'Exchange 12 administration functions will be built atop Monad, which would enable users to do everything from the command line that can be done from the graphical interface.'
2nd: The more time goes on, the more Windows takes on the features of unix.
3rd: Most every OS is some form of unix at this point except for Windows.
4th: Even Windows has a POSIX layer and unix-style command utilities for free as an add-on.
Re:the code of conduct for free software distribut
on
Drafting GPL3
·
· Score: 1
Probably. Other than the copyright notice, what's the difference between the BSD license and the public domain?
None as far as I can tell. I'd lump X11, BSD, and other similar licences into basically the same category basic category as public domain. I don't know why the other person thought there would be a big advantage to explicitly giving up all rights vs. giving up all but attribution and copyright announcement.
Re:the code of conduct for free software distribut
on
Drafting GPL3
·
· Score: 1
The beauty of the BSD license in this context is that the reference implementation can be mutated into any different implementation. So long as the intended invariants are maintained (proper ogg behavior, in this case) then derivations into proprietary instances should be allowed and encouraged, as a thriving ecosystem of competing implementations only helps the adoption of the format. The LGPL would only allow, for example, a non-free Ogg player to invoke the free LGPL library. Modifications to the library would have to be released as LGPL, which is a restriction that discourages forking of the reference implementation. (Forking in the case ofreference implementations is very desireable, including forking into non-free variants.)
True. I'll add one extra; usually the code but not the project name is granted under BSD style licences. Copyright isn't the same as trademark.
Because of that, if someone takes the code and mangles it to make an incompatable fork, the original authors can state that the fork does not follow the reference implementation and thus isn't *insert name here* compatable. They could force the fork to change any claims of being compatable with the original project. Usually doesn't happen.
Re:Does anyone else find it mildly strange....
on
Drafting GPL3
·
· Score: 1
Almost every time he writes something, I cringe. He's seen as (one of) the father(s) of the open-source movement, but his views are often far beyond even the mainstream OSS community, let alone the software community as a whole. RMS is to the OSS community like Ralph Nader is to Democrats, or Pat Buchanan is to Republicans: they sort of agree to some core tenets in concept, and don't want to be seen really disagreeing with him. However, they find themselves saying, "Well, yeah, but..." a lot of the time.
Well, I often feel the same way about RMS.
That said, like Nader, I'm occasionally forced to agree with his view of reality and say "The bastard got it right again!". I'm very glad he's around, even if I rarely find that he's 100% spot on.
Re:the code of conduct for free software distribut
on
Drafting GPL3
·
· Score: 4, Insightful
Well, quite frankly, BSD licensed software is not truly free -- what's the point in free software if you can turn it into a non-free product?
Don't get me wrong. BSD is a fine license all right, but nothing special compared to the protected freedom of GPL.
While I prefer the GPL over BSD (and similar licences) -- the GPL does not work well in all situations. For example, anything that requires a reference design that you want to be widely adopted. The Ogg codecs (Vorbis, Theora,...) for example.
All that is fine and dandy untill you get a curious user, like me, who either sees the ghost link (move the mouse over and your cursor and status bar will reflect there being a link) or views the HTML for some reason and sees it. But then again, I can just pull out tor or something if I get banned.
If someone is that inattentive, they get banned.
If you want implementation details -- and it looks like you indeed do -- I'll be glad to provide them to you for a fee. Are you that curious, or can you figure some of the basics out for yourself?
You make a robots.txt like with directories/pages they should not enter. You include links in pages that only a crawler would see normally. Only a crawler that ignored robots.txt or the exclusion in it would go those that trap pages and get banned.
To make this very clear; the links on the legitimate pages are not normally visible or say things like "." or "," or "This is a trap for sp@mmer$"...whatever. Color the text white on a white background. Dinner is served.
1. Create robots.txt, including references to the spam spider trap. Make sure that the legitimate references to normal pages are out numbered by a large margin.
2. When pages that could only be referenced in the spam spider trap are accessed, note the IP address.
3. Slowly respond or block connections from the originating IP address.
Bad guys are punished. Good guys are not. Low impact on system resources.
There's got to be a dozen filters out there that already do this. Anyone have experience using one?
Since when does failing to back up your hard drive make your system easier to hack into? If you're talking about them having poor data integrity that's one thing, but this doesn't seem to point to poor computer security.
(scratches head)
1. If you don't know what you had you don't know if what you have has been screwed with.
2. If you do get screwed with, it's critical to be able to restore from a known good system. Otherwise, game over; you have to rebuild from scratch and guess about what has/has not been compromised.
True, the access to MSH is not as convenient as it should be, but how difficult are those steps, really? I don't think that the procedure to get the files necessary is any more difficult than downloading, compiling, (searching for missing files), compiling again, (fixing miscellaneous inconsistincies) and then running a beta OSS program.
Your example is exaggerated, even for early versions of OSS. If it's too much of a pain, though, I figure that the developer needs to work out bugs as well so I wait for a later release. In either case, you can immediately do something about OSS while with Monad you have to agree to licences and wait.
As someone else posted, take a look at 4DOS/4NT to see how much Microsoft could have changed. The changes you mention are simple additions and not structural enhancements.
It's a pain to put in simple things like a timed pause or command seperation for @#!#$ sake; the main ways I've seen are running ping or other commands that have a timeout feature and use that. In Bash, it's 'sleep 5' for 5 seconds, 'sleep 3600' for an hour. To seperate commands in Bash, insert a ; between commands -- and the | still works!
Command.com vs. cmd.exe are very very similar from a UI perspective. cmd.exe has next to no substantial improvements. It's disapointing that they didn't start on this change with the introduction of NT.
Shells are in parts every bit as difficult to do right as GUIs. Some GUIs rightly adopt shell inovations suchas Konqueror's use of tab completion and wild cards, while shells can be improved upon in the GUI such as Konsole's Send Input To All Sessions.
It's not a mistake that bash (based on sh) has taken so many years to get to the place it is now.
That said, throwing money at the problem may not help if the wrong people are involved in the design. So far, it looks as if they have it right.
It will take years to iron out the problems and discover just what is good and what sucks about it. A staged default adoption is the right way to do it as long as they dump the silly mistakes that have never been corrected since DOS 2.x.
So far, it's still there. Very strange things going on though such as the number of people voting (21) and the number of stars (1 out of 5). I'm suspicious.
Hey, he responded! The 6 votes used to be 2.5 stars. Now, it's 1 star. Those folks were right; what an @$$. Here's what he said;
Thank you for your support! You could have used HTML to make the links work.
But I ask you, what the bigger risk? That I delete the mailing list I created, maintain and finance or that Nikon, Canon or all the other camera makers stop supporting some older cameras, leaving you alone with your collection of abandoned RAW files, even you paid for the camera?
Someone posted what looks like a copy of your links to the comment page. The interesting thing? 6 votes when the average number of votes for other messages is 1 or 0. Assuming that the person who posted it voted themselves up, it looks like all the other votes are against this message;
yes, and i'm sure the gimp-print people must be swearing up and down about canon's reluctance as they work hard to crank out epson and hp profiles.
I'll take that as sarcasm. If not...please ignore the following comments.
Both Epson and HP have provided drivers for printing under Unix-like systems. HP, specifically, has gone out of thier way to be helpful and contribute what they have to CUPS. The Gimp-print folks could use that at a minimum while contacting HP's reps for the CUPS drivers for more details if they want it.
"Automated testing is helpful, though mainly for known errors or conditions."
But in this case it was a known error. If the nature of the bug allowed it to be generated and verified using an automated test, you could add it to your regression test. Then if the bug showed up again, the regression test would catch it. This assumes that the automated test isn't dependent on the exact code snippet that caused the orginal problem but rather on the behavior.
Yep. If this specific variation was a known problem, then automated testing would do the trick; there would be no release of the defect in a shipping product since the bug would be caught before it was released as a minor revision to normal users.
It all depends, though...as you likely agree; The test would have to handle all variations on frame types and presentation -- including seperate windows vs. tabs and a simulated 'real' network environment with both 'good' and 'bad' sites. Map it out for yourself, and I'm sure that you can see the list combinations for this one case start to be substantial.
Automated testing is great when your target doesn't move much. If it is dynamic or requires close integration with the test suite to cover special features, it becomes difficult to implement. In the worst case, your test suite becomes as large as the system being tested if not substantially larger.
Slashdot Mirror
2nd: The more time goes on, the more Windows takes on the features of unix.
3rd: Most every OS is some form of unix at this point except for Windows.
4th: Even Windows has a POSIX layer and unix-style command utilities for free as an add-on.
None as far as I can tell. I'd lump X11, BSD, and other similar licences into basically the same category basic category as public domain. I don't know why the other person thought there would be a big advantage to explicitly giving up all rights vs. giving up all but attribution and copyright announcement.
True. I'll add one extra; usually the code but not the project name is granted under BSD style licences. Copyright isn't the same as trademark.
Because of that, if someone takes the code and mangles it to make an incompatable fork, the original authors can state that the fork does not follow the reference implementation and thus isn't *insert name here* compatable. They could force the fork to change any claims of being compatable with the original project. Usually doesn't happen.
Well, I often feel the same way about RMS.
That said, like Nader, I'm occasionally forced to agree with his view of reality and say "The bastard got it right again!". I'm very glad he's around, even if I rarely find that he's 100% spot on.
Don't get me wrong. BSD is a fine license all right, but nothing special compared to the protected freedom of GPL.
While I prefer the GPL over BSD (and similar licences) -- the GPL does not work well in all situations. For example, anything that requires a reference design that you want to be widely adopted. The Ogg codecs (Vorbis, Theora, ...) for example.
If someone is that inattentive, they get banned.
If you want implementation details -- and it looks like you indeed do -- I'll be glad to provide them to you for a fee. Are you that curious, or can you figure some of the basics out for yourself?
Not at all. The AC has it right;
You make a robots.txt like with directories/pages they should not enter. You include links in pages that only a crawler would see normally. Only a crawler that ignored robots.txt or the exclusion in it would go those that trap pages and get banned.
To make this very clear; the links on the legitimate pages are not normally visible or say things like "." or "," or "This is a trap for sp@mmer$"...whatever. Color the text white on a white background. Dinner is served.
No, that's the point. When the spider ignores robots.txt, they pick up the poisoned pages. Then, because they are doing something wrong, punish them.
Here's what would seem to work;
1. Create robots.txt, including references to the spam spider trap. Make sure that the legitimate references to normal pages are out numbered by a large margin.
2. When pages that could only be referenced in the spam spider trap are accessed, note the IP address.
3. Slowly respond or block connections from the originating IP address.
Bad guys are punished. Good guys are not. Low impact on system resources.
There's got to be a dozen filters out there that already do this. Anyone have experience using one?
That's odd. Not on the 2 systems here; XP Pro and XP Home. Sure you didn't install anything?
"The ministry of peace.
The ministry of truth.
The department of homeland security."
I still get the impression that the name implies a salute using a stiff palm raised high. Maybe with a little Vaugner playing in the background.
What moron thought that was a good name?
(scratches head)
1. If you don't know what you had you don't know if what you have has been screwed with.
2. If you do get screwed with, it's critical to be able to restore from a known good system. Otherwise, game over; you have to rebuild from scratch and guess about what has/has not been compromised.
Your example is exaggerated, even for early versions of OSS. If it's too much of a pain, though, I figure that the developer needs to work out bugs as well so I wait for a later release. In either case, you can immediately do something about OSS while with Monad you have to agree to licences and wait.
I knew it when I wrote it [explitive deleted].
It's a pain to put in simple things like a timed pause or command seperation for @#!#$ sake; the main ways I've seen are running ping or other commands that have a timeout feature and use that. In Bash, it's 'sleep 5' for 5 seconds, 'sleep 3600' for an hour. To seperate commands in Bash, insert a ; between commands -- and the | still works!
Command.com vs. cmd.exe are very very similar from a UI perspective. cmd.exe has next to no substantial improvements. It's disapointing that they didn't start on this change with the introduction of NT.
4DOS/4NT are quite nice. If I weren't so used to Bash, I'd be screaming for something similar under Linux.
It's not a mistake that bash (based on sh) has taken so many years to get to the place it is now.
That said, throwing money at the problem may not help if the wrong people are involved in the design. So far, it looks as if they have it right.
It will take years to iron out the problems and discover just what is good and what sucks about it. A staged default adoption is the right way to do it as long as they dump the silly mistakes that have never been corrected since DOS 2.x.
So THAT'S the reason the accents for tech support are so alien!
So far, it's still there. Very strange things going on though such as the number of people voting (21) and the number of stars (1 out of 5). I'm suspicious.
Like FreeBSD? They seem to be alive and kicking.
Thank you for your support! You could have used HTML to make the links work.
But I ask you, what the bigger risk? That I delete the mailing list I created, maintain and finance or that Nikon, Canon or all the other camera makers stop supporting some older cameras, leaving you alone with your collection of abandoned RAW files, even you paid for the camera?
And posted by Anonymous, that's so lame.
"My concern with OpenRAW is that Juergen Specht abused one stewardship -- the Japan Photography Mailing list -- by uniaterally deleting it. Who's to say he won't do it again?" (plus your links)
I would have to bet that Juergen Specht is a very sensitive individual and doesn't like criticism.
I'll take that as sarcasm. If not...please ignore the following comments.
Both Epson and HP have provided drivers for printing under Unix-like systems. HP, specifically, has gone out of thier way to be helpful and contribute what they have to CUPS. The Gimp-print folks could use that at a minimum while contacting HP's reps for the CUPS drivers for more details if they want it.
Wow! Thanks for the heads up!
But in this case it was a known error. If the nature of the bug allowed it to be generated and verified using an automated test, you could add it to your regression test. Then if the bug showed up again, the regression test would catch it. This assumes that the automated test isn't dependent on the exact code snippet that caused the orginal problem but rather on the behavior.
Yep. If this specific variation was a known problem, then automated testing would do the trick; there would be no release of the defect in a shipping product since the bug would be caught before it was released as a minor revision to normal users.
It all depends, though...as you likely agree; The test would have to handle all variations on frame types and presentation -- including seperate windows vs. tabs and a simulated 'real' network environment with both 'good' and 'bad' sites. Map it out for yourself, and I'm sure that you can see the list combinations for this one case start to be substantial.
Automated testing is great when your target doesn't move much. If it is dynamic or requires close integration with the test suite to cover special features, it becomes difficult to implement. In the worst case, your test suite becomes as large as the system being tested if not substantially larger.