it does; sometimes it spawns one or a flock of shells that eat up all the cpu and memory. We suspect it's due to file-name problems, and we've found a few of them. But it keeps biting us, even with bare find commands with no aliases or scripts involved. It's very difficult to diagnose and fix this sort of problem.
We're finding that every one of our shell scripts needs to be rewritten, even 20-year-old Bourne-shell scripts. This is not a pleasant thing to have to report about what is otherwise an excellent machine to work on.
Wrong. Drag Terminal to the dock, and then click on it there.
That way, you'll never have to do steps 1-6 again.
Oh, and you need to know that clicking on the Terminal icon a second time won't get you a second terminal. You need to go to a terminal window and use CMD-N, where CMD is that funny key with an Apple logo and/or a four-leaf-clover icon.
And Terminal is not all that much like an xterm, so you have a minor learning experience ahead. But it's not too bad.
Everyone is comfortable with windows, even if they don't like it.
Nonsense. First off, there are a lot of us who grew up with other kinds of systems. And for a second example, my wife just got herself a new Mac Powerbook. She has worked with DOS and Windows at work since they first came out, and should be as comfortable with them as anyone. She has more and more been saying how much she hates Microsoft, and explains that it's from long experience.
Her comments on the Mac? By the end of the first week, she was saying how much she loves having a computer that works. She carries it everywhere, even in her tennis bag. (And she complains about poor wireless coverage. "Here in the Boston area!" Two weeks ago, she had never used wireless networking.)
Yes, she has stumbled across the things that work differently. But she points out that every new version of Windows has the same kinds of incompatibilities. And the Mac's behavior makes a lot of sense to her, unlike all the things in Windows that are still baffling after all those years.
Yesterday, after she had complained about several baffling things that IE was doing, and commenting that the Safari that I'd installed wasn't all that much better, I spent maybe 3 minutes downloading and installing mozilla 1.4. After an hour of playing with it, she decided to trash IE and Safari, because mozilla was already so much better and easier to use. She'd already downloaded and tried out a couple of themes, set up the mail reader, and so on, which can be several days work on a new Windows box. She'd also primed the mail reader's junk-mail handler and watched it correctly classify several new pieces of spam. All within an hour.
But she doesn't much like the single titlebar across the top. Neither do I.
i'd have to play with mac in non-critical settings to be familiar with the platform to form my own opinion and figure out the gotchas.
There are some very serious gotchas. About a month ago, I copied all of a project to OSX to test portability problems. Nearly all the scripts, no matter what language, went totally berserk. It took a lot of testing to diagnose the cause of "impossible" behavior. It seems that there's some strange sort of case-insensitivity in a lot of OSX, perhaps in the file system itself. Here's a demo from a test box:
: mkdir tmp; cd tmp : touch FooBar foobar; ls -l total 0 -rw-r--r-- 1 jc staff 0 Aug 16 09:48 FooBar : rm * : touch foobar FooBar; ls -l total 0 -rw-r--r-- 1 jc staff 0 Aug 16 09:49 foobar :
So you can't create two files whose names differ only by capitalization. This shoots down much of the naming in this project. What happens seems to be the worst case: The program that attempts to create the second file succeeds as if the file doesn't exist, but it overwrites the first file.
After discovering this, I asked (with an incredulous tone of writing) in a Mac newsgroup, and got verification. After a few weeks of testing, I still don't understand the rules, but they aren't like any other unix/posix system I've dealt with.
This in itself is a showstopper for this project. Someone suggested that we could install a different file system that doesn't do this, but of course we can't ask customers to do that.
This also applies to the search path. One of the things we'd relied on was that the system libraries are pretty much full of things (executables, functions) whose names are all lower case. So we can pretty much avoid collisions with "system" things by using one or two upper-case letters in names. But this doesn't work on OSX. If your program execs FooBar or fooBar, and there's a foobar earlier in the search path, then foobar is what runs.
Hunting down such name conflicts and fixing them can be incredibly time consuming. The symptoms are bizarre in the extreme.
Now, one could argue that a lot of users like case insensitivity, and an OS can do whatever its builders like. True. But programmers need to know the rules, or they can't write software that works. In this case, the rules have changed from something simple and easy to understand, to something that (from a programmer's viewpoint) is complex and mysterious. File-name matching tests have results that are difficult to explain or predict. The rules don't seem to be documented anywhere that we can find. So software won't port to OSX easily, and software development constantly stumbles on code that gets (or writes) the wrong file. And since the search path is affected, scripts work very differently for a user with a different search path than those used by the developers.
Bad decisions like the one made by Lindows to run as root be default can lead to Linux having as bad a reputation as Microsoft.
But... but... Lindows is supposed to be emulating Windows. Running as root is required to do this job. After all, imagine the criticisms from Windows users if they didn't do this.
"What? I can't overwrite any file on the system unless I give some stupid root password (whatever that is)? They told me this would work just like Windows. They lied to me. I want my money back."
MS's management has told us and told us: Windows users don't want security. They want capability. So give them the power to do what they want. It's just those outdated unix systems that limit what their users can do to their own machine.
If SCO's claims are without merit, then they have placed themselves at a huge risk of a substantial judgment against them. Of all our sakes, I hope that this is the case.
Well, lotsa luck collecting. By the time the case is decided, SCO's officers will be sitting pretty in Rio, and the company will be bankrupt with a debt of a billion or so. How do you think you'll ever get any blood out of that turnip?
Exchanging illegal content over the internet is an illegal act,...
I'm always interested in how an unscrupulous web site could take advantage of such laws. This one's trivial. If you make the mistake of visiting one of my sites with javascript enabled, I can send you to jail quite easily.
One of my js demos is of image "preloading". It's useful: While you look at my main page, JS is silently downloading the images from the rest of my pages, so that subsequent pages come up faster. But I don't have to preload images from my site; I can use an arbitrary URL and "preload" any file from any site. And none of my pages has to ever use them. Think of this for a few seconds...
In much of the world, having images on your machine like the goatse pic or any random pronography is illegal. So what my web page does is "preload" a whole flock of these into your browser cache. My other pages don't actually reference them, but that's not a problem. The problem comes when I use my subpoena power to find out who has been caught by my JS and now has the images on their disk. I send a tipoff to the local authorities. You get a knock on your door, they examine the directories that my note told them about, and sure enough, the images are there. So it's off to jail with you.
Or I could use the same thing to "preload" a few commercial MP3's into your browser's cache. Then I tip off the recording company that you've downloaded them and have them on your disk.
Hmmm... Maybe I oughta teach this technique to a few of the more oppressive governments. They can set up "public information" pages that preload classified docs into your cache.
Needless to say, I like to turn off JS, Active-X, and other scripting goodies before I use any browser. Or I use a browser that doesn't have them in the first place. But most users just use IE or maybe Netscape out of the box, and don't know how to configure such things. So this law, like the DMCA, makes them vulnerable to just this sort of trick.
...will be wasted if you cannot defend your property in court.
But it will probably never happen. There have been any number of explanation from lawyers about why it's unlikely.
The basic scenario is that management tells their lawyers to prepare for a lawsuit challenging some GPL'd software that they want to convert to a proprietary product. The lawyers say "You have a choice here. You can obey the GPL now. Or you can challenge it, go to court, lose, and obey the GPL afterwards."
And, of course, one basic probblem is that if you challenge the GPL, until the case is settled, your rights to the software become the default rights that you have to any copyrighted material: You have no rights, other than what the copyright owner gives you. Since you've filed suit against the owner, they aren't going to give you anything. So for the duration of your court case, you can't use the software at all.
This is usually clear enough to convince even the densest PHB that maybe challenging the GPL isn't in the company's best interest.
SCO's lawyers are certainly smart and knowledgeable enough to understand all this. They have no intention of carrying a court case through to a decision. They are merely trying for maximum damage (and publicity) before they have to back out.
In Germany, this happened very quickly. In the US, it'll take years.
If anyone is still using plaintext to send passwords over their lan they are insane.
Well, a lot of people don't have any choice. Our cable ISP here, for example, provided the usual email accounts, and for a lot of customers, that is their only email. If you use it, you have no choice other than POP, and I haven't seen anything in several mailers that talks about encrypting the passwords. Our ISP doesn't actually block port 25, so you could run your own mailer. This isn't feasible for most customers, though, for several reasons. One is the dynamic IP addresses and insane hostnames. I've fixed that by using one of the many independent registration services, but to most customers, that would be utterly baffling and unusable. Another problem is that running your own email server is in fact in violation of the TOS in the ISP's contract, and they can legally block your port(s) or kick you off entirely at any time, without warning or recourse.
So for most non-geek customers, unencrypted POP passwords are the only option. There's probably no way they could even learn from the ISP that there's a problem; they certainly wouldn't get (or understand) any advice on how to fix it.
(Myself, I use an account at a school. It has been stable and usable for over 15 years now, unlike commercial email accounts that force you to change your address every 6 months whenever there's a merger, buyout, or corporate renaming. And I can use a plain-text mail reader, eliminating all problems with virii, worms and the like. But I'm not sure I'd recommend this to the typical non-geek.)
What I did say was that NO ONE PAYS ATTENTION to this stuff. No one reads anything outside of the sports section and the front page. No one is scouring the net looking for what information they can find about what is going on "behind the scenes".
Um, it appears that you do. And I do. And, from the messages here, it's obvious that at least a few/. readers do.
Politics isn't much influenced by the masses that don't pay attention. It's influenced by the activists of all sorts who get out there and do something.
Of course, making a reasonable, informed argument in a forum like this may seem like spitting into the wind. But you're doing something.
So you can really claim that NO ONE is paying attention. The question is whether there are enough of us to make a difference.
There have been occasional suggestions that anything that goes out of production (or support, for computers) for a year becomes public domain. This is usually suggested in terms of all those old movies that are rotting in company vaults while they sit on the copyright and refuse any public access to them.
It would make a lot of sense to lobby for such a copyright law for software. It would drag a lot of companies (kicking and screaming) into line with the public interest.
Well, around these parts (Bush/Cheney America), you're expected to do the proper, conservative, market-driven thing: You offer to sell it to the highest bidder.
Anything that doesn't earn you a good profit is a dead giveaway that you're one of those leftist, socialistic types. Probably a terrorist. In that case, you should report yourself to the Dept of Homeland Security.
Slashdot Mirror
Heh. Very good. I'd have used a lot more words:
...
...
Now if they could only supply some good programmings
A couple of years back, my wife got herself a present of a nice big screen to go with the DVD player, so she can watch all the old movies that she likes.
Then, about a month ago, she got herself a new Powerook G4, the one with the "superdrive" that does CD and DVD. She hasn't turned on the TV since then. She just takes the powerbook into the bedroom lies down, pops in the DVD, and she's happy. She joined Netflix so she can "rent" lots of old movies on DVD cheap, and has already bought a couple of them.
Television is so 2002
Finally I can see all those Benny Hill shows that I missed.
...
Talk about making English culture available to the world
That's a large part of what cost AT&T in the ATT/USL vs BSD case -- AT&T had incorporated BSD code without the BSD copyright notices, ...
/bin/true command. I saw several "releases" of this, and several times I had the fun of posting the entire program to newsgroups, openly challenging AT&T's lawyers to sue me for violating their claimed copyright. I never heard from them.
/bin/true looks like?
In fact, anyone who worked with Sys/V much back in the 80's saw a lot of evidence of this. It included a lot of shell scripts with known origins, that included AT&T copyright notice. One of the most blatant examples was their version of the
Yes, they really did put their copyright notice in a file that was otherwise just blank lines. This makes it obvious that they had just run a program that added their copyright notice to everything. This example alone would probably have been enough to shoot down all their other claimed copyrights in any courtroom.
I wonder what SCO's version of
On the off chance that someone wants to follow through with this idea, here's a traceroute that identifies their ISP: ...
12 p4-1-0-0.a00.dnvrco02.us.ra.verio.net (129.250.16.52) 127.103 ms 127.464 ms 131.445 ms
13 p1-0-2-0.a00.dnvrco02.us.ce.verio.net (198.173.159.254) 120.930 ms 117.026 ms 118.175 ms
14 *
That should be sufficient information to contact verio.net.
And thirdly, what corporate secret is more important than the due processes of democracy?
Uh, maybe the fact that there's a backdoor that enables rigging the election results?
If you think this is paranoid, you have no concept of how democracy works (or sometimes doesn't work).
To put this in some sort of historical perspective, I recently read an interesting observation about the coverage of the European exploration of the North American coast. The writer started with the observation that history books pay a lot of attention to the various expeditions from northern Europe during the 1500's. The Spanish and Portuguese are described as exploring South and Central America.
But there are records in Portuguese and Spanish archives showing that by around 1520, they had a regular (and highly profitable) fishing operation established on the Grand Banks, off the coast of New England and Nova Scotia. They had discovered the North Atlantic Gyre by then, and used it to get to the Grand Banks and back home fairly quickly. In English histories, the discovery of this circulation is generally attributed to Ben Franklin, 250 years later.
There's a lot of history of the English-speaking world simply ignoring exploration and advances in the rest of the world.
Here in the US, we almost never hear of ESA launches, unless there's some sort of disaster. Most Americans couldn't tell you what "ESA" stands for, and have no idea that any European countries other than Russia may have launched anything into space. Those few have heard of such things would tell you that the few launch attempts have mostly been failures.
Well, how I look at it is as a programmer. It seems to me that it would be fairly straightforward to write a routine that scans the disk for plain-text and html files, reads each, extracts anything formatted like an email address, and sends them all back to headquarters. There are a lot of spammers that would pay us for such lists, y'know.
Of course, you'd want the routine to record when it did this, in a binary format in some hidden file, so that it wouldn't do it more than once every few months. This would complicate the job just a little.
What defense do you have against this?
why does Linus Torvalds have to have 'S' at the end of his first and last name? I can't figure out where the apostrophy goes. ;)
;-)
At this point you take advantage of the tradition in English of using the plural and/or possessive form of the language that the word came from. Since Linus, despite his Norse-sounding last name, is a native of Finland and speaks Finnish, you can use the Finnish forms.
So the possessive form would be "Torvaldsin" or "Linusin".
See, there's actually a practical use for all those silly-sounding foreign inflections that they tried to teach you in school.
(And Finnish has lots and lots of them.)
Yeah; it's not a good idea to tell people that they have weak security. For a really good example, ask google about "Randal Schwarz". His story is going onto a decade now, and still isn't over.
Basically, he had done a lot of consulting work for Intel, and they gave him permanent free accounts on some machines to use as he wished when not on a contract. He saw a new company doc about how to deal with poor passwords. So he thought he'd help them out by nabbing a few password crackers off the Net and applying them to nearby machines. He found that some company VPs had easily-guessed passwords. While he was writing up a report, the sheriff showed up at his door with an arrest warrant. He is now a conviced felon.
Reading between the lines, it seems pretty clear that the people in the legal system think this is ridiculous, and it's really Intel who should be convicted and punished. But there seems to be little that can be done about it. As the judges read the laws, following the company's published guidelines and testing security is a felony, no matter how stupid that sounds. Telling people in the company that their VPs are violating the company's own security rules is also a crime.
So if you find problems, the best practice is to keep quiet about it.
Well, of course you can do anything in perl. First you write a routine that emulates a Turing machine ...
I mean, what were you guys thinking?
;-)
We were thinking "proof of concept", and "Hey, we keep hearing all these wonderful things about OSX; let's give it a try".
If you were developing tools that ran thru a web interface, why on earth were you going for Apple hardware?
That's not all that it does. The web interface is in fact an add-on, because the clients for this software are just now being dragged kicking and screaming into the 1990's, and their management doesn't have a clue about this newfangled web stuff.
, or else your team are less than competent.
Oh, I think we'd all agree to that, with big grins. We're a gang of mostly unix/linux hackers. A few have user-level experience with Macs, but nobody has ever done any Mac software development. So we're
approaching OSX as newbies, and understand quite well that we're not (yet) competent with it. One major question is whether we want to be.
So far, the concensus seems to be that OSX was an interesting diversion, but we should concentrate on making it run on linux. We do also have a couple of FreeBSD and Solaris boxes, and the software runs on them with just a "make install". But linux has the edge, because we can get things running there so quickly, and it seems to install on whatever hardware our managers decide to order this week.
I think I'll work on a personal project to try to port a lot of it to OSX. But it'll have to be an unofficial project, for my own edification, because the rest of the team has pushed the idea to a fairly low position on the stack. Not that anyone dislikes Apple, of course. We're all rooting for them. But we have to get things running, and the OSX wall seems just a bit too high for us at the moment.
Would be interesting to know how the system and software works, but then again, that information could be dangerous in the wrong hands."
... This sounds a lot like the explanations of why Open Source software is so much more secure and reliable than proprietary software. ;-)
Well, maybe, but if it can be kept secret by the authorities, they'll just "explain" it with reassuring PR, while not bothering to do any real fixes to the problems.
A lot of us have had far too much experience with big organizations to believe that secrecy will lead to solving the problems. The right way to prevent such things is to make the entire system public information. Then independent engineers can study it, point out the weakness, and suggest solutions, without worrying about losing their jobs if they go public with the bug reports.
(Hmmm
Creating a UFS partition would solve this problem.
;-)
No, it wouldn't. We can't ask customers to do that. They won't; they'll just go buy something else. (Actually, they're installing RedHat systems. Or rather, they're ordering the hardware, and we're doing the installation.
Apple provides documentation about porting Unix apps for those who RTFM
Two problems here: One is that it took us several weeks to stumble across the actual problem. There was nowhere that we could find any clues about the bizarre file-naming problem, just insane behavior from the code that got "The code can't do that!" responses from the people trying to debug it. When we finally noticed the file-name problem, it was easy to find documentation. But until you know what keywords to look for, you don't find anything.
Second, even with that link, and knowing what I'm looking for, several hours of study didn't turn up even the slightest clue that there was a file-naming problem lurking. I'd seen a few of those pages already, actually. If there's a warning anywhere about this problem, I still haven't found it.
Apple's documentation for porting software should start off with a very prominent list of all the software gotchas. Case insensitivity is a major problem, and should be one of the first things. If you're writing a single, monolithic executable, it's not a biggie. But if you're writing a flock of tools tied together by scripts and web pages, it's a real killer.
Further down the list should be the usual library incompatibilities. But we expect those, and know how to add code to our Configure script, and how to put #ifdef lines in the C code.
If only we could get find(1) to work sanely. Sometimes it does; sometimes it spawns one or a flock of shells that eat up all the cpu and memory. We suspect it's due to file-name problems, and we've found a few of them. But it keeps biting us, even with bare find commands with no aliases or scripts involved. It's very difficult to diagnose and fix this sort of problem.
We're finding that every one of our shell scripts needs to be rewritten, even 20-year-old Bourne-shell scripts. This is not a pleasant thing to have to report about what is otherwise an excellent machine to work on.
7. Double click on Terminal.
Wrong. Drag Terminal to the dock, and then click on it there.
That way, you'll never have to do steps 1-6 again.
Oh, and you need to know that clicking on the Terminal icon a second time won't get you a second terminal. You need to go to a terminal window and use CMD-N, where CMD is that funny key with an Apple logo and/or a four-leaf-clover icon.
And Terminal is not all that much like an xterm, so you have a minor learning experience ahead. But it's not too bad.
Everyone is comfortable with windows, even if they don't like it.
Nonsense. First off, there are a lot of us who grew up with other kinds of systems. And for a second example, my wife just got herself a new Mac Powerbook. She has worked with DOS and Windows at work since they first came out, and should be as comfortable with them as anyone. She has more and more been saying how much she hates Microsoft, and explains that it's from long experience.
Her comments on the Mac? By the end of the first week, she was saying how much she loves having a computer that works. She carries it everywhere, even in her tennis bag. (And she complains about poor wireless coverage. "Here in the Boston area!" Two weeks ago, she had never used wireless networking.)
Yes, she has stumbled across the things that work differently. But she points out that every new version of Windows has the same kinds of incompatibilities. And the Mac's behavior makes a lot of sense to her, unlike all the things in Windows that are still baffling after all those years.
Yesterday, after she had complained about several baffling things that IE was doing, and commenting that the Safari that I'd installed wasn't all that much better, I spent maybe 3 minutes downloading and installing mozilla 1.4. After an hour of playing with it, she decided to trash IE and Safari, because mozilla was already so much better and easier to use. She'd already downloaded and tried out a couple of themes, set up the mail reader, and so on, which can be several days work on a new Windows box. She'd also primed the mail reader's junk-mail handler and watched it correctly classify several new pieces of spam. All within an hour.
But she doesn't much like the single titlebar across the top. Neither do I.
i'd have to play with mac in non-critical settings to be familiar with the platform to form my own opinion and figure out the gotchas.
There are some very serious gotchas. About a month ago, I copied all of a project to OSX to test portability problems. Nearly all the scripts, no matter what language, went totally berserk. It took a lot of testing to diagnose the cause of "impossible" behavior. It seems that there's some strange sort of case-insensitivity in a lot of OSX, perhaps in the file system itself. Here's a demo from a test box:
: mkdir tmp; cd tmp
: touch FooBar foobar; ls -l
total 0
-rw-r--r-- 1 jc staff 0 Aug 16 09:48 FooBar
: rm *
: touch foobar FooBar; ls -l
total 0
-rw-r--r-- 1 jc staff 0 Aug 16 09:49 foobar
:
So you can't create two files whose names differ only by capitalization. This shoots down much of the naming in this project. What happens seems to be the worst case: The program that attempts to create the second file succeeds as if the file doesn't exist, but it overwrites the first file.
After discovering this, I asked (with an incredulous tone of writing) in a Mac newsgroup, and got verification. After a few weeks of testing, I still don't understand the rules, but they aren't like any other unix/posix system I've dealt with.
This in itself is a showstopper for this project. Someone suggested that we could install a different file system that doesn't do this, but of course we can't ask customers to do that.
This also applies to the search path. One of the things we'd relied on was that the system libraries are pretty much full of things (executables, functions) whose names are all lower case. So we can pretty much avoid collisions with "system" things by using one or two upper-case letters in names. But this doesn't work on OSX. If your program execs FooBar or fooBar, and there's a foobar earlier in the search path, then foobar is what runs.
Hunting down such name conflicts and fixing them can be incredibly time consuming. The symptoms are bizarre in the extreme.
Now, one could argue that a lot of users like case insensitivity, and an OS can do whatever its builders like. True. But programmers need to know the rules, or they can't write software that works. In this case, the rules have changed from something simple and easy to understand, to something that (from a programmer's viewpoint) is complex and mysterious. File-name matching tests have results that are difficult to explain or predict. The rules don't seem to be documented anywhere that we can find. So software won't port to OSX easily, and software development constantly stumbles on code that gets (or writes) the wrong file. And since the search path is affected, scripts work very differently for a user with a different search path than those used by the developers.
Bad decisions like the one made by Lindows to run as root be default can lead to Linux having as bad a reputation as Microsoft.
... but ... Lindows is supposed to be emulating Windows. Running as root is required to do this job. After all, imagine the criticisms from Windows users if they didn't do this.
But
"What? I can't overwrite any file on the system unless I give some stupid root password (whatever that is)? They told me this would work just like Windows. They lied to me. I want my money back."
MS's management has told us and told us: Windows users don't want security. They want capability. So give them the power to do what they want. It's just those outdated unix systems that limit what their users can do to their own machine.
If SCO's claims are without merit, then they have placed themselves at a huge risk of a substantial judgment against them. Of all our sakes, I hope that this is the case.
Well, lotsa luck collecting. By the time the case is decided, SCO's officers will be sitting pretty in Rio, and the company will be bankrupt with a debt of a billion or so. How do you think you'll ever get any blood out of that turnip?
Exchanging illegal content over the internet is an illegal act, ...
...
... Maybe I oughta teach this technique to a few of the more oppressive governments. They can set up "public information" pages that preload classified docs into your cache.
I'm always interested in how an unscrupulous web site could take advantage of such laws. This one's trivial. If you make the mistake of visiting one of my sites with javascript enabled, I can send you to jail quite easily.
One of my js demos is of image "preloading". It's useful: While you look at my main page, JS is silently downloading the images from the rest of my pages, so that subsequent pages come up faster. But I don't have to preload images from my site; I can use an arbitrary URL and "preload" any file from any site. And none of my pages has to ever use them. Think of this for a few seconds
In much of the world, having images on your machine like the goatse pic or any random pronography is illegal. So what my web page does is "preload" a whole flock of these into your browser cache. My other pages don't actually reference them, but that's not a problem. The problem comes when I use my subpoena power to find out who has been caught by my JS and now has the images on their disk. I send a tipoff to the local authorities. You get a knock on your door, they examine the directories that my note told them about, and sure enough, the images are there. So it's off to jail with you.
Or I could use the same thing to "preload" a few commercial MP3's into your browser's cache. Then I tip off the recording company that you've downloaded them and have them on your disk.
Hmmm
Needless to say, I like to turn off JS, Active-X, and other scripting goodies before I use any browser. Or I use a browser that doesn't have them in the first place. But most users just use IE or maybe Netscape out of the box, and don't know how to configure such things. So this law, like the DMCA, makes them vulnerable to just this sort of trick.
...will be wasted if you cannot defend your property in court.
But it will probably never happen. There have been any number of explanation from lawyers about why it's unlikely.
The basic scenario is that management tells their lawyers to prepare for a lawsuit challenging some GPL'd software that they want to convert to a proprietary product. The lawyers say "You have a choice here. You can obey the GPL now. Or you can challenge it, go to court, lose, and obey the GPL afterwards."
And, of course, one basic probblem is that if you challenge the GPL, until the case is settled, your rights to the software become the default rights that you have to any copyrighted material: You have no rights, other than what the copyright owner gives you. Since you've filed suit against the owner, they aren't going to give you anything. So for the duration of your court case, you can't use the software at all.
This is usually clear enough to convince even the densest PHB that maybe challenging the GPL isn't in the company's best interest.
SCO's lawyers are certainly smart and knowledgeable enough to understand all this. They have no intention of carrying a court case through to a decision. They are merely trying for maximum damage (and publicity) before they have to back out.
In Germany, this happened very quickly. In the US, it'll take years.
If anyone is still using plaintext to send passwords over their lan they are insane.
Well, a lot of people don't have any choice. Our cable ISP here, for example, provided the usual email accounts, and for a lot of customers, that is their only email. If you use it, you have no choice other than POP, and I haven't seen anything in several mailers that talks about encrypting the passwords. Our ISP doesn't actually block port 25, so you could run your own mailer. This isn't feasible for most customers, though, for several reasons. One is the dynamic IP addresses and insane hostnames. I've fixed that by using one of the many independent registration services, but to most customers, that would be utterly baffling and unusable. Another problem is that running your own email server is in fact in violation of the TOS in the ISP's contract, and they can legally block your port(s) or kick you off entirely at any time, without warning or recourse.
So for most non-geek customers, unencrypted POP passwords are the only option. There's probably no way they could even learn from the ISP that there's a problem; they certainly wouldn't get (or understand) any advice on how to fix it.
(Myself, I use an account at a school. It has been stable and usable for over 15 years now, unlike commercial email accounts that force you to change your address every 6 months whenever there's a merger, buyout, or corporate renaming. And I can use a plain-text mail reader, eliminating all problems with virii, worms and the like. But I'm not sure I'd recommend this to the typical non-geek.)
What I did say was that NO ONE PAYS ATTENTION to this stuff. No one reads anything outside of the sports section and the front page. No one is scouring the net looking for what information they can find about what is going on "behind the scenes".
/. readers do.
Um, it appears that you do. And I do. And, from the messages here, it's obvious that at least a few
Politics isn't much influenced by the masses that don't pay attention. It's influenced by the activists of all sorts who get out there and do something.
Of course, making a reasonable, informed argument in a forum like this may seem like spitting into the wind. But you're doing something.
So you can really claim that NO ONE is paying attention. The question is whether there are enough of us to make a difference.
I have said this a billion times before. No one outside of our geek culture has any idea what this is.
And I've told you a billion times: Don't exaggerate!
There have been occasional suggestions that anything that goes out of production (or support, for computers) for a year becomes public domain. This is usually suggested in terms of all those old movies that are rotting in company vaults while they sit on the copyright and refuse any public access to them.
...
It would make a lot of sense to lobby for such a copyright law for software. It would drag a lot of companies (kicking and screaming) into line with the public interest.
So get into the politics of it
Well, around these parts (Bush/Cheney America), you're expected to do the proper, conservative, market-driven thing: You offer to sell it to the highest bidder.
Anything that doesn't earn you a good profit is a dead giveaway that you're one of those leftist, socialistic types. Probably a terrorist. In that case, you should report yourself to the Dept of Homeland Security.