Not really - they are suited to different purposes. NTFS is significantly slower than ext2. On the other hand, as you kinda imply, NTFS is much more robust.
I personally value stability over speed, but it's a close call, especially since I have never lost data as a result of having to reboot without unmounting.
There is misinformation here - the current Linux support for NTFS is nothing to do with Jeff Merkey or his company. AFAIK Microsoft has not sent any such threats to the developers of the Linux driver.
You could do all the above things in Python - although I think you probably mean X client, since in X parlance applications are servers and that thing wot displays them is the client.
Writing an X client in Python - or Java - is a fairly stupid idea, incidentally. X is slow enough as it is without needing interpretive overhead care of the implementation language.
Java is like a master magician whereas Python is like a prentice mage.
I should stop replying to trolls.
Further to the above, there is at least one manufacturer whose CD-ROM drives will not allow extraction of CD-DA, a restriction placed deliberately as a result of the dubious (at least, according to this manufacturer) legality of the resultant data.
Once again, the studies conflict here. Which? UK (I think it was UK, but I've been using a mobile phone recently so can't be sure) did that study. Which? somewhere else did a similar test and found the opposite.
Even if that wasn't the case, your comment is still misleading: it applied to certain models of phones only.
Presumably the word `legacy' was intended to mean "Themes that I was using before I upgraded to KDE2". It's not legacy in the sense of outdated software per se, it's legacy in that you've got a new program to play with and so won't be using older stuff again.
As someone else pointed out, it's also a subtle joke - playing on the attitude that "Now I've got KDE2 I won't need anything else from anywhere ever, but I still want to use my old GTK themes!"
How can a library which is a superset of another library be "inferior"?
In a mathematical sense, that's a convincing argument. In the real world, it is impossible to maintain every benefit of a program and yet simultaneously add features.
For example, The Linux 2.2 kernel is more featureful than the Linux 2.0 kernel, but on the other hand it's bigger. If you were building embedded systems perhaps you'd choose the older kernel.
This sort of logic is the same sort of logic that made/makes large numbers of people and businesses run, select, or switch over to Microsoft Windows systems... 'gee, that _fill in your favorite superior computing platform here_ never caught on, why should we bother using it?'
...
There are of course lots of other issues, such as cost of media, installed base...
You seem (by mentioning Microsoft on Slashdot;) to be implying that switching to a technology just because it is popular is a flawed decision, but then you admit that it's an "issue".
The only reason people use floppies is because they're popular. Standards are important when they govern information interchange. Imagine if there were two different and incompatible varieties of audio CD. Or hundreds of different and incompatible varieties of HTML. Hang on, I just remembered something.:p
Well, I'm really not. I don't want to distribute information across the freenet - all the replication is wasteful, imo. Also, freenet does not break data into chunks.
Assuming data replication isn't a problem, however, freenet's probably an ideal carrier for this type of system.
Grr, and when I had this idea a while back I was sure it was original.
The idea I had: a Gnutella-like system, where information is distributed over several different computers. No individual computer that holds a segment of the information knows what it's for. To make sense of the information, you need to get two things: the decryption pad (which would be, as mentioned here, XORed with the data) and the "Key": a list of servers containing chunks of the information. These would be available from separate sites, or perhaps distributed across multiple sites.
Because no server containing the information also contains a key, and no server containing the information also contains a pad, and especially because there is no way to tell the information stored on the server from meaningless random garbage unless the pad is applied, no server would ever be liable for information it contained.
Anyway, I'm still interested in writing a system like this: wxWindows is the preferred implementation API. It could, in fact, be implemented alongside or on top of Gnutella itself. Email me if you'd like to be a part of it or would like to see it happen; if I get enough interest, I'll write it.
How To Write A Letter 1) Get Pen 2) Get Paper 3) Apply pen to paper 4) Move pen according to the following movement pattern: 5) (etc)
The above is certainly an algorithm. But try porting it to X-Windows. Just because an algorithm relies on certain other things being present (pen and paper in this case, specific computer architecture in the other) doesn't make it less of an algorithm.
Every who'se complained about the Fortran compiler, stop it: an algorithm takes inputs and produces outputs based on those inputs in a deterministic way. Quicksort is an algorithm, even though it *contains* other algorithms, such as those for iterating through a list and comparing elements. Fortran compiler is an algorithm, even though it contains many smaller algorithms.
Easily dismissed as the ramblings of an insecure, xenophobic fool those ideas are, aren't they?
Well, yes. If you're prepared to outsource IT dev jobs to someone in Kyrgyzstan you will almost certainly regret it later, when the app breaks, or you want new features, or you just need some help working it or making it integrate with the rest of the system.
The bulk of system development time is spent in testing, maintenance and modification. Initial coding's the easy part. IMO you'd have to be insane to outsource part of your project to unknown people you can't contact.
> No, because Napster's bot ban is unreasonable. > It smells suspiciously to me, like an attempt to > couple their client software with their service, > and I'm rather surprised that people here on > Slashdot would defend such a practice.
If it *is* such an attempt, it's a pretty poor one, considering the amount of Napster clones around for various other OSes. I've got no idea how you arrived at this conclusion, however: Napster's bot ban is similar to many IRC servers' bot bans: they're not limiting client software to theirs only, they're (presumably) stopping people from automatically searching for Metallica every 30 seconds, 24 hours a day, ie, to reduce load on their servers.
AFAIK Napster has never complained about people cloning their software, either. At least not recently. The conclusion you make is just nonsensical, and the comparisons you draw on aren't valid. Napster isn't limiting you to using a product produced by a certain company; they're specifically *prohibiting* you from using *any* product that performs a certain function (ie, automated searches). Think of it as a negative filter rather than a positive one. To use another analogy, it's the difference between "Don't kill people" (negative filter) and "You must walk down the street with a smile on your face at all times") (positive filter).
Of course, there's sometimes a fine line between a client and a bot. There are many modified Gnutella clients that send back `The RIAA is watching you!' or similar in response to every search, but they - presumably - also function as an otherwise normal Gnutella client.
Re:Whassa matter, the new kid scare you?
on
AtheOS
·
· Score: 1
What?
So far, I've seen no negative comments. The closest I got was someone asking if the thing was POSIX compliant. Everyone else has seemed excited and interested.
I certainly think you're not alone in wanting to try it. As an avid Linux user for the past verylongtime, I certainly want to.
Just because it's pulled rather than pushed doesn't mean it's not propagation.
When something "Propagates", it just means copies are made and stored in places other than the original location. In the same way that the Slashdot logo has propagated to hundreds of thousands of WWW caches all over the globe.
Slashdot Mirror
Not really - they are suited to different purposes. NTFS is significantly slower than ext2. On the other hand, as you kinda imply, NTFS is much more robust.
I personally value stability over speed, but it's a close call, especially since I have never lost data as a result of having to reboot without unmounting.
There is misinformation here - the current Linux support for NTFS is nothing to do with Jeff Merkey or his company. AFAIK Microsoft has not sent any such threats to the developers of the Linux driver.
Reference: Linux NTFS page
But to me, it's worse than fingernails on a blackboard.
Argh! Bastard!
Writing an X client in Python - or Java - is a fairly stupid idea, incidentally. X is slow enough as it is without needing interpretive overhead care of the implementation language.
Java is like a master magician whereas Python is like a prentice mage.
I should stop replying to trolls.
Further to the above, there is at least one manufacturer whose CD-ROM drives will not allow extraction of CD-DA, a restriction placed deliberately as a result of the dubious (at least, according to this manufacturer) legality of the resultant data.
Once again, the studies conflict here. Which? UK (I think it was UK, but I've been using a mobile phone recently so can't be sure) did that study. Which? somewhere else did a similar test and found the opposite.
Even if that wasn't the case, your comment is still misleading: it applied to certain models of phones only.
but He was busy writing Linux.
Presumably the word `legacy' was intended to mean "Themes that I was using before I upgraded to KDE2". It's not legacy in the sense of outdated software per se, it's legacy in that you've got a new program to play with and so won't be using older stuff again.
As someone else pointed out, it's also a subtle joke - playing on the attitude that "Now I've got KDE2 I won't need anything else from anywhere ever, but I still want to use my old GTK themes!"
How can a library which is a superset of another library be "inferior"?
In a mathematical sense, that's a convincing argument. In the real world, it is impossible to maintain every benefit of a program and yet simultaneously add features.
For example, The Linux 2.2 kernel is more featureful than the Linux 2.0 kernel, but on the other hand it's bigger. If you were building embedded systems perhaps you'd choose the older kernel.
This sort of logic is the same sort of logic that made/makes large numbers of people and businesses run, select, or switch over to Microsoft Windows systems... 'gee, that _fill in your favorite superior computing platform here_ never caught on, why should we bother using it?'
;) to be implying that switching to a technology just because it is popular is a flawed decision, but then you admit that it's an "issue".
:p
...
There are of course lots of other issues, such as cost of media, installed base...
You seem (by mentioning Microsoft on Slashdot
The only reason people use floppies is because they're popular. Standards are important when they govern information interchange. Imagine if there were two different and incompatible varieties of audio CD. Or hundreds of different and incompatible varieties of HTML. Hang on, I just remembered something.
Yes. Always.
Well, I'm really not. I don't want to distribute information across the freenet - all the replication is wasteful, imo. Also, freenet does not break data into chunks.
Assuming data replication isn't a problem, however, freenet's probably an ideal carrier for this type of system.
Grr, and when I had this idea a while back I was sure it was original.
The idea I had: a Gnutella-like system, where information is distributed over several different computers. No individual computer that holds a segment of the information knows what it's for. To make sense of the information, you need to get two things: the decryption pad (which would be, as mentioned here, XORed with the data) and the "Key": a list of servers containing chunks of the information. These would be available from separate sites, or perhaps distributed across multiple sites.
Because no server containing the information also contains a key, and no server containing the information also contains a pad, and especially because there is no way to tell the information stored on the server from meaningless random garbage unless the pad is applied, no server would ever be liable for information it contained.
Anyway, I'm still interested in writing a system like this: wxWindows is the preferred implementation API. It could, in fact, be implemented alongside or on top of Gnutella itself. Email me if you'd like to be a part of it or would like to see it happen; if I get enough interest, I'll write it.
Muh!
How To Write A Letter
1) Get Pen
2) Get Paper
3) Apply pen to paper
4) Move pen according to the following movement pattern:
5) (etc)
The above is certainly an algorithm. But try porting it to X-Windows. Just because an algorithm relies on certain other things being present (pen and paper in this case, specific computer architecture in the other) doesn't make it less of an algorithm.
Every who'se complained about the Fortran compiler, stop it: an algorithm takes inputs and produces outputs based on those inputs in a deterministic way. Quicksort is an algorithm, even though it *contains* other algorithms, such as those for iterating through a list and comparing elements. Fortran compiler is an algorithm, even though it contains many smaller algorithms.
Easily dismissed as the ramblings of an insecure, xenophobic fool those ideas are, aren't they?
Well, yes. If you're prepared to outsource IT dev jobs to someone in Kyrgyzstan you will almost certainly regret it later, when the app breaks, or you want new features, or you just need some help working it or making it integrate with the rest of the system.
The bulk of system development time is spent in testing, maintenance and modification. Initial coding's the easy part. IMO you'd have to be insane to outsource part of your project to unknown people you can't contact.
> No, because Napster's bot ban is unreasonable.
> It smells suspiciously to me, like an attempt to
> couple their client software with their service,
> and I'm rather surprised that people here on
> Slashdot would defend such a practice.
If it *is* such an attempt, it's a pretty poor one, considering the amount of Napster clones around for various other OSes. I've got no idea how you arrived at this conclusion, however: Napster's bot ban is similar to many IRC servers' bot bans: they're not limiting client software to theirs only, they're (presumably) stopping people from automatically searching for Metallica every 30 seconds, 24 hours a day, ie, to reduce load on their servers.
AFAIK Napster has never complained about people cloning their software, either. At least not recently. The conclusion you make is just nonsensical, and the comparisons you draw on aren't valid. Napster isn't limiting you to using a product produced by a certain company; they're specifically *prohibiting* you from using *any* product that performs a certain function (ie, automated searches). Think of it as a negative filter rather than a positive one. To use another analogy, it's the difference between "Don't kill people" (negative filter) and "You must walk down the street with a smile on your face at all times") (positive filter).
Of course, there's sometimes a fine line between a client and a bot. There are many modified Gnutella clients that send back `The RIAA is watching you!' or similar in response to every search, but they - presumably - also function as an otherwise normal
Gnutella client.
What?
So far, I've seen no negative comments. The closest I got was someone asking if the thing was POSIX compliant. Everyone else has seemed excited and interested.
I certainly think you're not alone in wanting to try it. As an avid Linux user for the past verylongtime, I certainly want to.
Just because it's pulled rather than pushed doesn't mean it's not propagation.
When something "Propagates", it just means copies are made and stored in places other than the original location. In the same way that the Slashdot logo has propagated to hundreds of thousands of WWW caches all over the globe.
To the - errm - misguided individual who attempted to disprove the theory of evolution using the 2nd law of thermodynamics: you're a git.
Sure, entropy will constantly increase in a closed system. But the Earth is not a closed system. Life requires energy, which is provided by the Sun.