I'm all against stupid translation (for those who don't know in France our academy has created the word cederome which is to be used instead of CD-ROM, yes I know it is incredibly stupid).
But the problem is that the english word "free" is ambiguous(sp?), so it may be better to use a non-ambigous term: "libre". Here, the translated word is ~superior to the original as it has a well defined meaning.
So why not use it in English too, American people has no problem when it comes to using foreign word, which unfortunately is not the case in France...
They didn't earn any money with the communicator because they couldn't sell it, thanks to Microsoft's competition.
So they had no problem using a very friendly licence, but Qt is the main product of the company which is selling it, so don't expect this anytime soon.
Here is a little summary of the reaction posted in the forum: - general reaction: great, good idea
- many people think that open software is good but that what is important is "standard format" ie using HTML/XML instead of.doc for exchanging document for example.
* for me, both is important and I do hope that KOffice and AbiWord will use the same file format in the future. I think that it is more important than the Gnome/KDE (in)compatibility.
- a minor nitpicking of the article 3 of the proposal: it would ban every copyrighted software, so even the GPL would be banned also.
* I think that it is only an error in the way the proposal is worded: Senators are seldom computer experts.
- some minor opposition: some are fearing that open software could mean job loss. * I find it quite amusing because the job threatened here are mostly on Microsoft' side, but anyway this is a weak conterargument (see the bazaar thingy).
As a French, what do I think about this proposal ? Well, I'm enthousiastly(sp?) for. It could boost free software, helping to improve it, the governement has enough money to spend on fixing the "rusty" parts of free software AND it would still be cheaper than buying closed source software.
BTW, in France we don't have the problem which keeps showing in English: does free means free as free speech or free as free beer, because we use two different word for it. Free as free speech is "libre" and free as free beer is "gratuit".
So why not talking about libre software or libertad software if we use the spanish word (after all spanish is the most spoken language, I think) ?
EROS seems to have a nice security model and have automic persistency support, it is GPL, small and seems to be quite fast according the small benchmark which has been made.
I don't know about it's real time capabilities though...
The guys who are working on HURD could, may be take a look at EROS, and report here what they think about it, this should be very interesting.
> [] but I think it's alive and kicking on the > client too. JDK 1.3 is going to make it even > better, with the incorporated HotSpot engine, > which I've already seen to speed up client-side > apps.
I thought that the HotSpot thing could only accelerate Server-side Java ?
The JDK1.3 will be really nice, because it will suppress many important bugs (you wouldn't believe how buggy Java is now, especially on the client side).
With fewer bugs, may be increased speed (note that in the volano benchmark, the much touted HotSpot comes behind the IBM JVM, so it is not the magic bullet, but anything that can help increasing speed is welcome), Java may become usable on the client side with the JDK 1.3...
PS: If you have enough memory, on the client side, Java is a pig, it needs lots of memory just to do very simple GUI.
Who designed a processor with a new CISC ISA recently? Nobody, everyone is designing RISC ISA or post-RISC ISA ie VLIW or EPIC or...
The fact that one CISC ISA, the 80x86 have managed to stay alive and well (:-)) means that backwards compatibility is sometimes more interesting than raw performance and is an amazing feat done by Intel (helped very much by Windows success).
Except maybe when movies talk about well-known topic, movies doesn't even try to be acurate, they try to give you cheap thrills... (especially Holliwood movies?)
If you would believe what movies say about skydive, you would believe that : a) you can talk to someone while falling b) a freefall last 5 minutes when you exit a plane a 4000 meters c) there is a low pull contest between skydivers d) every skydiver is mad and dirty .....
I'm sure that everyone who knows a particular topic is shocked to see how movies treat it. Come on, you really think that policemen usually do what you see on the screen ?
No, but movies have created some kind of myth about the police and another one about computers.
You are more shocked by the latter because you are more knowledgeable in computers, that's all...
I'm not an NT advocate, but please don't forget that even if NT works now only on Intel platform, the HAL is still useful because now on Intel platform, there is (will be) two ISA : 80x86 and the Itanium/Merced/EPIC ISA.
What I do find strange is that Microsoft had some trouble porting NT 2000 to the Itanium, I would think that thanks to the HAL, it would be very simple, or perhaps it is the different pointer size ? But isn't this kind of thing, that the HAL is supposed to hide ?
I'm (a bit) puzzled, anyone knowledgeable could answer ?
My original language is not english, but most of the time (99%), I have no problem at all reading all the post in/. except your articles which always seems very fuzzy and very long.
In France we say: - 'Ce qui se concoit bien, s'enonce clairement et les mots pour le dire viennent aisement' which means: - what you understand well, should be said clearly so as to be easily understandable.
I don't know if American and English people can easily read your articles, but please try to to think a little bit about us foreigner.
So 1) if it's quite clear in your mind what you want to say, try to express it clearly, make an effort to be easily readable. 2) if what you are trying to say isn't clear in your mind, why are you posting ?:-) Clarify your mind first and then go to step 1.:-):-)
The difference between Microsoft and Gnome is that Microsoft do much less release, their release are numbered by the year of the release.
Gnome release much more often, so they have to use the month! Why not say something like "The October Gnome Release v1.0.35", it gives more informations and it isn't that much longer.
BTW, I really do like the Linux version numbering scheme, odd subnumber for developer release and even subnumber for stable release. A good idea TM.
Frankly, they should also use version number, it is much more easy to follow the evolution of a product based on its version number (minor enhancement ? increase minor number) than on the month of the release...
Maybe they don't use version number because they fucked up the 1.0.0 version which should have been labeled 0.9.9 to say that it was feature complete but not yet stable. But that doesn't mean that using version number is a bad thing, it just mean that you have to be cautious about your version number..
To summarise the way Slashdot function: 1) someone gives a pointer on an 'interesting' facts 2)/. readers comments the news, discuss. There are very knowledgeable post, insightful post and totally shitty posts. 3) thanks to the moderation system, you can have a really good S/N ratio, if you're in a hurry or not too interested by the topic you can just choose your moderation level so as to keep just a few post.
Jane's will use the comment of Slashdot readers to modify/re-create an article on the subject, an interesting move...
I remenber that one time, there was a series of post about XWindows, what's wrong with it, should it be fixed or replaced by Berlin or something else... While it was an interesting topic, the sheer number of interesting post was too much to read even with the moderation system... And I hope that someday someone will have the guts to make a compilation on the topic.
IMHO community writing or summarisation of particularly interesting topics is what could take Slashdot to a new level, being even more interesting.
I wonder what will happen: an interesting topic show up, lots of interesting posts happens, a courageous man make a summary of the good points and post it on Slashdot on a new topic, so lots of posts be posted again... etc, etc. A nice recursive loop, which may generate a reference FAQ on this particular topic.
from Andrew: > Quantum computing. [] All electronic data transmission becomes insecure.
Bzzzt, only the exchange based on "traditional" cryptographic technique will become insecure.
Do not forget, that now some scientist are developing what you could name "quantum cryptography" whose security is based upon the law of physic as we know them...
So when quantum computers finally becomes available, I would expect that the "quantum cryptography" to be already used.
> Those space planes will be very important for fast, secure communication. 2020-2025 The NSA might already have this. Frankly, this looks stupid and paranoid, I know that prediction of the future shouldn't be taken too seriously but WTF?
And remenber the story on/. which said that Turing (a genius, by all means) predicted that in 2000 we would have a machine who would pass the Turing test... This shows that those kind of predictions have little value.
Oh well, let's dream a little bit, in 1990 I've read a book which predicted that the first universal molecal assembler would be build in 30 years, that's 2020. So let's wait and see...
>> there are no commercially proven clustering >> technologies to provide High Availability for Linux. > Can you say beowulf?
Mmm, I thought beowulf was a "computation farm" type cluster not a fail-over type cluster, am I mistaking ?
That's sad but I don't think that the reaction here in/. are objective, it is not because the article from Microsoft is biased that/. readers have to be biased as well. I think that a much better reaction would be a) filter the false from the truth b) when there is a real shortcoming in Linux c) show if this shortcoming is going to be fixed or not and when.
The article state that Linux: 1) lacks a journaling file system:it is worked on currently with the Ext3 FS, the JFS from SGI, but right now, it is true. 2) can't use more than 2 GB of memory : I think that this restriction will be partially lifted on 2.4, but right now it is true for the stable kernels. 3) can't do true asynchronous IO. Is-it true ? If yes, I think that it is really a shortcoming of Linux as it is a feature which can be important if it is needed...
I think that a page which would show the shortcomings of Linux and wheter they will be fixed and when is needed, the problem is that it is really hard to do from an unbiased point of view.
Don't misunderstand me, I like Linux, but I really don't like the game Linux is better than NT, it is stupid, Linux has to be good in itself and a good way to improve a product is to know its "defaults" and working hard in order to fix those...
[] > By comparison, metric units tend to be either too large or too small:
That's funny: I've never heard someone complain about the metric system here in France with units being either too small or too large. And no, I don't think that it is because it was invented here.
The only reason of your problem is that you are not used to the metrical system, that's why you feel it is "unnatural", trust me I don't have at all this problem.
> With computers to help us, there's no real reason to HAVE to change anymore.
It is not a computer problem but a human problem... What do you think ? That at the NASA, they don't use computers.
> Converting would be very expensive, cause many more NASA-type foul-ups, and offer little or nothing in return.
Any change is painful, but after the change is completed, it would prevent this kind of error.
> It makes about as much sense for the US to convert to Metric units as it does for the rest of the world to adopt the English language.
A language is different from a metric system, it is easier to change the metric system that the whole language, and BTW what is the language I'm typing in right now ? Esperanto ? No, it's English so in a way the rest of the world IS adopting English.
I don't claim at all, that the metric system is the perfect system, but if look at it without the NIH syndrome, it is quite coherent, and useful. (me i would like to have a system in which the speed of light is 1:-)
Oh, BTW I do remenber another mistake from the NASA when they used terrestrial miles instead of nautical miles (or vice versa) while they were trying to rescue a spaceship...
In Europe, we are on the verge of changing of currency unit, it won't be easy to adjust to the Euro, but it'm not too worried. The US could change its unit system if there was a real will.
> This bug that you refer to about slow printing is referenced in the Java Developer Connection's Bug Parade.
Yes, I know, trust me, I used to check the Bug Parade quite often to work around bugs in the JDK and even submited two bugs. (or one ? can't remenber). I found it very useful to work around the bugs.
I used the "fix" (disabling the double buffering before printing), but still printing was slow, and it was really hard to format the bill the way I wanted... Maybe I shouldn't have used the JDK directly but there wasn't any IDE with the Java2 platform.
Moral of the story, do not believe the hype from Sun and magazines even techie one such as Dr Dobbs.
Maybe sometimes, Java will mature into a usable tool for making applications (and not applets), but before I try it again, this time I will check carefuly the bug parade...
You couldn't have used a worst example: Printing with Java.
I've tried to used Java, six month ago for a small shop. My problems began when I tried to print bills, with the JDK 1.1.7, the printing system was crude to say the least, and I didn't manage to generate "not too ugly" bills so as Java2 just went out and reviews said that it was really good, I decided to use it.
It turned out that the printing wasn't compatible with the old way, so I decided to use the new printing API whose design was much better, but it turned out that the implementation was quite poor: it was sloooooow, a memory pig and sometimes it just didn't worked.
So maybe you can print raw text with java, but even with Java 2.0 you couldn't print even simple page.
For the GUI, AWT design is quite poor, Swing is much better but it is a bit slow and memory intensive. There is a bug in the Java bug's parade which was there for more than a year the last time I checked: users were furious.
So for the two example, you have given Java is quite poor...
I've been burned badly by Java and won't use it again for a long time: too many bugs.
I don't understand at all, how it is possible that magazine are so stupid to hype any novelty without talking about its downside. There was a fairly thorough review of Java2 in Dr Dobbs, and they didn't talk about these problem with printing and said that it was totally upward compatible...:-((
Slashdot Mirror
He was always naming its robots with a surname even if the production name of hte robot was quite stern.
I'm all against stupid translation (for those who don't know in France our academy has created the word cederome which is to be used instead of CD-ROM, yes I know it is incredibly stupid).
But the problem is that the english word "free" is ambiguous(sp?), so it may be better to use a non-ambigous term: "libre".
Here, the translated word is ~superior to the original as it has a well defined meaning.
So why not use it in English too, American people has no problem when it comes to using foreign word, which unfortunately is not the case in France...
They didn't earn any money with the communicator because they couldn't sell it, thanks to Microsoft's competition.
So they had no problem using a very friendly licence, but Qt is the main product of the company which is selling it, so don't expect this anytime soon.
Here is a little summary of the reaction posted in the forum:
.doc for exchanging document for example.
- general reaction: great, good idea
- many people think that open software is good but that what is important is "standard format" ie using HTML/XML instead of
* for me, both is important and I do hope that KOffice and AbiWord will use the same file format in the future. I think that it is more important than the Gnome/KDE (in)compatibility.
- a minor nitpicking of the article 3 of the proposal: it would ban every copyrighted software, so even the GPL would be banned also.
* I think that it is only an error in the way the proposal is worded: Senators are seldom computer experts.
- some minor opposition: some are fearing that open software could mean job loss.
* I find it quite amusing because the job threatened here are mostly on Microsoft' side, but anyway this is a weak conterargument (see the bazaar thingy).
As a French, what do I think about this proposal ?
Well, I'm enthousiastly(sp?) for.
It could boost free software, helping to improve it, the governement has enough money to spend on fixing the "rusty" parts of free software AND it would still be cheaper than buying closed source software.
BTW, in France we don't have the problem which keeps showing in English: does free means free as free speech or free as free beer, because we use two different word for it.
Free as free speech is "libre" and free as free beer is "gratuit".
So why not talking about libre software or libertad software if we use the spanish word (after all spanish is the most spoken language, I think) ?
EROS seems to have a nice security model and have automic persistency support, it is GPL, small and seems to be quite fast according the small benchmark which has been made.
I don't know about it's real time capabilities though...
The guys who are working on HURD could, may be take a look at EROS, and report here what they think about it, this should be very interesting.
Offtopic and stupid
Flamebait and stupid
ROFL
Why I am not a moderator when I would like to be ?
Moderator, please moderate the precedent post as funny (or sad if you are this poor guy)
> [] but I think it's alive and kicking on the
> client too. JDK 1.3 is going to make it even
> better, with the incorporated HotSpot engine,
> which I've already seen to speed up client-side
> apps.
I thought that the HotSpot thing could only accelerate Server-side Java ?
The JDK1.3 will be really nice, because it will suppress many important bugs (you wouldn't believe how buggy Java is now, especially on the client side).
With fewer bugs, may be increased speed (note that in the volano benchmark, the much touted HotSpot comes behind the IBM JVM, so it is not the magic bullet, but anything that can help increasing speed is welcome), Java may become usable on the client side with the JDK 1.3...
PS: If you have enough memory, on the client side, Java is a pig, it needs lots of memory just to do very simple GUI.
Did you read the article or Hennesy & Patterson ?
The CISC ISA are very different from the VLIW/EPIC ISAs.
> 1) this guy can't financially support himself, much less a family. ("cool" adds nothing to a job resume, you know.)
This seem to imply that women should better go for the rich men only.
Who designed a processor with a new CISC ISA recently? Nobody, everyone is designing RISC ISA or post-RISC ISA ie VLIW or EPIC or ...
The fact that one CISC ISA, the 80x86 have managed to stay alive and well (:-)) means that backwards compatibility is sometimes more interesting than raw performance and is an amazing feat done by Intel (helped very much by Windows success).
Except maybe when movies talk about well-known topic, movies doesn't even try to be acurate, they try to give you cheap thrills... (especially Holliwood movies?)
If you would believe what movies say about skydive, you would believe that :
a) you can talk to someone while falling
b) a freefall last 5 minutes when you exit a plane a 4000 meters
c) there is a low pull contest between skydivers
d) every skydiver is mad and dirty
.....
I'm sure that everyone who knows a particular topic is shocked to see how movies treat it.
Come on, you really think that policemen usually do what you see on the screen ?
No, but movies have created some kind of myth about the police and another one about computers.
You are more shocked by the latter because you are more knowledgeable in computers, that's all...
I'm not an NT advocate, but please don't forget that even if NT works now only on Intel platform, the HAL is still useful because now on Intel platform, there is (will be) two ISA : 80x86 and the Itanium/Merced/EPIC ISA.
What I do find strange is that Microsoft had some trouble porting NT 2000 to the Itanium, I would think that thanks to the HAL, it would be very simple, or perhaps it is the different pointer size ? But isn't this kind of thing, that the HAL is supposed to hide ?
I'm (a bit) puzzled, anyone knowledgeable could answer ?
My original language is not english, but most of the time (99%), I have no problem at all reading all the post in /. except your articles which always seems very fuzzy and very long.
:-) Clarify your mind first and then go to step 1. :-) :-)
In France we say:
- 'Ce qui se concoit bien, s'enonce clairement et les mots pour le dire viennent aisement' which means:
- what you understand well, should be said clearly so as to be easily understandable.
I don't know if American and English people can easily read your articles, but please try to to think a little bit about us foreigner.
So 1) if it's quite clear in your mind what you want to say, try to express it clearly, make an effort to be easily readable.
2) if what you are trying to say isn't clear in your mind, why are you posting ?
No, seriously, you should try to be more readable
The difference between Microsoft and Gnome is that Microsoft do much less release, their release are numbered by the year of the release.
Gnome release much more often, so they have to use the month! Why not say something like "The October Gnome Release v1.0.35", it gives more informations and it isn't that much longer.
BTW, I really do like the Linux version numbering scheme, odd subnumber for developer release and even subnumber for stable release. A good idea TM.
Frankly, they should also use version number, it is much more easy to follow the evolution of a product based on its version number (minor enhancement ? increase minor number) than on the month of the release...
Maybe they don't use version number because they fucked up the 1.0.0 version which should have been labeled 0.9.9 to say that it was feature complete but not yet stable. But that doesn't mean that using version number is a bad thing, it just mean that you have to be cautious about your version number..
To summarise the way Slashdot function: /. readers comments the news, discuss. There are very knowledgeable post, insightful post and totally shitty posts.
... etc, etc.
1) someone gives a pointer on an 'interesting' facts
2)
3) thanks to the moderation system, you can have a really good S/N ratio, if you're in a hurry or not too interested by the topic you can just choose your moderation level so as to keep just a few post.
Jane's will use the comment of Slashdot readers to modify/re-create an article on the subject, an interesting move...
I remenber that one time, there was a series of post about XWindows, what's wrong with it, should it be fixed or replaced by Berlin or something else... While it was an interesting topic, the sheer number of interesting post was too much to read even with the moderation system... And I hope that someday someone will have the guts to make a compilation on the topic.
IMHO community writing or summarisation of particularly interesting topics is what could take Slashdot to a new level, being even more interesting.
I wonder what will happen: an interesting topic show up, lots of interesting posts happens, a courageous man make a summary of the good points and post it on Slashdot on a new topic, so lots of posts be posted again
A nice recursive loop, which may generate a reference FAQ on this particular topic.
from Andrew:
/. which said that Turing (a genius, by all means) predicted that in 2000 we would have a machine who would pass the Turing test... This shows that those kind of predictions have little value.
> Quantum computing. [] All electronic data transmission becomes insecure.
Bzzzt, only the exchange based on "traditional" cryptographic technique will become insecure.
Do not forget, that now some scientist are developing what you could name "quantum cryptography" whose security is based upon the law of physic as we know them...
So when quantum computers finally becomes available, I would expect that the "quantum cryptography" to be already used.
> Those space planes will be very important for fast, secure communication. 2020-2025 The NSA might already have this.
Frankly, this looks stupid and paranoid, I know that prediction of the future shouldn't be taken too seriously but WTF?
And remenber the story on
Oh well, let's dream a little bit, in 1990 I've read a book which predicted that the first universal molecal assembler would be build in 30 years, that's 2020. So let's wait and see...
>> there are no commercially proven clustering
/. are objective, it is not because the article from Microsoft is biased that /. readers have to be biased as well.
:it is worked on currently with the Ext3 FS, the JFS from SGI, but right now, it is true.
>> technologies to provide High Availability for Linux.
> Can you say beowulf?
Mmm, I thought beowulf was a "computation farm" type cluster not a fail-over type cluster, am I mistaking ?
That's sad but I don't think that the reaction here in
I think that a much better reaction would be a) filter the false from the truth b) when there is a real shortcoming in Linux c) show if this shortcoming is going to be fixed or not and when.
The article state that Linux:
1) lacks a journaling file system
2) can't use more than 2 GB of memory : I think that this restriction will be partially lifted on 2.4, but right now it is true for the stable kernels.
3) can't do true asynchronous IO. Is-it true ? If yes, I think that it is really a shortcoming of Linux as it is a feature which can be important if it is needed...
I think that a page which would show the shortcomings of Linux and wheter they will be fixed and when is needed, the problem is that it is really hard to do from an unbiased point of view.
Don't misunderstand me, I like Linux, but I really don't like the game Linux is better than NT, it is stupid, Linux has to be good in itself and a good way to improve a product is to know its "defaults" and working hard in order to fix those...
> Rachael :) )
> (This is my first ever post, so be gentle
Not bad for a first post: Score:4, Insightful
:-)
I've already seen a project where the design was really polished, but the implementation was done by a beginner which was more or less on her own.
The result was of course a piece of sh...
Was it her fault ? No, of course, we all began one day and all beginners should be closely supervised with his code reviewed, etc, etc.
To summarize: design is important, successfull projects do spend a lot of time in the design phase but good design alone is NOT enough...
[]
:-)
> By comparison, metric units tend to be either too large or too small:
That's funny: I've never heard someone complain about the metric system here in France with units being either too small or too large. And no, I don't think that it is because it was invented here.
The only reason of your problem is that you are not used to the metrical system, that's why you feel it is "unnatural", trust me I don't have at all this problem.
> With computers to help us, there's no real reason to HAVE to change anymore.
It is not a computer problem but a human problem... What do you think ? That at the NASA, they don't use computers.
> Converting would be very expensive, cause many more NASA-type foul-ups, and offer little or nothing in return.
Any change is painful, but after the change is completed, it would prevent this kind of error.
> It makes about as much sense for the US to convert to Metric units as it does for the rest of the world to adopt the English language.
A language is different from a metric system, it is easier to change the metric system that the whole language, and BTW what is the language I'm typing in right now ? Esperanto ? No, it's English so in a way the rest of the world IS adopting English.
I don't claim at all, that the metric system is the perfect system, but if look at it without the NIH syndrome, it is quite coherent, and useful. (me i would like to have a system in which the speed of light is 1
Oh, BTW I do remenber another mistake from the NASA when they used terrestrial miles instead of nautical miles (or vice versa) while they were trying to rescue a spaceship...
In Europe, we are on the verge of changing of currency unit, it won't be easy to adjust to the Euro, but it'm not too worried. The US could change its unit system if there was a real will.
> This bug that you refer to about slow printing is referenced in the Java Developer Connection's Bug Parade.
...
Yes, I know, trust me, I used to check the Bug Parade quite often to work around bugs in the JDK and even submited two bugs. (or one ? can't remenber). I found it very useful to work around the bugs.
I used the "fix" (disabling the double buffering before printing), but still printing was slow, and it was really hard to format the bill the way I wanted... Maybe I shouldn't have used the JDK directly but there wasn't any IDE with the Java2 platform.
Moral of the story, do not believe the hype from Sun and magazines even techie one such as Dr Dobbs.
Maybe sometimes, Java will mature into a usable tool for making applications (and not applets), but before I try it again, this time I will check carefuly the bug parade
You couldn't have used a worst example: Printing with Java.
:-((
I've tried to used Java, six month ago for a small shop. My problems began when I tried to print bills, with the JDK 1.1.7, the printing system was crude to say the least, and I didn't manage to generate "not too ugly" bills so as Java2 just went out and reviews said that it was really good, I decided to use it.
It turned out that the printing wasn't compatible with the old way, so I decided to use the new printing API whose design was much better, but it turned out that the implementation was quite poor: it was sloooooow, a memory pig and sometimes it just didn't worked.
So maybe you can print raw text with java, but even with Java 2.0 you couldn't print even simple page.
For the GUI, AWT design is quite poor, Swing is much better but it is a bit slow and memory intensive. There is a bug in the Java bug's parade which was there for more than a year the last time I checked: users were furious.
So for the two example, you have given Java is quite poor...
I've been burned badly by Java and won't use it again for a long time: too many bugs.
I don't understand at all, how it is possible that magazine are so stupid to hype any novelty without talking about its downside.
There was a fairly thorough review of Java2 in Dr Dobbs, and they didn't talk about these problem with printing and said that it was totally upward compatible...