The point is that the page has not been removed/altered/etc. and that before modding accusatory messages up, people should take 5 seconds and check the link.
Exactly. I'm no expert on PCI bus operations but if they aren't that complex, then the binaries won't differ much.
How far should this go? I mean if there's a common function (say swap(x, y) for a overy simple example) in a piece of GPL code, am I not allowed to write my own swap function with the same name and signature? And considering that there aren't that many ways to implement it, am I obligated to avoid an implementation that produces the same object code?
They're not releasing that piece of source as part of some GPL requirements, they're releasing to show that it's not covered by the GPL. From the press release (empahsis mine):
For the avoidance of doubt, the hardware abstraction layer (roughly
analogous to a PC's BIOS) has it's PCI allocation and bridge setup
based in part on the following functions from the Linux kernel sources
I admit that it could probably be worded better, but it sounds like they could have took the function names/possibly signatures and wrote their own code. Get the source and find out. However, if the experts in this matter can still show that the object form is too close to the GPL output, then there may be something to worry about.
Since email has no geographic reference point associated with it, it's pretty hard to force someone sending email to pay for several different drop lists from different states just in case someone they're mailing is on one of them.
Aldebaran's great, OK, Algol's pretty neat, Betelgeuse's pretty girls, Will knock you off your feet. They'll do anything you like, Real fast and then real slow, But if you have to take me apart to get me there, Then I don't want to go.
Singing, Take me apart, take me apart, What a way to roam, And if you have to take me apart to get me there, I'd rather stay at home.
Sirius is paved with gold So I've heard it said By nuts who then go on to say "See Tau before you're dead." I'll gladly take the high road Or even take the low, But if you have to take me apart to get me there, Then I, for one, won't go.
Singing, Take me apart, take me apart, You must be off your head, And if you try to take me apart to get me there, I'll stay right here in bed.
- Douglas Adams "The Restaurant at the End of the Universe"
I don't think they're trying to lock anyone in, I just think they realize that generating thousands or millions of INSERT statements is a really crappy way to move a database. Using a specialized bulk extract/import tool is usually a lot easier and a lot faster.
All of the major database have utilities to import/export data in user-defined formats, so I can't see that Oracle not providing an instant "generate sql script" command is keeping anyone from moving anything off of Oracle. Besides if you really want to do it, you can get free Toad. Works great.
The new move, however, could alienate companies, including SCO customers, according to Giga Information Group analyst Stacey Quandt.
You think so? I mean if I were evaluating technology to purchase, I know I always want to select the company that appears to be ready to sue anything that moves since that's the only way they're going to survive.
</sarcasm>
Most of the code in the NT kernel is already built from BSD code.
Ummm, no.
The kernel is not BSD-based. It was written with the help of DEC (who really wanted a more "mainstream" OS for the Alpha platform), and is much more VMS-like than anything.
Now, the original TCP/IP stack and utilites were based on BSD, so in that area it's been BSD-like.
Also, I think many Mac users would disagree with the on the availability of drivers for OS X.
That's the kind of response I was looking for in the first place. I'll admit that I do like CE, and I think it's the best OS (possibly even product) that MS has produced. I've also had good luck with it in my use.
Out of curiosity, do you remember what API it was that had that behavior? Want to make sure I watch out for it.
WinCE (for automotive, whatever...) is certainly the worst choice. QNX, VxWorksAE, or Embedded(orRT)Linux would certainly be better - but the fundamental problem remains - this is HARD to get right.
So, why is CE the worst choice? This is not trolling, I really want to know. I've used several RTOS's, including CE, and can't think of anything that would necessarily impair it vs. the others you mentioned.
You mean like them putting a file on the net with a spiked URL in one of the id3 tags that goes to a server that is logging when it gets a hit to that URL? So when your new cool player tries to pull up that page it's looging your IP?
Searched, found registration 2623790 for trademark PCI owned by PCI-SIG. First line on the registration details under the logo:
Word Mark PCI
So, yes, they do have a registered trademark on the word PCI. To register a work mark, it doesn't need to be universally distinct, just distinct enough in the particular class you register it in.
No, the MS JVM (1.1.4 vintage) runs great on Windows. It's the Sun JVM that runs like crap. Of course it's faster on Windows than most of Sun's boxes. Go figure.
PCI is their registered trademark. They have a right to restrict its use in the areas for which it's been registered.
However, I'm sure they could care less if people reference PCI in the course of descriptions, discussions, etc. I think they had a problem with potential (as unlikely as it is) confusion with his site and an official PCI-SIG source of information.
A simple disclaimer on his site would have probable been fine. But he chose to kill the site instead.
To: Alan.Deikman@znyx.com
Cc: mcohen@schwabe.com
Subject: I can't believe you guys shut down the free PCI device table!!!
I'll bet he can't believe it either. Especially since all they wanted was the logo removed and it was Jim that decided to take down the site in protest.
Actually it is. They had it registered in September. So, all he really had to do was remove logo, change the references to PCI(R) (posting thing strips out entities appearently), and put a disclaimer that PCI is a registered trademark of PCI-SIG and that this site is not affiliated with or endorsed by PCI-SIG.
That probably would have cleared everything up. But it's much more fun to have a fit.
Slashdot Mirror
I still don't think we've got a conspiracy here.
The point is that the page has not been removed/altered/etc. and that before modding accusatory messages up, people should take 5 seconds and check the link.
How far should this go? I mean if there's a common function (say swap(x, y) for a overy simple example) in a piece of GPL code, am I not allowed to write my own swap function with the same name and signature? And considering that there aren't that many ways to implement it, am I obligated to avoid an implementation that produces the same object code?
Parent is score 3, Informative when it's completely false? It's not that hard to paste the link and try it!
Or, maybe, they saw your post and then quickly put it back...
For the avoidance of doubt, the hardware abstraction layer (roughly analogous to a PC's BIOS) has it's PCI allocation and bridge setup based in part on the following functions from the Linux kernel sources
I admit that it could probably be worded better, but it sounds like they could have took the function names/possibly signatures and wrote their own code. Get the source and find out. However, if the experts in this matter can still show that the object form is too close to the GPL output, then there may be something to worry about.
Aldebaran's great, OK,
Algol's pretty neat,
Betelgeuse's pretty girls,
Will knock you off your feet.
They'll do anything you like,
Real fast and then real slow,
But if you have to take me apart to get me there,
Then I don't want to go.
Singing, Take me apart, take me apart,
What a way to roam,
And if you have to take me apart to get me there,
I'd rather stay at home.
Sirius is paved with gold
So I've heard it said
By nuts who then go on to say
"See Tau before you're dead."
I'll gladly take the high road
Or even take the low,
But if you have to take me apart to get me there,
Then I, for one, won't go.
Singing,
Take me apart, take me apart, You must be off your head,
And if you try to take me apart to get me there,
I'll stay right here in bed.
- Douglas Adams "The Restaurant at the End of the Universe"
Archive date: May 11, 1996
http://web.archive.org/web/19960511013133/http://w ww.altavista.digital.com/
Check out that graphical navbar at the top.
http://web.archive.org/web/19961030020512/www.berk eley.edu/about/
All of the major database have utilities to import/export data in user-defined formats, so I can't see that Oracle not providing an instant "generate sql script" command is keeping anyone from moving anything off of Oracle. Besides if you really want to do it, you can get free Toad. Works great.
You think so? I mean if I were evaluating technology to purchase, I know I always want to select the company that appears to be ready to sue anything that moves since that's the only way they're going to survive.
</sarcasm>
Ummm, no.
The kernel is not BSD-based. It was written with the help of DEC (who really wanted a more "mainstream" OS for the Alpha platform), and is much more VMS-like than anything.
Now, the original TCP/IP stack and utilites were based on BSD, so in that area it's been BSD-like.
Also, I think many Mac users would disagree with the on the availability of drivers for OS X.
Out of curiosity, do you remember what API it was that had that behavior? Want to make sure I watch out for it.
It's so obvious. How silly of me to ask the question.
You can get the source for Windows CE. Look here. Even the licensing terms aren't too horrible (for MS).
So, why is CE the worst choice? This is not trolling, I really want to know. I've used several RTOS's, including CE, and can't think of anything that would necessarily impair it vs. the others you mentioned.
Nahh, never happen...
Word Mark PCI
So, yes, they do have a registered trademark on the word PCI. To register a work mark, it doesn't need to be universally distinct, just distinct enough in the particular class you register it in.
However, I'm sure they could care less if people reference PCI in the course of descriptions, discussions, etc. I think they had a problem with potential (as unlikely as it is) confusion with his site and an official PCI-SIG source of information.
A simple disclaimer on his site would have probable been fine. But he chose to kill the site instead.
I'll bet he can't believe it either. Especially since all they wanted was the logo removed and it was Jim that decided to take down the site in protest.
That probably would have cleared everything up. But it's much more fun to have a fit.