Linus interprets the GPL for Linux as allowing non-GPL apps and device drivers, and there are a lot of non-GPL apps for Linux.
<speculation>
RMS will interpret the GPL for Hurd as allowing only GPL apps and device drivers. So even if Hurd gets to be big there would never be an Oracle/Hurd etc.
Or in other words Hurd will never be as big a Linux
</speculation>
The thing that really really made Java take of big style way back in 96 was when Microsoft announced they'd support it in their browser. When MS announced support EVERYONE else did within the next few weeks.
If there is an official MS Office for Linux then Linux will officially have "made it" on the desktop. It is the rubber stamp that will boost support for the desktop beyond anything else. When MS support Linux with a version of Office then we will suddenly see hundreds of other apps and drivers pop-up in the me-too rush.
This was news 11 years ago when Manchester University started work on Amulet - an asynch version of the ARM chip.
And much as I think the project is very cool the fact that many PDAs use StrongARMs and not Amulets tells a story.
From the site:
The power/performance qualities of AMULET were sufficiently encouraging to continue the line of investigation. As the performance of AMULET2 still lagged that of contemporary synchronous devices the next task was to improve the MIPS rate, without sacrificing excess power. This has been addressed by AMULET3i. This macrocell can deliver over 100 MIPS on a 0.35m commodity process - and the processor core is capable of more in the future. The first AMULET3i application is the DRACO DECT radio controller. However AMULET3 based systems will be developed for other application area; we're particularly interested in contactless smart cards for reasons which will gradually be made apparent.
100 MIPS - imagine that!
Don't get me wrong - I still think that the asynch idea is cool, but I don't think that Clockless, or asynchronous, chips work very much faster... than their synchronous equivalents
There is no way we are even close to this being viable.
What counts as a page view?
When I have read all the HTML?, What if the graphics don't load, but are vital to the page. What if I write a bot that reads 99% of the page but then closes the socket. What if my browser crashes? There are so many what if's to make it totally impractical from a viewing POV.
Who pays?
At home, how do I separate my kids browsing from mine, and similarly how does a company separate work browsing from allowed spare time browsing. And what about contested payments? Am I expected to compare my browser logs with my bank statement? There are so many what if's to make it totally impractical from a financial POV.
American airlines Airbus A300.
Flying from JFK to the dominican republic
Between 5-10 miles out of JFK, over Queens
At least 4 buildings on fire.
All 3 airports in NY now closed.
No news on accident/terrorist
It won't work on the net well, and there are problems with the idea in the real world too.
In effect the sea floor and positions of sender and reciever are acting as a secret key. They 'encrypt' the messages and you can only decrypt if you know the secret key in enough detail - i.e. you are the reciever, and the working with the sender. However the snooper in *theory* could decode the signal if he knew enough about the sender/reciever/sea bed, and could do some farily complex maths. How complex the maths is says if it will work in practice. But given that computer can model huricanes, I would guess that modeling the sea bed is plauible.
In the virtual world though all bets are off. The terrain is very mappable, and fairly simple. So if the problems of varing ping times can be worked out the encryption is very easily broken.
I wonder if the sea bed version stops working if the tide changes.
So I guess we should all be hoping that the crackers/carders don't have enough money to spend on the fees to join CERT. - Clearly the idea kind of breaks down if we are selling sploits to the bad guys before we tell the good guys how to fix the problem they didn't know they had.
So lets hope the.coms have more money to spare then the carders.;-)
But then again this news is about as relevant as CERT is themselves these days.
The book is about using OSS with Oracle, and not Oracle becoming OSS.
From the page: "In Oracle & Open Source, we concentrate on: Where to get today's open source base technologies and application tools; how to install them; how to connect them to Oracle; and how to modify them, should you wish to do so, to suit your own requirements"
In fact they "didn't try to address the future of open source, particularly in terms of its relationship to commercial software"
But I guess you cant expect the/. editors to check up on the facts can you. And after all you are only here for pro Linux/OSS and anti MS bias aren't you...
From the linked page: Another difference is in the markup languages the two use, i-Mode uses c-HTML which is a subset of HTML while WAP uses WML which is a subset of XML. Although c-HTML is similar to HTML and easier for Web designers to use, XML is considered to be the Internet language of the future as HTML has limited capabilities.
Hmmm.
WML is not a sub-set of XML, it is an implmentation of XML. I think the idea of using XHTML instead of a totally new XML dialect (WML) is a good one. (recent/. story)
The replacement for HTML may one day be XHTML, but not pure XML. XML needs some form of implementation to make it work.
--
I'm not nearly so impressed with this book, as everyone else seems to be. I have 2 issues:
Schneier explains a lot of stuff, like what authentication is, what a private key is and so on, that most geeks will know backwards. I'm willing to bet that anyone that read his Counterpane newsletters will probably not learn a huge amount of new stuff.
He sometimes jumps into geek-speak (for example use of grok) which assumes the reader is a geek. I would guess that most people that know what grok means don't need 80% of the explainations in the book.
I usually love Schneier's stuff, I just think that the market for this book isn't knowledgeable geeks.
The only reasonable way to "test" security is to perform security reviews. This is an expensive, time-consuming, manual process. It's not enough to look at the security protocols and the encryption algorithms. A review must cover specification, design, implementation, source code, operations, and so forth. And just as functional testing cannot prove the absence of bugs, a security review cannot show that the product is in fact secure.
The theory looks sound, but one of the biggest things to do when optimizing any Java program is to minimize the calls to new(). I recently sped a program up by 150% in a snap simply by killing a few new()s. Swing and LotusXSL have had similar experiences.
I think that part of the problem is that all of this is new, so there is more to do. HotSpot the trendy JIT from Sun in places IS ALREADY FASTER THAN C, but whenever it comes to Object creation, things slow down a lot.
So why in theory should new() be quick when in fact it is slow? IMHO the problem is not with the memory claiming, its with all the other stuff that the JVM has to do.
When I call new Foo() the JVM:
Checks to see if the bytecode for Foo already exists, and if not it loads it, verifies it, and calls the class init method. This is very very slow, but should only be done once.
Allocs new memory. Probably very quick
Calls the hierachy of constructors of all Foos superclasses. Quite slow.
(For advanced garbage collectors) Place the object on the 'recent items' list. Probably quick
So I guess the complexity of the system as a whole is the problem here.
It's Christmas, so there is no news, so the Journos need to make some. On top of that this it the first really 'E-ed up' Christmas for most people, so a good scare story is bound to get the hits in.
From my perspective as a purchaser, the hit rate, whilst not 100%, is not low enough to stop me buying online, and is certainly not 25%.
From my perspective working for a financial organization doing E-Commerce, the failure rate I see is very small fraction of the figure quoted.
IMHO, Blackdown are up against big business deals where the suits make decisions and the PR people dust over the cracks later.
Sun seem to have a very tight deal with Inprise - Noticed that JBuilder Enterprise is only available for Solaris? Yet it is 100% Java, so to get it working on Windows all you have to do is follow the/bin/sh install script mentally on your Win box. I know, I've done it. Yet Inprise are prepared to loose all the Windows JBuilder Enterprise customers for a few months by telling people that it is Solaris only. That means Sun and Inprise are deeply in bed.
I wonder if deals like this will feature in a future DoJ vs Sun monopoly case? Granted Sun are not currently as badly behaved as MS, but that is only because they don't have the opportunity. I seem to recall McNally saying a few years ago something along the lines that he approved of Gates' business methods - he'd do the same thing given the option. Maybe I remember it wrong though.
The bit that I don't understand is that Inprise are partly owned by Microsoft (IIRC). So how does that work?
And i want my Java NOW, dammit! So... what sort of licensing hell awaits GPL'd projects that want to use Sun's classes? Yikes!
No Hell.
Using SUN's classes puts you under NO distribution restrictions for your own code. You may not have a licence to redistribute SUN's code, but then anyone that has a JVM has those classes anyway.
... grep the internet looking for remarks made by famous people against OSS, insert a few comments, and mould it into a stick with which to beat ESR, and OSS.
The author slags Linus for being a dictator, and claims Linux is not demorcratic. But the author also slags OSSs need for strong leadership. Personally I think giving the masses a vote on whether driver X implements feature Y in a sensible way is just plain stupid - Democracy does not work everywhere.
More useful would be solutions to the problems which most/.ers know about.
How do you keep s/n ratio up on a dev. list?
How do you keep a core team together?
How do we avoid yet more pointless bitching?
Don't get me wrong I don't think OSS is a silver bullet either, but we need to be a little more constructive in our criticism of it.
<joke> I think we need ESR to vigourously defend himself, and maybe some choice words from Bruce wouldn't go amiss too... </joke>
Unmetered - You just have to know how.
on
ISP War in the UK
·
· Score: 2
http://www.screaming.net have been doing unmetered calls in the UK on evenings and at weekends for a while. Screaming Net is reportedly slow however. (it is a partnership between retailer Tempo, and telco LocalTel, just like Freeserve is a pertnership between Dixons and Energis)
http://www.08004u.co.uk/Are now doing unmetered calls 24x7. However you do have to pay 50UKP per month subscription. One day our telecoms regulator Oftel will get its finger out and sort BT out.
Unmetered - You just have to know how.
on
ISP War in the UK
·
· Score: 0
http://www.screaming.net have been doing unmetered calls in the UK on evenings and at weekends for a while. Screaming Net is reportedly slow however. (it is a partnership between retailer Tempo, and telco LocalTel, just like Freeserve is a pertnership between Dixons and Energis) http://www.08004u.co.uk/Are now doing unmetered calls 24x7. However you do have to pay 50UKP per month subscription. One day our telecoms regulator Oftel will get its finger out and sort BT out.
At least since Echelon both the UK and the USA and all the other countries in the same game have been illegally tapping communications. So I guess it is no surprise that they want to carry on.
The real problem is that most people don't care enough to change the way they vote as a result of this stuff.
2 key problems to solve:
Stopping this sound like a James Bond film - an urbam myth to be ignored.
Getting people to realize the potential dangers of a government that act outside the law.
The current safety may be only because although the NSA/GCHQ/etc act in a way that is above the law, the people that make governing decisions are not to the same extent.
So which is more likely - Clinton/Blair etc finding a way to act illegally without a Starr report, or making the NSA act in a sensible/lawful manner?
Commerce Department yanks ICANN's chain, backhands NSI
On 9 July the Commerce Department sent a 32-page letter [1] to the ICANN board and the House Commerce Committee, responding to committee chairman Tom Bliley's questions on ICANN's recent actions [2]. Here's the NY Times's coverage [3] of this letter (free registration and cookies required). Commerce Department officials said that ICANN should
hold all meetings in public,
drop a proposed $1-per-domain-name fee until a permanent ICANN board can vote on it, and
draw up binding contracts with domain-name services that would bar ICANN from going beyond their mission.
Commerce did not let NSI entirely off the hook, either. While chastising ICANN for a threat, issued in its Berlin meeting, to cancel NSI's authority to issue domain names, the Commerce letter states baldly that unless NSI signs ICANN's operating agreement, Commerce will in fact terminate that authority. NSI must stop at once claiming the.com,.net. and.org domain-name databases as their intellectual property, Commerce insists.
Congress has now scheduled the investigative hearing promised by Bliley. The Subcommittee on Oversight and Investigations will convene "Domain Name System Privatization: Is ICANN Out of Control?" on Thursday, July 22, 1999 at 11:00 a.m. in the Rayburn House Office Building, room 2322.
On 16 July Commerce again extended the deadline [4] for the end of the open domain registration test. The test had already been extended once [5] because of protracted wrangling among NSI, ICANN, and the test registrars. The new target date for wider participation in competitive registration is 6 August.
Slashdot Mirror
<speculation>
RMS will interpret the GPL for Hurd as allowing only GPL apps and device drivers. So even if Hurd gets to be big there would never be an Oracle/Hurd etc.
Or in other words Hurd will never be as big a Linux
</speculation>
Yes.
Do I have any intention of using it?
No.
The thing that really really made Java take of big style way back in 96 was when Microsoft announced they'd support it in their browser. When MS announced support EVERYONE else did within the next few weeks.
If there is an official MS Office for Linux then Linux will officially have "made it" on the desktop. It is the rubber stamp that will boost support for the desktop beyond anything else. When MS support Linux with a version of Office then we will suddenly see hundreds of other apps and drivers pop-up in the me-too rush.
And much as I think the project is very cool the fact that many PDAs use StrongARMs and not Amulets tells a story.
From the site:
The power/performance qualities of AMULET were sufficiently encouraging to continue the line of investigation. As the performance of AMULET2 still lagged that of contemporary synchronous devices the next task was to improve the MIPS rate, without sacrificing excess power. This has been addressed by AMULET3i. This macrocell can deliver over 100 MIPS on a 0.35m commodity process - and the processor core is capable of more in the future. The first AMULET3i application is the DRACO DECT radio controller. However AMULET3 based systems will be developed for other application area; we're particularly interested in contactless smart cards for reasons which will gradually be made apparent.
100 MIPS - imagine that!
Don't get me wrong - I still think that the asynch idea is cool, but I don't think that Clockless, or asynchronous, chips work very much faster ... than their synchronous equivalents
What counts as a page view?
When I have read all the HTML?, What if the graphics don't load, but are vital to the page. What if I write a bot that reads 99% of the page but then closes the socket. What if my browser crashes? There are so many what if's to make it totally impractical from a viewing POV.
Who pays?
At home, how do I separate my kids browsing from mine, and similarly how does a company separate work browsing from allowed spare time browsing. And what about contested payments? Am I expected to compare my browser logs with my bank statement? There are so many what if's to make it totally impractical from a financial POV.
In short:
It won't work.
Summary:
American airlines Airbus A300.
Flying from JFK to the dominican republic
Between 5-10 miles out of JFK, over Queens
At least 4 buildings on fire.
All 3 airports in NY now closed.
No news on accident/terrorist
In effect the sea floor and positions of sender and reciever are acting as a secret key. They 'encrypt' the messages and you can only decrypt if you know the secret key in enough detail - i.e. you are the reciever, and the working with the sender. However the snooper in *theory* could decode the signal if he knew enough about the sender/reciever/sea bed, and could do some farily complex maths. How complex the maths is says if it will work in practice. But given that computer can model huricanes, I would guess that modeling the sea bed is plauible.
In the virtual world though all bets are off. The terrain is very mappable, and fairly simple. So if the problems of varing ping times can be worked out the encryption is very easily broken.
I wonder if the sea bed version stops working if the tide changes.
Soon they'll decide if the shouting is too loud and abandon the idea ..., or not.
The same thing happened with
XP subscriptions: http://slashdot.org/articles/01/05/06/0038258.shtm l
Spamming: http://slashdot.org/articles/00/09/28/1341249.shtm l
Passport: http://slashdot.org/yro/00/07/29/1228209.shtml --
So lets hope the .coms have more money to spare then the carders. ;-)
But then again this news is about as relevant as CERT is themselves these days.
From the page: "In Oracle & Open Source, we concentrate on: Where to get today's open source base technologies and application tools; how to install them; how to connect them to Oracle; and how to modify them, should you wish to do so, to suit your own requirements"
In fact they "didn't try to address the future of open source, particularly in terms of its relationship to commercial software"
But I guess you cant expect the /. editors to check up on the facts can you. And after all you are only here for pro Linux/OSS and anti MS bias aren't you ...
Not to knock the work being done in China, but they are not the first Maglev.
Birmingham (UK, not AL) between the NEC exhibition centre and the train station. I'd guess it has been there for over 10 years. They announced plans in the middle of last year for a version 2.
--
Now might be a good time to d/l all those service packs ;-)
--
Another difference is in the markup languages the two use, i-Mode uses c-HTML which is a subset of HTML while WAP uses WML which is a subset of XML. Although c-HTML is similar to HTML and easier for Web designers to use, XML is considered to be the Internet language of the future as HTML has limited capabilities.
Hmmm. /. story)
WML is not a sub-set of XML, it is an implmentation of XML. I think the idea of using XHTML instead of a totally new XML dialect (WML) is a good one. (recent
The replacement for HTML may one day be XHTML, but not pure XML. XML needs some form of implementation to make it work. --
I'm not nearly so impressed with this book, as everyone else seems to be. I have 2 issues:
I usually love Schneier's stuff, I just think that the market for this book isn't knowledgeable geeks.
--
Order in which JDK 1.3 is/will be released for various platforms:
Has to be a tad embarrasing for Sun!
--
... Security is not something that can be tested for.
Makes sense if you think about it. And it blows a truck right through the "you need a formal spec to test against" premise.
I think Schnier makes much more sense from a theoretical point of view.
From http://www.counterpane.com/crypto- gram-9911.html
The only reasonable way to "test" security is to perform security reviews. This is an expensive, time-consuming, manual process. It's not enough to look at the security protocols and the encryption algorithms. A review must cover specification, design, implementation, source code, operations, and so forth. And just as functional testing cannot prove the absence of bugs, a security review cannot show that the product is in fact secure.
No mention of a formal spec.
Go Bruce!
I recently sped a program up by 150% in a snap simply by killing a few new()s. Swing and LotusXSL have had similar experiences.
I think that part of the problem is that all of this is new, so there is more to do. HotSpot the trendy JIT from Sun in places IS ALREADY FASTER THAN C, but whenever it comes to Object creation, things slow down a lot.
So why in theory should new() be quick when in fact it is slow? IMHO the problem is not with the memory claiming, its with all the other stuff that the JVM has to do.
When I call new Foo() the JVM:
- Checks to see if the bytecode for Foo already exists, and if not it loads it, verifies it, and calls the class init method.
- Allocs new memory. Probably very quick
- Calls the hierachy of constructors of all Foos superclasses. Quite slow.
- (For advanced garbage collectors) Place the object on the 'recent items' list. Probably quick
So I guess the complexity of the system as a whole is the problem here.This is very very slow, but should only be done once.
From my perspective as a purchaser, the hit rate, whilst not 100%, is not low enough to stop me buying online, and is certainly not 25%.
From my perspective working for a financial organization doing E-Commerce, the failure rate I see is very small fraction of the figure quoted.
Sun seem to have a very tight deal with Inprise - Noticed that JBuilder Enterprise is only available for Solaris? Yet it is 100% Java, so to get it working on Windows all you have to do is follow the /bin/sh install script mentally on your Win box. I know, I've done it. Yet Inprise are prepared to loose all the Windows JBuilder Enterprise customers for a few months by telling people that it is Solaris only. That means Sun and Inprise are deeply in bed.
I wonder if deals like this will feature in a future DoJ vs Sun monopoly case? Granted Sun are not currently as badly behaved as MS, but that is only because they don't have the opportunity.
I seem to recall McNally saying a few years ago something along the lines that he approved of Gates' business methods - he'd do the same thing given the option. Maybe I remember it wrong though.
The bit that I don't understand is that Inprise are partly owned by Microsoft (IIRC). So how does that work?
No Hell.
Using SUN's classes puts you under NO distribution restrictions for your own code. You may not have a licence to redistribute SUN's code, but then anyone that has a JVM has those classes anyway.
So writing GPLed Java code is legally ok.
IANAL of course.
The author slags Linus for being a dictator, and claims Linux is not demorcratic. But the author also slags OSSs need for strong leadership.
Personally I think giving the masses a vote on whether driver X implements feature Y in a sensible way is just plain stupid - Democracy does not work everywhere.
More useful would be solutions to the problems which most /.ers know about.
Don't get me wrong I don't think OSS is a silver bullet either, but we need to be a little more constructive in our criticism of it.
<joke> I think we need ESR to vigourously defend himself, and maybe some choice words from Bruce wouldn't go amiss too ... </joke>
Screaming Net is reportedly slow however. (it is a partnership between retailer Tempo, and telco LocalTel, just like Freeserve is a pertnership between Dixons and Energis)
http://www.08004u.co.uk/Are now doing unmetered calls 24x7. However you do have to pay 50UKP per month subscription. One day our telecoms regulator Oftel will get its finger out and sort BT out.
http://www.screaming.net have been doing unmetered calls in the UK on evenings and at weekends for a while. Screaming Net is reportedly slow however. (it is a partnership between retailer Tempo, and telco LocalTel, just like Freeserve is a pertnership between Dixons and Energis) http://www.08004u.co.uk/Are now doing unmetered calls 24x7. However you do have to pay 50UKP per month subscription. One day our telecoms regulator Oftel will get its finger out and sort BT out.
The real problem is that most people don't care enough to change the way they vote as a result of this stuff.
2 key problems to solve:
The current safety may be only because although the NSA/GCHQ/etc act in a way that is above the law, the people that make governing decisions are not to the same extent.
So which is more likely - Clinton/Blair etc finding a way to act illegally without a Starr report, or making the NSA act in a sensible/lawful manner?
Echelon Links, And More, And More
There is no mention of any nice Linux extras, just that it uses RH6.0.
Commerce Department yanks ICANN's chain, backhands NSI
On 9 July the Commerce Department sent a 32-page letter [1] to the ICANN board and the House Commerce Committee, responding to committee chairman Tom Bliley's questions on ICANN's recent actions [2]. Here's the NY Times's coverage [3] of this letter (free registration and cookies required). Commerce Department officials said that ICANN should
hold all meetings in public,
drop a proposed $1-per-domain-name fee until a permanent ICANN board can vote on it, and
draw up binding contracts with domain-name services that would bar ICANN from going beyond their mission.
Commerce did not let NSI entirely off the hook, either. While chastising ICANN for a threat, issued in its Berlin meeting, to cancel NSI's authority to issue domain names, the Commerce letter states baldly that unless NSI signs ICANN's operating agreement, Commerce will in fact terminate that authority. NSI must stop at once claiming the .com, .net. and .org domain-name databases as their intellectual property, Commerce insists.
Congress has now scheduled the investigative hearing promised by Bliley. The Subcommittee on Oversight and Investigations will convene "Domain Name System Privatization: Is ICANN Out of Control?" on Thursday, July 22, 1999 at 11:00 a.m. in the Rayburn House Office Building, room 2322.
On 16 July Commerce again extended the deadline [4] for the end of the open domain registration test. The test had already been extended once [5] because of protracted wrangling among NSI, ICANN, and the test registrars. The new target date for wider participation in competitive registration is 6 August.
[1] http://www.ntia.doc.gov/n tiahome/domainname/blileyrsp.htme t.html
[2] http://www.news.com/N ews/Item/Textonly/0,25,38200,00.html?pfv
[3] http://www.ny times.com/library/tech/99/07/biztech/articles/10n
[4] http://www.zdnet.co m/zdnn/filters/bursts/0,3422,2295115,00.html
[5] http://tbtf.com/archive/1999-07-08.html #s01
TBTF Is required reading for anyone with a clue.