Unfortunately, what I think we really need is a GPL or BSD licenced JVM. Without one, you will never see a java application packaged into Debian and other free operating systems. For this to happen it must support AWT. JET looks good, but it fails the license test.
So a month later hes backlogged in unmerged branges that hes not had a chance to look at yet? How exactly does that help?
Well, if he's now using CVS, he could allow people he trusts (the 5 or so people he talks about) to merge the less trusted branches back into the main trunk.
But don't ask me to tell Linus how to maintain his own kernel:)
Re:I know Linus doesn't like it...
on
Linus Does Not Scale
·
· Score: 2, Interesting
It simply won't help the problem... source management isn't the problem. Making sure useless crap doesn't make it into the tree is the problem. CVS doesn't have any sort of means to make sure that what goes in is quality code.
No, but CVS does allow branching of the tree. Couldn't different patches just be maintained on different branches, and merged into the trunk when it is finally accepted?
If they change the standards, they'll have to release the code, which means interoperation will be trivial. Don't like the AOL standard's? No worries, don't use them, just means you won't be able to communicate with the AOL users.
I am getting so annoyed at this. Everyone *thinks* that this was done by normal moderation means. I *know* it wasn't.
There were multiple posts, about 20 or so, just discussing the moderation system. They were scored between -1 and 5. I was refreshing the page about once every three minutes. It takes time for a post's moderation to change - moderations aren't done all that often. Then *BAM*, every post went to -1. In just a couple of minutes (a single page refresh). It wasn't until someone posted something that couldn't be bitchslapped to -1 for being offtopic (it was ontopic) that the 'users' had to keep this thread in check by hand.
The problem is, it now *looks* like it was moderated down using the standard moderation system (you know, those 5 points they give you). It *wasn't*.
Re:The first Slashdot troll post investigation
on
KaZaa Suspends Downloads
·
· Score: -1, Offtopic
Actaully, I was hoping they were going to let the moderations run a little longer, just to see where everyone's post would end up -- but then again, strictly speaking, its offtopic.
You see, the moderation system works - its all offtopic... oh, well, its only karma
You don't need to send the virus (lantern) signature. Its just a matter of *not* putting the virus signature in the list of detectable virii. If they find a variant trojan in the wild, they find a signature in the variant not present in the original. This simply means that McAfee and Symantec will know the Magic Lantern signature, they just won't tell anyone what it is.
The one thing that gnutella clients need is support for downloading from multiple hosts simultaneously. I haven't yet found a client that supports this, or is it some limitation on the network?
I don't think its a limitation of the protocol as such, but there is no information as to the MD5 (or other hash) of the files you are downloading. So you don't really know that the multiple files are the same. If you knew that these 'N' hosts all had the same file (from the hash or MD5), you could then request whatever part of the file you wanted from each host and build the resulting final file, and check that it matches its MD5. Without a hash algorithm, you could still 'assume' it was the same file by name and filesize, and try the simultaneous approach, but assume makes an ass out of u and me!
Because the Kazaa networks now go through a single host before logging on, introducing a single point of failure/tracking on the network.
I agree that the Kazaa/Morpheus/giFT technology is superior, but the giFT engine has to be bought up to speed (super-server support), and reliance on the Kazaa/Morpheus networks themselves need to be dropped.
I agree with everything you say except the part about the dll being easy to port. I'll beleive it when I see a kde port. Even then, it won't work across the internet, because the native part can't be d/loaded. Perhaps if the native part was built into the JVM it really would become the Java UI alternative.
Which works great on Redhat, Mandrake and the rest...
Also, tools like this only work well if you have a pool of.deb packages to load from, doesn't seem right for a single product, does it? Not if you consider that it requires a lot of work to get the dependencies right in the package itself, and then you have to match it to a given distribution.
Package management utilities are great for installing software from a given distribution, but they aren't great for cross-distribution software. Debian creates.deb packages for the different versions of Debian. Redhat makes their own RPMs for their system, and Mandrake creates yet another set of RPMs for Mandrake. Although you can run Redhat RPMs on a Mandrake system, and vice-versa (.deb too?), I've found it isn't as stable.
What about compressed XML. Its small, cause the compression deals with the redundancy, and its robust / reasonably self documenting and all the other good things about XML.
:) - No, I don't advocate any of these things, but some people have it good, and don't give a damn about the rest of the people on the planet. Its just a lack of empathy for their fellow human being. If they had actaully suffered, and were refugee's themselves, well, I guess they wouldn't be thinking like this anyway, but you *have* to be able to look at the world from different points of view.
Just for the record, I don't agree with my country's use of detention camps either. Its exactly the same phenomonen.
Perhaps if you look at the [...] bit, the post would make more sense. The point is, treat other people poorly, get treated poorly.
Any Linux user who tries a stunt like that deserves a seg fault (or worse)
This is something that annoys me - programs that just go 'Segmentation Fault' in the night... What do you do with a seg fault? How do you find the cause of a seg fault? What causes them? Unmatched libraries and application code? Null / Uninitialised pointer exceptions?
If anyone can enlighten me, I'd really appreciate it. Gabber was giving me the old seg fault, and that's just not enough information to fix it!
Better still is Java's error handling. I plug Java a lot, because it's what I use now and there are Open Source versions of Java such as Kaffe - its not the only language on the planet (but it should be:) ).
Anyway, throw an exception in Java, it should have a name that makes sense, eg NullPointerException. Then you also get access to a stack trace that can be logged. If you have JIT turned off, you get line numbers as well as the usual class and method names. Really nice for working out what went wrong. With a well designed app and exception handling, your program can carry on, after dealing with the exception, and provide a very useful log for improving the source code.
Someone should fork Kaffe and continue work on it, or produce an OS JVM written in Java, like IBM's Jallapeno Java RVM.
Slashdot Mirror
Talk about mood stabilisers
+1 vt 4 tny blair pls. thnks, cyal8r 8)
Unfortunately, what I think we really need is a GPL or BSD licenced JVM. Without one, you will never see a java application packaged into Debian and other free operating systems. For this to happen it must support AWT. JET looks good, but it fails the license test.
So a month later hes backlogged in unmerged branges that hes not had a chance to look at yet? How exactly does that help?
:)
Well, if he's now using CVS, he could allow people he trusts (the 5 or so people he talks about) to merge the less trusted branches back into the main trunk.
But don't ask me to tell Linus how to maintain his own kernel
It simply won't help the problem... source management isn't the problem. Making sure useless crap doesn't make it into the tree is the problem. CVS doesn't have any sort of means to make sure that what goes in is quality code.
No, but CVS does allow branching of the tree. Couldn't different patches just be maintained on different branches, and merged into the trunk when it is finally accepted?
If they change the standards, they'll have to release the code, which means interoperation will be trivial. Don't like the AOL standard's? No worries, don't use them, just means you won't be able to communicate with the AOL users.
I am getting so annoyed at this. Everyone *thinks* that this was done by normal moderation means. I *know* it wasn't.
There were multiple posts, about 20 or so, just discussing the moderation system. They were scored between -1 and 5. I was refreshing the page about once every three minutes. It takes time for a post's moderation to change - moderations aren't done all that often. Then *BAM*, every post went to -1. In just a couple of minutes (a single page refresh). It wasn't until someone posted something that couldn't be bitchslapped to -1 for being offtopic (it was ontopic) that the 'users' had to keep this thread in check by hand.
The problem is, it now *looks* like it was moderated down using the standard moderation system (you know, those 5 points they give you). It *wasn't*.
Still don't believe me? -- Jamie admits modding offtopic posts with unlimmited mod points
Actaully, I was hoping they were going to let the moderations run a little longer, just to see where everyone's post would end up -- but then again, strictly speaking, its offtopic.
You see, the moderation system works - its all offtopic... oh, well, its only karma
Thankyou... you learn something new everyday. Even if it is from an AC :)
You don't need to send the virus (lantern) signature. Its just a matter of *not* putting the virus signature in the list of detectable virii. If they find a variant trojan in the wild, they find a signature in the variant not present in the original. This simply means that McAfee and Symantec will know the Magic Lantern signature, they just won't tell anyone what it is.
Why is this so hard to understand?
So, the website gets 1 penny from the mirror operator? That's what I call Return on Investment!!
You can always accomplish any "normal" task if the OS provides you a good API or DDK.
Except that this isn't a "normal" task, is it?
The one thing that gnutella clients need is support for downloading from multiple hosts simultaneously. I haven't yet found a client that supports this, or is it some limitation on the network?
I don't think its a limitation of the protocol as such, but there is no information as to the MD5 (or other hash) of the files you are downloading. So you don't really know that the multiple files are the same. If you knew that these 'N' hosts all had the same file (from the hash or MD5), you could then request whatever part of the file you wanted from each host and build the resulting final file, and check that it matches its MD5. Without a hash algorithm, you could still 'assume' it was the same file by name and filesize, and try the simultaneous approach, but assume makes an ass out of u and me!
Gnucleus has had the ability to map gnutella nodes for months now. Runs under win32/WINE.
Because the Kazaa networks now go through a single host before logging on, introducing a single point of failure/tracking on the network.
I agree that the Kazaa/Morpheus/giFT technology is superior, but the giFT engine has to be bought up to speed (super-server support), and reliance on the Kazaa/Morpheus networks themselves need to be dropped.
I think I left out the ':)'...
I agree with everything you say except the part about the dll being easy to port. I'll beleive it when I see a kde port. Even then, it won't work across the internet, because the native part can't be d/loaded. Perhaps if the native part was built into the JVM it really would become the Java UI alternative.
Which works great on Redhat, Mandrake and the rest...
.deb packages to load from, doesn't seem right for a single product, does it? Not if you consider that it requires a lot of work to get the dependencies right in the package itself, and then you have to match it to a given distribution.
.deb packages for the different versions of Debian. Redhat makes their own RPMs for their system, and Mandrake creates yet another set of RPMs for Mandrake. Although you can run Redhat RPMs on a Mandrake system, and vice-versa (.deb too?), I've found it isn't as stable.
Also, tools like this only work well if you have a pool of
Package management utilities are great for installing software from a given distribution, but they aren't great for cross-distribution software. Debian creates
Ugh, I'm bored and you're a Moron.
They should do an AWT port of it to make it cross-platform.
So, does this mean we'll be seeing a free Java UI toolkit that doesn't suck?
Cool!
Open source, clean room JVM implementation - Kaffe!!
Looks like a great case for using something like SwarmCast.
On a completely offtopic note, a swarmcast like system would work great with the P2P file sharing programs, would it?
What about compressed XML. Its small, cause the compression deals with the redundancy, and its robust / reasonably self documenting and all the other good things about XML.
:) - No, I don't advocate any of these things, but some people have it good, and don't give a damn about the rest of the people on the planet. Its just a lack of empathy for their fellow human being. If they had actaully suffered, and were refugee's themselves, well, I guess they wouldn't be thinking like this anyway, but you *have* to be able to look at the world from different points of view.
Just for the record, I don't agree with my country's use of detention camps either. Its exactly the same phenomonen.
Perhaps if you look at the [...] bit, the post would make more sense. The point is, treat other people poorly, get treated poorly.
Any Linux user who tries a stunt like that deserves a seg fault (or worse)
This is something that annoys me - programs that just go 'Segmentation Fault' in the night... What do you do with a seg fault? How do you find the cause of a seg fault? What causes them? Unmatched libraries and application code? Null / Uninitialised pointer exceptions?
If anyone can enlighten me, I'd really appreciate it. Gabber was giving me the old seg fault, and that's just not enough information to fix it!
Cheers!
Better still is Java's error handling. I plug Java a lot, because it's what I use now and there are Open Source versions of Java such as Kaffe - its not the only language on the planet (but it should be :) ).
Anyway, throw an exception in Java, it should have a name that makes sense, eg NullPointerException. Then you also get access to a stack trace that can be logged. If you have JIT turned off, you get line numbers as well as the usual class and method names. Really nice for working out what went wrong. With a well designed app and exception handling, your program can carry on, after dealing with the exception, and provide a very useful log for improving the source code.
Someone should fork Kaffe and continue work on it, or produce an OS JVM written in Java, like IBM's Jallapeno Java RVM.