Also, I find that the documentation is pretty good, but they are in desperate need of a Zope Cookbook. Do you know about http://www.zopelabs.com/cookbook?
For one, the new model includes some features that Linux/UNIX don't readily support - and has some features that are very interesting. Basically, the ACL's are still around - which is nice and all. But most interestingly applications are run with permissions - and not just in the sense of the running as a user. Specific fine-grained controls are possible (though I am unclear as to if they are currently implemented.. I haven't found them yet!) that detail which resources the application has permissions to access - regardless of user context.
This is not quite correct. See for instance here or look up capabilites and unix and posix in google (you may have to search a bit). Or surf here to learn that linux also has capabilities. This is also NAMI (not a microsoft innovation).
What will slow (but not kill) the PC industry is exactly what has been said before... nobody wants to upgrade because their computers already to what they want them to. Anybody with a Pentium class machine can run Win9x reasonably well, email, Word, internet. If that's all you're doing, why the hell would you upgrade??
What you state is certainly right, but it's IMO not a complete characterisation (sp?) of the PC market.
Take for instance BEOS. I'm *not' a BEOS advocate (I don't even know how to capitalize it right) - but I had it installed one time.
Beos already did a lot of things better than windows, but as the market is, it was more or less clear that beos would fail. This is not solely MS' fault, the whole market has the problem of hardware compability, software compability, application file format compability, distribution problems (preinstallations on new hardware) etc.
To abuse the particularly loved car analogy, if every new market entrepreneur had to build it's own " compatible" roads,...
What MS does is to do abuse it's power to keep these barriers as high as possible.
Of course, he only evers goes after smaller fish, but doesn't dare trying the same with bigger companies (unfortunately, or he'd at last get his nose bloody).
this is *not* true. He went against microsoft (with the "explorer" trademark) and won - well, at least MS settled out of court, supposedly paying quite a dime.
He also went against heise (again "explorer") - and lost in the first trial...
Say what you want against him, but he doesn't fear big fishes.
Just a comment, something which I vaguely remember reading on ntbugtraq.
I wouldn't trust "Run As" too much. There may be some nasties w.r.t applications which have several threads running, most importantly IE.
That means if you have an IE window open as user penguin42, doing a runas on a new IE might _not_ run as testuser (the new IE is just a new thread of the running one, with the same privileges).
Reviews and audits are also done with bridges, but nobody in their right mind would come to the conclusion that you could abandon margins because of good peer review of the design.
Simply put, in bridge building you have always the *additional* security of safety margins, without waiving the principles of peer review.
OTOH, there you are right, CS has the possibility of easily producing pre-/beta- stuff, something bridge designers can't do to that extend.
But it seems that this isn't enough to make OSs as reliable as bridges;-).
don't be silly.
WTF do (ex-)linux companies have to do with the quote you posted.
I think this quote has a point.
If we go into comparing, let's say, building bridges and os programming, I think we _can_ see the differences in methodologies one needs.
With bridges, we have a well known and accepted theory of their statics, a relativly narrow expectation what we expect a bridge to do, and we can, by using tolerances of a wide margin, account for the fact that something unexpected happens.
In an os, there is not really a broadly accepted theory (micro- vs macro-kernel, VM, filesystems, implemetation language) - at least when we look how different realisations we see in practice.
What do we expect am OS to do, or more precisly, what do we expect an OS to do well?
latency vs throughput, single vs massivly multi cpu, graphics in kernel vs graphics in userspace...
Seems we have no real consensus here.
At last, and this is perhaps the most important factor - we can't make an OS more failsafe (or performing better) by introducing margins anywhere. Due to the binary nature of CS it doesn't make sense to use redundancy for many aspects of an OS.
It either works or fails.
you're right, it might not exactly have been the "OS/2" bluescreen, but fact is the machine was rebooting and didn't come up - at least not up to it's normal functionality. I suspected a reboot because of a driver. Spontanous reboots when something in kernelspace goes wrong happen under win* and linux, why not with OS/2?
So, the difference to a BSOD is not too big, that was what I meant(sp?)
Well, it was the OS/2 Warp equivalent of a bsod. The machine rebooted, I saw BIOS messages.
Then I went to another ATM, got my money, and out of curiosity, went again to the broken ATM.
OS/2 was still booting, it looked very ugly and hacky (some errors in autoexec.bat or whatever it's called in OS/2) - it also told me several times to hit enter or something.
Then something funny happened, it told me (as I was looking at the screen) to insert a disk. It beeped and *lighted the slot where you insert the card*!
Interestingly enough the fact that the US bombed a factory that was producing medicine for in a poor country that is torn apart by famine, disease and strife is one of the rallying cries that Bin Laden used to recruit and swell the ranks of Al-Qaeda. no, the bombing makes sense, if you assume that
a) it was a quite important factory (market leader of malaria treatments in north africa).
b) they suspected bin laden and his network of (partly-)owning that factory
See:
http://globalization.about.com/library/weekly/aa 10 2899.htm
http://www.apbnews.com/newscenter/majorcases/bin la den/1999/09/24/missile0924_01.html
Well, this one little thing: Virtual hosts... If there's multiple DNS names pointing on same server, the server hicups if it sees someone wanting just the page and not providing the host name in Host: header.
Well, two little things (one for *nix, one for windows nt):
I can give two examples, things that I really hate with IE and which IMO clearly exist for websites to dictate what I do:
If you configured IE to ask whether it should accept cookies, it asks. But where's in the dialog which pops up to get IE to _ignore_ cookies from this page. If you check "don't ask again" it will accept cookies from now on. This is wrong.
If you disable ActiveX, IE will always pop up a window which informs you that this page may look sucky because of your current security settings.
Maybe I'm just to dense to configure IE correctly, but then it's too hard to find, IMO.
Make no mistake, the Taliban is a really vile group of people who are as bad as they are made out to be. But also don't make the mistake of thinking that the U.S. government would have cared about them, their treatment of their citizens or Afghanistan in general if it weren't for 9/11.
Please mod parent up. It just isn't right to rate this as flamebait.
Don't believe me? Just take a better look at some of our allies in the war against terrorism.
Notice the *identical* number of hits for C, C++ and C#? Obviously a flaw in the search on dice. However, it's probably safe to say that none of C, C++ or C# produces more than that number of hits legitimately.
Really?
Just click on the first link of the C# results and tell me where you find C# in the job description.
Hint: Nowhere
Conclusion: Your method is completly flawed, presumably because their search engine just ignores '#' and '++'.
Btw., even if it worked, your C# and C++ results would be a subset of your 'C' results.
Interesting, I hate that feature (really - using mostly ms-win oss on my desktops).
It confuses the shit out of me when helping colleagues on their machine and even on my own. I prefer to organize things like I want them to have to avoid cluttering, not some pseudo-intelligent algorithm. Hierachical organization of apps is the key - unfortunatly mainly MS-Apps install themselfes directly in the start menu...
Show me where, in the Constitution, one has a right to private communication with one's lawyer.
I'm not an us citizen, but show me where in your constitution it is explicitly forbidden to make a law that muslims have to wear visible yellow signs on their clothes.
No, nobody want's to do that, but it shows your argument is wrong.
Slashdot Mirror
Also, I find that the documentation is pretty good, but they are in desperate need of a Zope Cookbook.
Do you know about
http://www.zopelabs.com/cookbook?
For one, the new model includes some features that Linux/UNIX don't readily support - and has some features that are very interesting. Basically, the ACL's are still around - which is nice and all. But most interestingly applications are run with permissions - and not just in the sense of the running as a user. Specific fine-grained controls are possible (though I am unclear as to if they are currently implemented.. I haven't found them yet!) that detail which resources the application has permissions to access - regardless of user context.
This is not quite correct. See for instance here or look up capabilites and unix and posix in google (you may have to search a bit). Or surf here to learn that linux also has capabilities. This is also NAMI (not a microsoft innovation).
What will slow (but not kill) the PC industry is exactly what has been said before... nobody wants to upgrade because their computers already to what they want them to. Anybody with a Pentium class machine can run Win9x reasonably well, email, Word, internet. If that's all you're doing, why the hell would you upgrade??
...
What you state is certainly right, but it's IMO not a complete characterisation (sp?) of the PC market.
Take for instance BEOS. I'm *not' a BEOS advocate (I don't even know how to capitalize it right) - but I had it installed one time.
Beos already did a lot of things better than windows, but as the market is, it was more or less clear that beos would fail. This is not solely MS' fault, the whole market has the problem of hardware compability, software compability, application file format compability, distribution problems (preinstallations on new hardware) etc.
To abuse the particularly loved car analogy, if every new market entrepreneur had to build it's own " compatible" roads,
What MS does is to do abuse it's power to keep these barriers as high as possible.
Of course, he only evers goes after smaller fish, but doesn't dare trying the same with bigger companies (unfortunately, or he'd at last get his nose bloody).
...
this is *not* true. He went against microsoft (with the "explorer" trademark) and won - well, at least MS settled out of court, supposedly paying quite a dime.
He also went against heise (again "explorer") - and lost in the first trial
Say what you want against him, but he doesn't fear big fishes.
Thanks for your answer. I have similar problems.
And as nice as xml-rpc, soap, webdav and anythingelseunderthesun-over-http is, soon nobody will be able to be sure what really passes his DMZ config.
I know that problem...
Out of curiosity, have you thought of proxys + DMZ? Are the clients in your network too diverse?
Just a comment, something which I vaguely remember reading on ntbugtraq.
I wouldn't trust "Run As" too much. There may be some nasties w.r.t applications which have several threads running, most importantly IE.
That means if you have an IE window open as user penguin42, doing a runas on a new IE might _not_ run as testuser (the new IE is just a new thread of the running one, with the same privileges).
No, IMO the principles are not the same.
;-).
Reviews and audits are also done with bridges, but nobody in their right mind would come to the conclusion that you could abandon margins because of good peer review of the design.
Simply put, in bridge building you have always the *additional* security of safety margins, without waiving the principles of peer review.
OTOH, there you are right, CS has the possibility of easily producing pre-/beta- stuff, something bridge designers can't do to that extend.
But it seems that this isn't enough to make OSs as reliable as bridges
AFAIK, your "dumb" place the place where gnome should go, according to the LSB.
Ok, I'm a bit late, but anyway:
l o. html
You are so right.
Esp. when you go to google and type in:
postgres 7.1 blob jdbc
you get the following link on the first page of search results:
http://www.postgresql.org/idocs/index.php?jdbc-
don't be silly.
...
WTF do (ex-)linux companies have to do with the quote you posted.
I think this quote has a point.
If we go into comparing, let's say, building bridges and os programming, I think we _can_ see the differences in methodologies one needs.
With bridges, we have a well known and accepted theory of their statics, a relativly narrow expectation what we expect a bridge to do, and we can, by using tolerances of a wide margin, account for the fact that something unexpected happens.
In an os, there is not really a broadly accepted theory (micro- vs macro-kernel, VM, filesystems, implemetation language) - at least when we look how different realisations we see in practice.
What do we expect am OS to do, or more precisly, what do we expect an OS to do well?
latency vs throughput, single vs massivly multi cpu, graphics in kernel vs graphics in userspace
Seems we have no real consensus here.
At last, and this is perhaps the most important factor - we can't make an OS more failsafe (or performing better) by introducing margins anywhere. Due to the binary nature of CS it doesn't make sense to use redundancy for many aspects of an OS.
It either works or fails.
dumb question:
if you are running your own mailserver, why don't you generate artificial bounces for spam mails?
you're right, it might not exactly have been the "OS/2" bluescreen, but fact is the machine was rebooting and didn't come up - at least not up to it's normal functionality. I suspected a reboot because of a driver. Spontanous reboots when something in kernelspace goes wrong happen under win* and linux, why not with OS/2?
So, the difference to a BSOD is not too big, that was what I meant(sp?)
lately I saw an BSOD'ed ATM.
Well, it was the OS/2 Warp equivalent of a bsod. The machine rebooted, I saw BIOS messages.
Then I went to another ATM, got my money, and out of curiosity, went again to the broken ATM.
OS/2 was still booting, it looked very ugly and hacky (some errors in autoexec.bat or whatever it's called in OS/2) - it also told me several times to hit enter or something.
Then something funny happened, it told me (as I was looking at the screen) to insert a disk. It beeped and *lighted the slot where you insert the card*!
some are quite good
you know nessus?
just strip the slashcode has inserted into the links...
forget it, just click here and here
funny thing is that the second server just seemed to die.
woah, your app (tora) look really nice. hmmmm, you know, postgres is a nice database, too ...
Seriously, tora looks like a very cool effort
Interestingly enough the fact that the US bombed a factory that was producing medicine for in a poor country that is torn apart by famine, disease and strife is one of the rallying cries that Bin Laden used to recruit and swell the ranks of Al-Qaeda.
a 10 2899.htm
n la den/1999/09/24/missile0924_01.html
no, the bombing makes sense, if you assume that
a) it was a quite important factory (market leader of malaria treatments in north africa).
b) they suspected bin laden and his network of (partly-)owning that factory
See:
http://globalization.about.com/library/weekly/a
http://www.apbnews.com/newscenter/majorcases/bi
Take al ook at mupad
It's some sort of mathematica lookalike, superior in some cases and they have free versions.
It's been a while since I used it, but it was great.
Well, two little things (one for *nix, one for windows nt):
Maybe I'm just to dense to configure IE correctly, but then it's too hard to find, IMO.
Make no mistake, the Taliban is a really vile group of people who are as bad as they are made out to be. But also don't make the mistake of thinking that the U.S. government would have cared about them, their treatment of their citizens or Afghanistan in general if it weren't for 9/11.
Please mod parent up. It just isn't right to rate this as flamebait.
Don't believe me? Just take a better look at some of our allies in the war against terrorism.
Notice the *identical* number of hits for C, C++ and C#? Obviously a flaw in the search on dice. However, it's probably safe to say that none of C, C++ or C# produces more than that number of hits legitimately.
Really?
Just click on the first link of the C# results and tell me where you find C# in the job description.
Hint: Nowhere
Conclusion: Your method is completly flawed, presumably because their search engine just ignores '#' and '++'.
Btw., even if it worked, your C# and C++ results would be a subset of your 'C' results.
Interesting, I hate that feature (really - using mostly ms-win oss on my desktops).
It confuses the shit out of me when helping colleagues on their machine and even on my own. I prefer to organize things like I want them to have to avoid cluttering, not some pseudo-intelligent algorithm. Hierachical organization of apps is the key - unfortunatly mainly MS-Apps install themselfes directly in the start menu...
Show me where, in the Constitution, one has a right to private communication with one's lawyer.
I'm not an us citizen, but show me where in your constitution it is explicitly forbidden to make a law that muslims have to wear visible yellow signs on their clothes.
No, nobody want's to do that, but it shows your argument is wrong.