The Portland Pattern Repository is a fascinating reference site, even if you're not a programmer. It's also of interest as the very first WikiWiki site.
Then again, it's sort of redundant to mention the PPR in this story, since it tends to turn up in Google searches.
Bartleby.com has all kinds of reference books available, and lots of other stuff online. Like all of the Harvard Classics.
I'm suprised I'm the first to mention Project Gutenberg. Maybe people don't think of it as a reference site, since mostly it's good for downloading whole public-domain texts. But various volunteers (including me) are now in the process of proofreading the famous Britannica 11th Edition for inclusion in the Gutenberg set. At the current rate, it'll be done in about a year. But perhaps with more volunteers...
You know, flame all you want but we all know everyone started in this programming language.
Say what? You mean every programmer started out hacking interpreted DOS programs? There were no programmers before DOS was invented? Nobody ever started out with Turbo Pascal? People continued to start out with QBasic even after Windows took over and MS stopped providing QBasic by default? Well, OK.
Actually, the current USPS does a very good job, all things considered. They have every right to resent the term "snail mail", though they're probably stuck with it.
But it hasn't always been this way. Horror stories about lost mail and general screwups were legion. UPS wouldn't even be in business if the Parcel Post hadn't been a complete disaster for years and years.
I'm old enough to remember when the USPS was the federal Post Office Department and the Postmaster-General was appointed by the President. The job always went to the chairman of the party in power, because of all the patronage jobs. Needless to say, efficiency and quality of service were not top priorities. Most of the USPS's unfortunate reputation dates from this period.
Good point. But I doubt if Scott is going to rush to the jeweler. Sun has spent a fortune developing and promoting Java, and on failed Java-based businesses. It'll take a lot more than a few certification fees for the whole thing to show anything like a profit!
The cert. hassles might have helped raise tempers. But leaving JBoss won't make it any easier for these guys to get their code certified. Well, maybe a little, if you assume the working relationship between Fleury and his developers went sour. Java certification is a pretty controversial area right now, and there's a lot to suggest that Sun's process is pretty arbitrary.
The Slashdot story and the Blog buzz all say this is about a bunch of developers wanting their own JBoss fork. The Core Developers party line is that they just don't want to be tied to JBoss to the exclusion of competing products. Any other explanations?
To be fair, Fleury was a little muddled about that point. Perhaps he was reluctant to come right out and admit that he has no legal hold on the Core people unless the infringe the JBoss trademark.
It's interesting to note that the trademark currently belongs to Fleury personally. The USPTO trademark database he registered it a couple years ago. Before that, it belonged to a California company called Telkel, which appears to be defunct.
Oh wait, Fleury registered the trademark in 2001, but cites "first use" a year earlier. This is perfectly legitimate (its use that establishes your ownership of a trademark, not registration) except that in 2000 Telkel was still in business. IANAL, but I'd guess that the Core people could challenge Fleury's ownership of the trademark if they wanted to. Probably not worth the hassle though.
It's also interesting to note that the jboss.org domain belongs to JBoss Group LLC, "a privately-held services company". Not, as the.org suggest, a voluntary alignment of open-source zealots.
You're describing hassles that all trademark owners go through. One of the requirements of maintaining a trademark is that you "actively enforce" it. The ads give them something to point to if anybody ever claims that "Xerox" is generic. To the best of my knowledge, nobody has.
Perhaps that's all there is to the Open Group/Apple lawsuit -- OG making noise so they can continue to claim that "Unix" is not generic.
Such a move is probably a necessary part of any policy. But that's not a solution in itself. Rules without enforcement are worse than useless. And severe penalties just make the joke a sad one.
Gee, here we are throwing nasty threats at our employees. What happened to trust?
You're right and wrong. From a technical point of view, it doesn't make any sense to deny that Linux or BSD is a flavor of Unix. But I don't see any logic to the claim that the Unix trademark has gone generic. I've never seen it used in official documents by anybody who hadn't licensed it. (Sun, HP, and all the other Unix biggies have licenses.) Even the online community usually avoids the issue by saying "*nix" or "Unix-like" instead of "Unix". (I don't, because terms like *nix offend my tech-writer's compulsive nitpickiness. But I'm definitely in a tiny minority.) As far as I can see, the Unix trademark is better enforced than such common trademarks as "Kleenex" and "Xerox", and there's no sign that they're in trouble.
On the other hand, it's not clear to me that Apple has violated the trademark. They are a little sloppy when they talk about OS X's Unix origins -- they really should make it clear that they have no license for the Unix trademark -- but it's perfectly legitimate for them to claim that OS X is derived from Unix.
Really this is about the Open Group struggling to hang onto the shreds of its dwinding relevence. Sun and HP still go through the motions of certifying their right to use the Unix trademark, but they don't make a big thing about it. And Linux continues to eat into the Unix marketplace, even though it isn't certified as compliant with any Unix specification. It probably could be, if anybody were willing to spend the money. But nobody is, and nobody cares -- which is bad for Open Group.
I think I agree with the attitude you're expressing. But you're kind of oversimplifying the issue.
In a really, well-run company, the CIO will tell the CEO, "we have a problem with rogue APs". The CEO tells the VPs, who tell the department managers. The managers bring it up in department meetings. Because the managers have good working relationships with all their subordinates, they figure out who has APs and which ones need to be hardened. Problem solved, and no Big Brother nonsense necessary.
In the real world, no company is that well run. This manager or VP doesn't get along with his or her subordinates. That one is a control freak. This employee doesn't see what the big deal is, and won't let anybody look at his AP. That one never goes to department meetings, doesn't take orders from anybody, and has so much seniority that...
Oops, the trauma of my last job is showing! Point is, not all problems end up being solved by management/worker trust and collaboration. It's certainly desireable that you solve as many problems that way as you can. But there's always something you end up having to enforce with rules and snooping, and other nasty stuff. When that sort of thing gets out of hand, the company is probably in deep trouble. But you always have to deal with some of it.
They're not sitting on them. I'll bet all those IP numbers are in use, and probably there are LA County employees crying for more. Why didn't they notice that their numbers had been hijacked? Presumably they have a firewall that cuts off their address space from the rest of the Internet.
I've always thought it was dumb that public IP numbers are so widely used for networks for which public access is not only unnecessary, but actually avoided. Why spend zillions on firewall software when you can get the same effect just by using a private IP space? I guess the changeover costs are a killer.
Let's see, you've got a color display (which probably isn't viewable unless you have the backlight on), a 200 mhz processor, and a fairly powerful microwave transceiver. You're going to run down that battery awfully fast! Perhaps there's no SSH client because this is just the wrong platform for it.
OK, I'm not a big expert, but isn't the XML behind SOAP pretty well nailed down? Tools like XMLSpy are for designing new XML applications, not for implementing ones that are already well-defined. For that, you want specialized tools. I know my former employer, Borland, bundles SOAP components and design tools with several of its IDEs. Can't comment on their quality -- never worked in that area.
XML is a big area. Are you using it for data transfer (they don't call it "the ASCII of the Internet for nothing!) or to author content? And what kind of data or content?
If you're engineering a lot of schemas, transforms, and other XML weirdnesses, XMLSpy is probably the most cost-effective tool there is, despite its high cost. But I've always found it inadequate for content editing. Which is my interest, since I'm a tech writer.
Everybody with a little Java knowledge and access to a component library has written a half-assed XML content editor. But only two editing tools seem to be worth bothering with: XMetal and FrameMaker.
XMetal is very well designed. I was able to make it grok Docbook just by pointing it a the correct DTD. Unfortunately, it's Windows only. And it now belongs to Corel, which doesn't make one optimistic about its future.
Adobe now provides the "structured" version of FrameMaker for no extra charge when you buy the regular SKU. And FrameMaker runs on Windows and most Unixes. (Not Linux, alas.) Problems: the worst UI design in the universe. And defining new XML applications is a nightmare.
Yeah, if I had a choice, I'd choose Fido over T-Mobile. I don't really have a choice, unless my cell phone to have a Canadian area code!
The point of my calculations was to guess at how quickly US$1/M would add up, assuming your usage patterns are typical. That is a lot of money for just one megabyte. But then again GPRS is a pretty slow channel, and you'd probably want to browse in text only mode anyway. On that scale, 1M is a lot of content.
But some of the professors only want you to post to the public discussion groups and never have you post to the private portfolio, basically this means they don't have to do anything accept scan the conferences and give out more assignments. They don't have to look over your work and give you any feedback. I bet it takes less than an hour a week to do this.
I've run into this myself. But not in online courses -- in evening classes offered by a University Extension.
I think Extensions and online schools have a certain tendency to attract both bad teachers and bad students. Bad teachers tend to be working professionals who teach part time for a few extra bucks. Since they don't have an academic career, they feel free to pull all kinds of crap.
Bad students are people who are doing the "continuing education" bit at their employer's expense. Now, most people who do this take it seriously, but there are a lot who are just going through the motions, and do the minimum amount of work needed to pass the class. For obvious reasons, bad teachers and bad students quickly come to an understanding, and the whole class suffers as a result.
I'm convinced that a lot of "continuing" and online institutions tolerate this -- they just want the tuition. But if your institution has a lot of teachers who really care about what they're doing, you have to assume that the school cares too. Which means you have to make a lot of noise about any bad teachers you encounter. If nothing else, you'll get a tuition refund for the classes you've been wasted. Helps to document everything they do wrong.
I own a headset for my cell, but I never use it. All that cable is too bulky for my pocket. And it gets tangled up so easily. And it's a pain to stop and plug it in. Now, a Bluetooth headset, I could just stick it in my pocket with cell, than put it on when the phone rings.
I'd also like to share data between my cell and Palm. Have a cable for it, but it's too cumbersome. Might be worth the trouble to sync my phone books, but it's too awkard for web browsing. So I use the ridiculous WML browser built into my phone. Now if both my phone and my Palm had Bluetooth.
I probably will never get a bluetooth card for my laptop. Then again, I could share data with other people at a conferance table. Or print. Yeah, I can do that now -- have to mess with an Ethernet cable, make sure my DHCP client is configured for the local network, hope there's a hub slot free, hope the dongle on my Ethernet card doesn't break again, hope everybody I need to share data with is also properly set up...
Well, you have a point -- if you assume that ink or gestures is the only way to interact with a tablet.
I'm sort of physically impaired, so I'm not a good example. But my experience with ink and gestures on the PDA suggests that not everybody finds them productive.
I started out with a Newton. Of course, early Newtons had terrible handwriting recognition. Which they more or less fixed -- about the time Jobs pulled the plug on the product. But even if I had one of the later Newton's, I'm not sure I'd rely on handwriting input that much -- for me, it'd still be too slow and unreliable.
Similar issues with gesture-based inputs, like Graffiti and Jot. Once you get past the initial learning phase, they're a lot easier to use than handwriting input. But you have to have really good coordination to use them with any speed or accuracy. And I don't have that.
Which is why I use Fitaly Stamp on my Palm. Shouldn't be that hard to implement on Linux.
It's true that the article doesn't have any actual Java code (and is generally very unclear!). But it does talk about class files. Javascript doesn't use class files. Java does.
Given the flakiness of the source, and the absence of really good examples, I suspect this whole thing is bogus. It sounds like something that might work with the Java 1.0 security model, but not with the current version. Perhaps this guy noticed a change in Javascript implementations, and jumped to the conclusion that it re-opened an old security hole, not realizing it had been plugged elsewhere.
First prescription drugs, and now this! I don't know any U.S. provider that does flat-rate GPRS.
If I'm calculating this right, you often transfer around 1M an hour, which would cost C$50 (US$36) if you didn't have a flat-rate plan. OK, T-Mobile has a GPRS plan that costs US$20 (forgive me if I stop converting) for the first 5M and US$5 for each addition 1M. Quite a bit cheaper!
Another difference in the market: you can get a T68i for US$99 if you sign up for T-Mobile through Amazon. Also a Nokia 3650 for free. Curiously, these phones are not available from the T-Mobile web site.
A Bluetooth adapter for a PDA goes for about US$100. Not free, but not unreasonable if you already own a compatible PDA (which I do) and would be spending Us$30-40 for GPRS anyway.
Slashdot Mirror
Then again, it's sort of redundant to mention the PPR in this story, since it tends to turn up in Google searches.
Bartleby.com has all kinds of reference books available, and lots of other stuff online. Like all of the Harvard Classics.
I'm suprised I'm the first to mention Project Gutenberg. Maybe people don't think of it as a reference site, since mostly it's good for downloading whole public-domain texts. But various volunteers (including me) are now in the process of proofreading the famous Britannica 11th Edition for inclusion in the Gutenberg set. At the current rate, it'll be done in about a year. But perhaps with more volunteers...
So Fleury founded both Telkel and JBoss Group? That's interesting, since Telkel was in California and JG is in Georgia.
But it hasn't always been this way. Horror stories about lost mail and general screwups were legion. UPS wouldn't even be in business if the Parcel Post hadn't been a complete disaster for years and years.
I'm old enough to remember when the USPS was the federal Post Office Department and the Postmaster-General was appointed by the President. The job always went to the chairman of the party in power, because of all the patronage jobs. Needless to say, efficiency and quality of service were not top priorities. Most of the USPS's unfortunate reputation dates from this period.
Good point. But I doubt if Scott is going to rush to the jeweler. Sun has spent a fortune developing and promoting Java, and on failed Java-based businesses. It'll take a lot more than a few certification fees for the whole thing to show anything like a profit!
The Slashdot story and the Blog buzz all say this is about a bunch of developers wanting their own JBoss fork. The Core Developers party line is that they just don't want to be tied to JBoss to the exclusion of competing products. Any other explanations?
Probably there's some truth to all these stories.
It's interesting to note that the trademark currently belongs to Fleury personally. The USPTO trademark database he registered it a couple years ago. Before that, it belonged to a California company called Telkel, which appears to be defunct.
Oh wait, Fleury registered the trademark in 2001, but cites "first use" a year earlier. This is perfectly legitimate (its use that establishes your ownership of a trademark, not registration) except that in 2000 Telkel was still in business. IANAL, but I'd guess that the Core people could challenge Fleury's ownership of the trademark if they wanted to. Probably not worth the hassle though.
It's also interesting to note that the jboss.org domain belongs to JBoss Group LLC, "a privately-held services company". Not, as the .org suggest, a voluntary alignment of open-source zealots.
Perhaps that's all there is to the Open Group/Apple lawsuit -- OG making noise so they can continue to claim that "Unix" is not generic.
Gee, here we are throwing nasty threats at our employees. What happened to trust?
On the other hand, it's not clear to me that Apple has violated the trademark. They are a little sloppy when they talk about OS X's Unix origins -- they really should make it clear that they have no license for the Unix trademark -- but it's perfectly legitimate for them to claim that OS X is derived from Unix.
Really this is about the Open Group struggling to hang onto the shreds of its dwinding relevence. Sun and HP still go through the motions of certifying their right to use the Unix trademark, but they don't make a big thing about it. And Linux continues to eat into the Unix marketplace, even though it isn't certified as compliant with any Unix specification. It probably could be, if anybody were willing to spend the money. But nobody is, and nobody cares -- which is bad for Open Group.
In a really, well-run company, the CIO will tell the CEO, "we have a problem with rogue APs". The CEO tells the VPs, who tell the department managers. The managers bring it up in department meetings. Because the managers have good working relationships with all their subordinates, they figure out who has APs and which ones need to be hardened. Problem solved, and no Big Brother nonsense necessary.
In the real world, no company is that well run. This manager or VP doesn't get along with his or her subordinates. That one is a control freak. This employee doesn't see what the big deal is, and won't let anybody look at his AP. That one never goes to department meetings, doesn't take orders from anybody, and has so much seniority that...
Oops, the trauma of my last job is showing! Point is, not all problems end up being solved by management/worker trust and collaboration. It's certainly desireable that you solve as many problems that way as you can. But there's always something you end up having to enforce with rules and snooping, and other nasty stuff. When that sort of thing gets out of hand, the company is probably in deep trouble. But you always have to deal with some of it.
A lot of which is covered in these articles, none of which Simonker bothered to link. Slashdot editors are getting lazier and lazier.
I've always thought it was dumb that public IP numbers are so widely used for networks for which public access is not only unnecessary, but actually avoided. Why spend zillions on firewall software when you can get the same effect just by using a private IP space? I guess the changeover costs are a killer.
Let's see, you've got a color display (which probably isn't viewable unless you have the backlight on), a 200 mhz processor, and a fairly powerful microwave transceiver. You're going to run down that battery awfully fast! Perhaps there's no SSH client because this is just the wrong platform for it.
Ah, ignorance. It gives you such a superior feeling! I quite miss it.
OK, I'm not a big expert, but isn't the XML behind SOAP pretty well nailed down? Tools like XMLSpy are for designing new XML applications, not for implementing ones that are already well-defined. For that, you want specialized tools. I know my former employer, Borland, bundles SOAP components and design tools with several of its IDEs. Can't comment on their quality -- never worked in that area.
If you're engineering a lot of schemas, transforms, and other XML weirdnesses, XMLSpy is probably the most cost-effective tool there is, despite its high cost. But I've always found it inadequate for content editing. Which is my interest, since I'm a tech writer.
Everybody with a little Java knowledge and access to a component library has written a half-assed XML content editor. But only two editing tools seem to be worth bothering with: XMetal and FrameMaker.
XMetal is very well designed. I was able to make it grok Docbook just by pointing it a the correct DTD. Unfortunately, it's Windows only. And it now belongs to Corel, which doesn't make one optimistic about its future.
Adobe now provides the "structured" version of FrameMaker for no extra charge when you buy the regular SKU. And FrameMaker runs on Windows and most Unixes. (Not Linux, alas.) Problems: the worst UI design in the universe. And defining new XML applications is a nightmare.
The point of my calculations was to guess at how quickly US$1/M would add up, assuming your usage patterns are typical. That is a lot of money for just one megabyte. But then again GPRS is a pretty slow channel, and you'd probably want to browse in text only mode anyway. On that scale, 1M is a lot of content.
I think Extensions and online schools have a certain tendency to attract both bad teachers and bad students. Bad teachers tend to be working professionals who teach part time for a few extra bucks. Since they don't have an academic career, they feel free to pull all kinds of crap.
Bad students are people who are doing the "continuing education" bit at their employer's expense. Now, most people who do this take it seriously, but there are a lot who are just going through the motions, and do the minimum amount of work needed to pass the class. For obvious reasons, bad teachers and bad students quickly come to an understanding, and the whole class suffers as a result.
I'm convinced that a lot of "continuing" and online institutions tolerate this -- they just want the tuition. But if your institution has a lot of teachers who really care about what they're doing, you have to assume that the school cares too. Which means you have to make a lot of noise about any bad teachers you encounter. If nothing else, you'll get a tuition refund for the classes you've been wasted. Helps to document everything they do wrong.
He's just pining for the fjords!
I'd also like to share data between my cell and Palm. Have a cable for it, but it's too cumbersome. Might be worth the trouble to sync my phone books, but it's too awkard for web browsing. So I use the ridiculous WML browser built into my phone. Now if both my phone and my Palm had Bluetooth.
I probably will never get a bluetooth card for my laptop. Then again, I could share data with other people at a conferance table. Or print. Yeah, I can do that now -- have to mess with an Ethernet cable, make sure my DHCP client is configured for the local network, hope there's a hub slot free, hope the dongle on my Ethernet card doesn't break again, hope everybody I need to share data with is also properly set up...
I'm sort of physically impaired, so I'm not a good example. But my experience with ink and gestures on the PDA suggests that not everybody finds them productive.
I started out with a Newton. Of course, early Newtons had terrible handwriting recognition. Which they more or less fixed -- about the time Jobs pulled the plug on the product. But even if I had one of the later Newton's, I'm not sure I'd rely on handwriting input that much -- for me, it'd still be too slow and unreliable.
Similar issues with gesture-based inputs, like Graffiti and Jot. Once you get past the initial learning phase, they're a lot easier to use than handwriting input. But you have to have really good coordination to use them with any speed or accuracy. And I don't have that.
Which is why I use Fitaly Stamp on my Palm. Shouldn't be that hard to implement on Linux.
Given the flakiness of the source, and the absence of really good examples, I suspect this whole thing is bogus. It sounds like something that might work with the Java 1.0 security model, but not with the current version. Perhaps this guy noticed a change in Javascript implementations, and jumped to the conclusion that it re-opened an old security hole, not realizing it had been plugged elsewhere.
If I'm calculating this right, you often transfer around 1M an hour, which would cost C$50 (US$36) if you didn't have a flat-rate plan. OK, T-Mobile has a GPRS plan that costs US$20 (forgive me if I stop converting) for the first 5M and US$5 for each addition 1M. Quite a bit cheaper!
Another difference in the market: you can get a T68i for US$99 if you sign up for T-Mobile through Amazon. Also a Nokia 3650 for free. Curiously, these phones are not available from the T-Mobile web site.
A Bluetooth adapter for a PDA goes for about US$100. Not free, but not unreasonable if you already own a compatible PDA (which I do) and would be spending Us$30-40 for GPRS anyway.