> Hard disks that are faster, not bigger. If I need more space, I'll add more spindles. How about giving me a disk
> that can push 50 or 100 MB/sec from the platters?
Me, I'd like to get rid of HDD entirely: they are slow, fragile and noisy..
MRAM would be perfect, but I doubt that we will have gigabytes of MRAM anytime soon:-(
> Cheap SMP. I'll take my dual 550 over a single 1 GHz any day of the week. How about 8x500 MHz on the
> desktop, instead of 1x4GHz which is still crippled by 1 CPU hogging app?
You will have "soon" some kind of parallelism on your desktop: it won't be SMP but SMT (or in Intel buzzword-speaking: "hyperthreading").
While we're talking about changing unit definitions..
In France we're using the word octet which means eight bits.
I think that octet is interesting because
- it is always 8 bits. A byte is usually 8 bits but not always.
- as it abreviates to KO, MO (or now KiO and MiO) you have less risks of confusion between kilobytes and kilobits..
It is effective, but it is UGLY (and that's an understatement).
It is so ugly that words cannot convey the meaning: it is as ugly as the "bombing" of the world trade center..
Excuse me not to get so excited about a plane designed for carpet bombing..
Carpet bombing is IMPRECISE so there are many "colateral damages", an military term for innocent civilians ie also innocent children, women and men mutilated and killed..
> I work for a Major Automotive Manufacturer's ITM department, and Perl has saved our lives.
Yes, Perl is a usefull tool, but it isn't the only one..
Why couldn't have you used Python instead?
Being more readable than C/C++ is not an impresive feat!
I hate Java myself: too much bugs in the standard library, the design of Swing is nice but its implementation is buggy and slow.
> s. You can write some really horrible, incomprehensible stuff in English, if you want to. But you can also write Shakespere. French - even the French of Voltaire and Hugo - is still pretty much French.
Mmm, the French slang (l'argot) can be quite hard to understand too and I'm French!
But frankly the fact that now English rules and that French is less and less spoken has NOTHING to do about which language is the more flexible!
It is a consequence of the economic domination of the USA..
If in the future China rule the world (economicaly speaking) then people would more and more use Chinese..
In France, you can't buy nazy items, cannot say things pro-racism..
I'm french so I'm problably LESS aware of the things forbidden in France which are legal elsewhere..
Is oral and anal sex still "banned" in some states of the USA?
I've always found funny than the "land of freedom" would ban those things..
Australia is banning a videogame ??
What is "big news" here??
So explain to me, why Alan Cox and the other kernel hackers have installed some code in order to detect when a kernel has a non-free driver used?
Because they refuse to troubleshoot problems caused by NVidia non-free drivers (as said by Alan Cox himself).
The driver is working for you, and to be honest, I had once a TNT (bought because I believed that the driver was going to be open-source) and the driver was also working for me.
But is-it working with SMP? or some other bizarre situations?
I wouldn't bet on it!
When a light beam is used to modify another light beam, it is usually made with some materials which reacts to lights.
The material will be heated by the light, so it will generate heats..
Light doesn't interact with light directly, so a light-based chip would be really a light - non-linear material - light chip.
Usually the interesting effect in those materials used to modulate light are only a "second order" effect, which means that you have to use quite intense lights to have something usefull.
Intense light --> heat.
No, thanks:-)
French WERE catholics would be far closer to the truch..
Now a large part of French are atheist/agnostic, there is only 60% of french people who call themselves catholic, and only 10% of French goes to church from time to time..
So catholicism is still a strong force in France, but it is going weaker and weaker: the young generation is even less catholic..
L4 is fast, but I believe that it has some limitations on the number of threads, size of messages, etc..
These limits were made to go faster..
The port of Hurd over the L4 microkernel is complicated due to these restrictions..
> I'm horrified at how IDE has flourished. It's the worst possible standard for a drive interface.
Now let's see, where I have seen this before?
- 80x86 one of the worst ISA (braindead floating points ISA, too few registers, unduly complicated) won against all the other because it was cheaper.. IBM considered going with the 68000, but it was quite expensive..
- Microsoft vs Apple: Windows won, they were cheaper and still able to get the job done..
Who earns more money now?
Do you see a trend here?
As long as it get the job done, the cheapest technology will win, even if it is "ugly" from a technical point of view..
>About Word...
>Destroying the on disk copy of a document
>before successfully writing out the new copy is
>just plain stupid.
[]
>You write the new copy under a fake name,
> swap it atomically (even over file severs)
>with the original file, then delete the fake
>named file (which now contains the old data).
And what happens when you're editing a file on a floppy disk?
There is some good chance that you won't be able to have both the old and the new version of the file at the same time on the floppy disk..
So while I agree that programs should go out of their way to make sure that you don't ever loose data, I beleive that it is VERY difficult to do it right..
Couldn't it be targeted also at scientific computing?
A lot of these code are in Fortran, but C is also used..
I doubt very much that targetting games developer on Linux will get you very far, but multimedia processing could be interesting though.
But there is competition of course, from other compilers of course, but also from libraries: a good hand-optimised library can get you quite interesting performances..
Basically you're saying that X should move to a higher level, no?
I agree and maybe you would be interested to have a look at Berlin.
See http://www.berlin-consortium.org
It is a work in progress not ready at all for use, but its design is interesting..
They didn't choose to evolve X but instead they use the Fresco framework for basis.
They have a press release, see here: http://research.sun.com/features/async/
(I'm sorry, I can't use HTML: the lameness filter don't want to allow the posting otherwise.)
I imagine the "perfect" laptop:
- an OLED screen (no need for backlighting)
- an asynchronous processor (low power)
- no HDD, but plenty of MRAM (this RAM is persistent)
>Java handles backward compatibility by
>'depreciation'. That's what the original poster
>meant. I think you knew it.
Yes in theory Java is backward compatible by tagging old interface 'deprecated' which cause warnings at the compilation..
In my practice, it doesn't work very well: during the 1.1.x -> 1.2.0 transition they added a new printing interface and 'deprecated' the old interface.
The old interface was supposed to work, unfortunately it didn't and the new printing interface was incroyably slow and memory consuming..
The result? You couldn't print anything moderately complex with 1.2.0 and the much needed corrections of the rest were only on 1.2.0..
Well you know there are intermediate steps: usually other RISC architecture have either 32 or 64 registers..
And having 128 registers has other drawbacks that long context switches: there is a size versus speed trade-off..
These problems get only worse when you use SMT: a four-way SMT CPU needs four-times more register at least..
This plus the in-order design:I don't believe that Intel will ever use SMT for the IA-64, more likely they'll go CMT..
A part from the "mine is bigger than yours" factor I'm wondering what kind of speed-up you have when you go from 64 to 128 number of registers..
It is probably very limited apart from very specials problems..
> It's a combination windowing system with
> toolkits for a consisten user interface?
>...and I thought X was bloated. No thanks, guys.
The X server is quite lightweight, but the clients are not: think at the number of toolkit you use simultaneously: Qt, GTK+, Tk, Lesstiff..
This is memory bloat!!
Worse, those toolkit has usually some troubles working with the others: cut-copy-paste problems sometimes, poor look&feel integration etc..
And think about communications between the client and the server:
- with X you have LOTS of very low level communications between the client and the server (draw a rectangle here, etc..).
Have you done XLib programming?
If no, you'd be surprised to see how many events the X server send to the clients..
- with Berlin, usually a client would use higher level primitives that the server which would manage: less bandwith usage, improved latency.
X main's advantage is that it works now, but I feel that Berlin's design is cleaner IMHO.
I'm French and like everywhere the governement has a very "fuzzy" attitude.
On one hand, the government has created an agency for improving the usage of "logiciels libres" withing the administrations, but on the other hand the government is on the verge of supporting the creation of software's patents, thanks to the intense lobbying of the big industrials..
There is a petition to fight against the creation of software patents in Europe, but apparently it has no effect:-(
Usually VM are stack-based because it is simpler, I think.
But the Tao-VM (used by these "Amiga" guys) is based on an unbounded set of register, I heard that they claim that it is the fastest VM..
Have you looked at the Tao-VM ?
I remember of flamewars in rec.comp.arch about register-based architecture vs stack-based architecture of CPU.
I think that there is the same design issue for VM..
You're kidding right?
X has some nice capabilities but I wouldn't call it nicely designed.
This is this "nice design" which explain that I have usually 3 toolkit in memory at the same time?? (Qt, Lesstiff, GTK, Tk sometimes, etc.)
Talk about unnecessary memory usage! And these toolkits don't integrates together very well (look and feel, sometimes cut/copy/paste doesn't work very,etc.)
And have you looked at XLib programming have you seen how low level the communication are between the client and the server?? How "chatty" the communication is between the client and the server??
Berlin is well designed (but not ready for prime time), X design is at best so-so.
>>Realistically, no hardware vendor could afford to do without Microsoft.
> Thats bullshit as well! Anyone of the large vendors could go head to head with MS any day of the week.
> IBM was prepared to do it, but chickened out at the last second.
>Compaq had at the time revenues easily topping that of MS.
>Dell is a freaking-gigantic monolith.
Your assertion are pointless: there is a cutthroat competition between PC hardware makers!!
The day one of those hardware makers make something which goes against Microsoft, his rebate on Microsoft software would be suppressed and instantly its PC sold with Microsoft software would be higher priced than those of its competitors: he would be dead in no time (or more likely he would have to do what Microsoft wants him to do to regain its rebate).
So even if the revenues of the PC makers are above those of Microsoft, they are very vulnerable to Microsoft decisions because of
1) the competition between PC makers
2) the Microsoft monopoly
3) the Microsoft rebates
I really hope that Be will sue Microsoft, IMHO they have a really strong point so Be should win..
But it's just my opinion of course, Microsoft have so much money and power that I suspect that there won't be any outcome of a trial: if they see that they will loose the trial, they would go for an out-of-court settlement..
Slashdot Mirror
The TAO VM (the VM which is used in the "new" Amiga OS) is register based (with an "infinite" number of registers).
So the 68k emulator is not the only VM to be register based..
Now, I've never seen interesting paper which compares both approach, I don't think that one approach is necessarily better than the other..
PS (off-topic):
The story in the acme journal linked in the FAQ is really impressive..
http://use.perl.org/~acme/journal
> Hard disks that are faster, not bigger. If I need more space, I'll add more spindles. How about giving me a disk
:-(
> that can push 50 or 100 MB/sec from the platters?
Me, I'd like to get rid of HDD entirely: they are slow, fragile and noisy..
MRAM would be perfect, but I doubt that we will have gigabytes of MRAM anytime soon
> Cheap SMP. I'll take my dual 550 over a single 1 GHz any day of the week. How about 8x500 MHz on the
> desktop, instead of 1x4GHz which is still crippled by 1 CPU hogging app?
You will have "soon" some kind of parallelism on your desktop: it won't be SMP but SMT (or in Intel buzzword-speaking: "hyperthreading").
While we're talking about changing unit definitions..
In France we're using the word octet which means eight bits.
I think that octet is interesting because
- it is always 8 bits. A byte is usually 8 bits but not always.
- as it abreviates to KO, MO (or now KiO and MiO) you have less risks of confusion between kilobytes and kilobits..
It is effective, but it is UGLY (and that's an understatement).
It is so ugly that words cannot convey the meaning: it is as ugly as the "bombing" of the world trade center..
Excuse me not to get so excited about a plane designed for carpet bombing..
Carpet bombing is IMPRECISE so there are many "colateral damages", an military term for innocent civilians ie also innocent children, women and men mutilated and killed..
I don't see your points.
> I work for a Major Automotive Manufacturer's ITM department, and Perl has saved our lives.
Yes, Perl is a usefull tool, but it isn't the only one..
Why couldn't have you used Python instead?
Being more readable than C/C++ is not an impresive feat!
I hate Java myself: too much bugs in the standard library, the design of Swing is nice but its implementation is buggy and slow.
> s. You can write some really horrible, incomprehensible stuff in English, if you want to. But you can also write Shakespere. French - even the French of Voltaire and Hugo - is still pretty much French.
Mmm, the French slang (l'argot) can be quite hard to understand too and I'm French!
But frankly the fact that now English rules and that French is less and less spoken has NOTHING to do about which language is the more flexible!
It is a consequence of the economic domination of the USA..
If in the future China rule the world (economicaly speaking) then people would more and more use Chinese..
In France, you can't buy nazy items, cannot say things pro-racism..
I'm french so I'm problably LESS aware of the things forbidden in France which are legal elsewhere..
Is oral and anal sex still "banned" in some states of the USA?
I've always found funny than the "land of freedom" would ban those things..
Australia is banning a videogame ??
What is "big news" here??
[]
>The NVidia card "just works" with Linux
[]
So explain to me, why Alan Cox and the other kernel hackers have installed some code in order to detect when a kernel has a non-free driver used?
Because they refuse to troubleshoot problems caused by NVidia non-free drivers (as said by Alan Cox himself).
The driver is working for you, and to be honest, I had once a TNT (bought because I believed that the driver was going to be open-source) and the driver was also working for me.
But is-it working with SMP? or some other bizarre situations?
I wouldn't bet on it!
When a light beam is used to modify another light beam, it is usually made with some materials which reacts to lights.
The material will be heated by the light, so it will generate heats..
Light doesn't interact with light directly, so a light-based chip would be really a light - non-linear material - light chip.
Usually the interesting effect in those materials used to modulate light are only a "second order" effect, which means that you have to use quite intense lights to have something usefull.
Intense light --> heat.
You said "French are catholics".
:-)
No, thanks
French WERE catholics would be far closer to the truch..
Now a large part of French are atheist/agnostic, there is only 60% of french people who call themselves catholic, and only 10% of French goes to church from time to time..
So catholicism is still a strong force in France, but it is going weaker and weaker: the young generation is even less catholic..
AFAIK BeOS has not a microkernel.
L4 is fast, but I believe that it has some limitations on the number of threads, size of messages, etc..
These limits were made to go faster..
The port of Hurd over the L4 microkernel is complicated due to these restrictions..
> I'm horrified at how IDE has flourished. It's the worst possible standard for a drive interface.
Now let's see, where I have seen this before?
- 80x86 one of the worst ISA (braindead floating points ISA, too few registers, unduly complicated) won against all the other because it was cheaper.. IBM considered going with the 68000, but it was quite expensive..
- Microsoft vs Apple: Windows won, they were cheaper and still able to get the job done..
Who earns more money now?
Do you see a trend here?
As long as it get the job done, the cheapest technology will win, even if it is "ugly" from a technical point of view..
Do you know some recent benchmarks between the FreeBSD filesystem with softupdates and Linux with ReiserFS or XFS or Ext3FS?
Or are you just making up the "better performance than journaling" ?
Such rather strong assertion needs proofs, not just theoretical justifications..
At its beginning the Alpha was a pure speed daemon, a very "simple" (in order, simple dispatching) but with a very high clock design.
Then its design became more brainiac, now it is an out of order design: they choose to increase the Instruction Level Parallelism over the frequency.
So the Alpha reduced its advance in clock speed..
>About Word...
>Destroying the on disk copy of a document
>before successfully writing out the new copy is
>just plain stupid.
[]
>You write the new copy under a fake name,
> swap it atomically (even over file severs)
>with the original file, then delete the fake
>named file (which now contains the old data).
And what happens when you're editing a file on a floppy disk?
There is some good chance that you won't be able to have both the old and the new version of the file at the same time on the floppy disk..
So while I agree that programs should go out of their way to make sure that you don't ever loose data, I beleive that it is VERY difficult to do it right..
You have an "auto-vectorising" compiler, good.
Couldn't it be targeted also at scientific computing?
A lot of these code are in Fortran, but C is also used..
I doubt very much that targetting games developer on Linux will get you very far, but multimedia processing could be interesting though.
But there is competition of course, from other compilers of course, but also from libraries: a good hand-optimised library can get you quite interesting performances..
Basically you're saying that X should move to a higher level, no?
I agree and maybe you would be interested to have a look at Berlin.
See http://www.berlin-consortium.org
It is a work in progress not ready at all for use, but its design is interesting..
They didn't choose to evolve X but instead they use the Fresco framework for basis.
I just wanted to add that this technique while very nice is not entirely free.
The "voltage spike" used to lower the response time means that there is an increase in power consumption (sp?).
So laptop users may not want this feature enabled while they are traveling..
Of course it depends it the increase of power consumption is large or not..
They have a press release, see here: http://research.sun.com/features/async/
(I'm sorry, I can't use HTML: the lameness filter don't want to allow the posting otherwise.)
I imagine the "perfect" laptop:
- an OLED screen (no need for backlighting)
- an asynchronous processor (low power)
- no HDD, but plenty of MRAM (this RAM is persistent)
>Java handles backward compatibility by
>'depreciation'. That's what the original poster
>meant. I think you knew it.
Yes in theory Java is backward compatible by tagging old interface 'deprecated' which cause warnings at the compilation..
In my practice, it doesn't work very well: during the 1.1.x -> 1.2.0 transition they added a new printing interface and 'deprecated' the old interface.
The old interface was supposed to work, unfortunately it didn't and the new printing interface was incroyably slow and memory consuming..
The result? You couldn't print anything moderately complex with 1.2.0 and the much needed corrections of the rest were only on 1.2.0..
Talk about good backward compatibility!!
Well you know there are intermediate steps: usually other RISC architecture have either 32 or 64 registers..
And having 128 registers has other drawbacks that long context switches: there is a size versus speed trade-off..
These problems get only worse when you use SMT: a four-way SMT CPU needs four-times more register at least..
This plus the in-order design:I don't believe that Intel will ever use SMT for the IA-64, more likely they'll go CMT..
A part from the "mine is bigger than yours" factor I'm wondering what kind of speed-up you have when you go from 64 to 128 number of registers..
It is probably very limited apart from very specials problems..
> It's a combination windowing system with ...and I thought X was bloated. No thanks, guys.
> toolkits for a consisten user interface?
>
The X server is quite lightweight, but the clients are not: think at the number of toolkit you use simultaneously: Qt, GTK+, Tk, Lesstiff..
This is memory bloat!!
Worse, those toolkit has usually some troubles working with the others: cut-copy-paste problems sometimes, poor look&feel integration etc..
And think about communications between the client and the server:
- with X you have LOTS of very low level communications between the client and the server (draw a rectangle here, etc..).
Have you done XLib programming?
If no, you'd be surprised to see how many events the X server send to the clients..
- with Berlin, usually a client would use higher level primitives that the server which would manage: less bandwith usage, improved latency.
X main's advantage is that it works now, but I feel that Berlin's design is cleaner IMHO.
I'm French and like everywhere the governement has a very "fuzzy" attitude.
:-(
On one hand, the government has created an agency for improving the usage of "logiciels libres" withing the administrations, but on the other hand the government is on the verge of supporting the creation of software's patents, thanks to the intense lobbying of the big industrials..
There is a petition to fight against the creation of software patents in Europe, but apparently it has no effect
Usually VM are stack-based because it is simpler, I think.
But the Tao-VM (used by these "Amiga" guys) is based on an unbounded set of register, I heard that they claim that it is the fastest VM..
Have you looked at the Tao-VM ?
I remember of flamewars in rec.comp.arch about register-based architecture vs stack-based architecture of CPU.
I think that there is the same design issue for VM..
You're kidding right?
X has some nice capabilities but I wouldn't call it nicely designed.
This is this "nice design" which explain that I have usually 3 toolkit in memory at the same time?? (Qt, Lesstiff, GTK, Tk sometimes, etc.)
Talk about unnecessary memory usage! And these toolkits don't integrates together very well (look and feel, sometimes cut/copy/paste doesn't work very,etc.)
And have you looked at XLib programming have you seen how low level the communication are between the client and the server?? How "chatty" the communication is between the client and the server??
Berlin is well designed (but not ready for prime time), X design is at best so-so.
>>Realistically, no hardware vendor could afford to do without Microsoft.
> Thats bullshit as well! Anyone of the large vendors could go head to head with MS any day of the week.
> IBM was prepared to do it, but chickened out at the last second.
>Compaq had at the time revenues easily topping that of MS.
>Dell is a freaking-gigantic monolith.
Your assertion are pointless: there is a cutthroat competition between PC hardware makers!!
The day one of those hardware makers make something which goes against Microsoft, his rebate on Microsoft software would be suppressed and instantly its PC sold with Microsoft software would be higher priced than those of its competitors: he would be dead in no time (or more likely he would have to do what Microsoft wants him to do to regain its rebate).
So even if the revenues of the PC makers are above those of Microsoft, they are very vulnerable to Microsoft decisions because of
1) the competition between PC makers
2) the Microsoft monopoly
3) the Microsoft rebates
I really hope that Be will sue Microsoft, IMHO they have a really strong point so Be should win..
But it's just my opinion of course, Microsoft have so much money and power that I suspect that there won't be any outcome of a trial: if they see that they will loose the trial, they would go for an out-of-court settlement..