I've been very happy with the performance of Firefox 3 beta 3 already, and added JavaScript performance is certainly nice. This will mean I'll try to use my VIA EPIA as desktop again, it's 1.2 GHz processor slowed down the internet experience just enough to make it the lesser option.
Now to restore flash again, it's still dysfunctional, and it is starting to get on my nerves. Sites that require flash are annoying, but not annoying enough that I never want to visit them again. Oh well, we'll have to wait for the official version I guess.
It's not likely that hordes of Slashdotters all over the planet are crawling out of their basements to flock to this restaurant wearing their neatest t-shirts now is it?
Eclipse displays these keys in the menu's and drop down menu's. No need to look at the keyboard for that. And their keyboard map is very nice as well, hell, it displays when you hit shift-ctrl-l, it displays this in the menu, and their preferences page is brilliant. And you can use this functionality for free if you make an Eclipse framework based application.
If you don't use a shortcut enough to remember its position, what use is the keyboard shortcut anyway? Use the menus.
"But if the stuff printed on each key changed when you press the Ctrl key? The user will be exposed to so much more functionality! And that's not even mentioning Function keys, or modal software (like vi), or..."
And what are you going to do, try every combination until you find copy or paste? Of course not, the answer to this is to display the shortcuts when an action is taken. This action would normally be a mouse click in a menu or after the menu pops up after a right mouse click. A nice screen with all the shortcuts displayed also helps.
The problem is that many programs won't display these shortcuts, mainly because it is not provided by the GUI builders as a default option. If you would use the Eclipse framework to build an application, you get this and much other functionality basically for free. This is how it should be.
Finding keyboard shortcuts by pressing the keys is like trying to find a CLI command in the shell without being able to search the manual pages. Besides, if these programs won't display the keyboard shortcuts in the menu's, they probably won't be displayed on this keyboard either.
What this keyboard might be good at is for displaying special characters after pressing an alternative alt key. Maybe. And possibly games (which would have to be written for it).
This is what smart card IC's do although it is more scrambling than actual encryption (at least most of the time, many manufacturers don't release the specs for their implementation).
Bah, that's the same as saying that you only need assembler to be able to create any application. Sure, you can make a demo program, and with enough time you can do pretty astounding things with it, but it breaks down for any modern application development. Of course you can argue that students need some basic CLI handling and C++ classes, but lets not carried away and stop showing them the applications that most programmers/designers use.
I used UltraEdit + the JDK for the start of my career, and back then you didn't have the IDE functionality that we're relying on now. Then came along Eclipse 2.1 and I've never looked back. I presume it will be the case for most C++ engineers when they switch to Visual Studio or Eclipse for their needs. There are just way too many advantages.
It took a long time to start up on my company PC as well. But this was a McAfee infected PC, which had enabled on access scanning on read operations. And it scanned in ZIP files as well. This seems to include all jars and javadoc archives. Disabling McAfee for the java folder and zip files removed the long wait, and the minute long waits at 100% CPU whilst I was typing (Eclipse displays javadoc for methods that you are trying to use in real time).
It now also does not delete > 600 MB zipped backup files with a single infected file in it somewhere (in my "INFECTED, DO NOT EXECUTE" folder that was backed up as well). Gods, I hate McAfee.
The thing we need is encryption in the CPU. VIA and Sun already provide processors with build in encryption (and several embedded processors can do the same). It is really time for AMD and Intel to provide support for symmetric primitives, if possible with a well designed, documented and compatible instructions and registers (the one provided by VIA has some drawbacks, I think). A hardware key store that can be used by these instructions might also be an interesting option, although some people would no doubt argue against that (since it might enable much stronger DRM).
Of course, it would be important for the applications to be able to make use of this interface. The most standardized way is to use a PKCS#11 interface, which would in turn use some driver or device. Applications could then use either the low level stuff, the PKCS#11 interface (a dynamic library) or, of course, openssl. TrueCrypt would probably opt for the low level interface. This would probably require some rework since they've opted to program their own version of AES. And their are two pitfalls: a crypto processor on the CPU means that you need to stream all the data to the CPU, and hard disk encryption normally uses a specific AES mode of operation that might be a bit harder to support.
I believe AMD is currently looking into ways of embedding crypto into their processors. It's really weird this has not happened yet since, in my opinion, it is one of the easier ways to speed up current computers. It might also do wonders for their server sales (https/ssl/tls).
While they're at it, why not add a few commonly used codecs (within encryption) such as base-64 or even LZH? Check out the numbers posted on the VIA and Sun sites to have an idea of what a HW capable CPU might be capable of. These numbers are always best case scenarios, but these CPU's can saturate a harddisk and a gigabit network interface, even at the same time.
Sometimes the scale of this planet amazes me. But maybe that's because I'm from the Netherlands, which is one of the most populated places in the world. It can be pretty hard to go somewhere and not meet someone (especially in Amsterdam, where I live:) ).
Breaking news: Weird sparking coming from basements all over the US.
People have reported that sparking and loud screams have been reported all over the US. It has been thought that a posting on Slashdot caused the sparking when thousands of nerds drooled over their laptop computers. In other unrelated news, the amount of rugged laptop sales has skyrocketed.
I'm a bit afraid that I will be loosing most of the benefits of a dual CPU (better GUI response) when applications are going to start using all those cores. Even now, if the applications are run from the same disk things get really messy. Of course, flash SSD's may alleviate some of that last problem.
I've had it with my video screwing up when running an IO or CPU sensitive task. The problem is that prioritizing tasks - for both CPU and IO - on current O/S does not really work, and it is starting to hurt.
They don't have to enforce this, they will just have to rely on the fact that watching horror movies does not generate any side effects. How are you going to tell if those Chinese people are watching horror movies or not? If watching horror movies generate no visible side effect, why would you ban them? Well, publicity of course.
In my opinion, the whole idea that you would download the thing is idiotic. I mean, somebody - probably unauthenticated at that time - sends me a text file, and my HTML/XML implementation is downloading files from the URL's in the text file? It could point to a huge XML file and my implementation would happily start to download the thing. It could even point to some child porn for all that matter. It would be much better if the default would be to lookup a file in the filesystem (cached from disk) for the specific URI (a hashtable comes to mind).
I've even seem this problem with XML signature in the Java libraries. The default implementations will even be stupid enough to download the XML schema from the internet! And then the problem becomes that you cannot specify which algorithms etc. the file should adhere to. These are all things you, as a developer, should think about, according to the builders of the libraries.
As long as the libraries and documentation doesn't take care of these major issues, don't expect the developers to think about them, that's all I'm saying. Anyway, why would we rewrite those handlers over and over? That's the idea of libraries, isn't it?
"It can also be used for free publicity: require a "sponsored by [company]" tag somewhere in exchange for releasing the code."
Then again, if it is crappy coded, it might backfire. And if you stick your name to it you might be harassed a bit by users of the code. But I could imagine that it might be possible to get this done, and I imagine it could be worth it for the company concerned. It would also matter if the employee is important within the company, and if he's proven he can handle such responsibility (the coder should *really* ask himself this question first).
I agree with another poster is that starting of by writing letters is not the best thing to do. Try to get hold of the person responsible at a birthday celebration or other informal meeting would be the best place to start things off.
As for the GP, it seems this is exactly the right place to get advice (if only informally) of lawyers. Unfortunately I'm not one, otherwise it would be three in a row:)
"I can pay the wages of a gas pump attendant for 10 years for the price of one gas pumping robot."
Yes, and the pump attendant will be there 24/7? So you might need about 4 pump attendants, and hope they don't get ill. Not that I am necessarily in favor of robots like these, but there seems to be something wrong with your calculations. And, come to think of it, your wages.
I'm waiting for several years now for a system that is completely silent, uses very low power and does not heat my room. And can be used and accessed all the time. And of course, one that does not make the performance penalties that VIA makes in their current EPIA offerengs (otherwise I would be there).
Fortunately this seem to be going to happen in the very near future. Chipsets and CPU's are partially powering down where ever possible, and with a flash SSD's there is no spin-up or (loud) rattling when an indexing service turns on. With voltage scaling like this, there is no need for a separate low power CPU within such a system. Screens are already powering up very quickly and OLED screens are on the horizon as well.
Who needs all those problems with hibernate and suspend if your system is on all the time, while staying below a few volts?
Good, I'm switching off my main computer now, I cannot sleep with the fans and hard disk rattling away. I'll have to read the replies tomorrow after powering up for a few minutes (sleep & hibernate are broken due to some peripherals).
Because although the items in the list are quite interesting, the list isn't. It's simply a list of items that the author encountered and thought of as interesting enough to put in the list. So enjoy the stories in the list and purge the list itself from memory. Especially the mention of the ark of course, since there is very little evidence that the thing existed in the first place, let alone that it had such special capabilities.
Come to think of it, the vials containing fluid blood of St Januarius in an Italian church were more of a gadget, since they definitely contained interesting chemical compounds (but, as expected, probably not blood).
If you look at the screen shot you see two different threads running in parallel, so I guess we are currently out of luck. It would be great if they could make them run in sequence indeed. Nothing breaks down hard disk speeds and responsiveness like two simultaneous copies. Of course, you might want to check which devices are being used by a copy. If you have a copy reading/writing/dev/hda and another copy reading/writing/dev/hdb, you might still want to run those in parallel. But even if they cannot do that, pretty please give me a choice to do those copies in sequence.
Sleepyhead, we were talking WAN instead of LAN, not WLAN. Strangely enough, it might not matter. VPN is not just point to point, it can be used to create a secure virtual LAN as well. Actually, that's in the name, isn't it.
Slashdot Mirror
I've been very happy with the performance of Firefox 3 beta 3 already, and added JavaScript performance is certainly nice. This will mean I'll try to use my VIA EPIA as desktop again, it's 1.2 GHz processor slowed down the internet experience just enough to make it the lesser option.
Now to restore flash again, it's still dysfunctional, and it is starting to get on my nerves. Sites that require flash are annoying, but not annoying enough that I never want to visit them again. Oh well, we'll have to wait for the official version I guess.
It's not likely that hordes of Slashdotters all over the planet are crawling out of their basements to flock to this restaurant wearing their neatest t-shirts now is it?
True, but the older SRAM really lived up to its name, I could restart the computer and be pretty sure that everything would still be there.
Eclipse displays these keys in the menu's and drop down menu's. No need to look at the keyboard for that. And their keyboard map is very nice as well, hell, it displays when you hit shift-ctrl-l, it displays this in the menu, and their preferences page is brilliant. And you can use this functionality for free if you make an Eclipse framework based application.
If you don't use a shortcut enough to remember its position, what use is the keyboard shortcut anyway? Use the menus.
"But if the stuff printed on each key changed when you press the Ctrl key? The user will be exposed to so much more functionality! And that's not even mentioning Function keys, or modal software (like vi), or..."
And what are you going to do, try every combination until you find copy or paste? Of course not, the answer to this is to display the shortcuts when an action is taken. This action would normally be a mouse click in a menu or after the menu pops up after a right mouse click. A nice screen with all the shortcuts displayed also helps.
The problem is that many programs won't display these shortcuts, mainly because it is not provided by the GUI builders as a default option. If you would use the Eclipse framework to build an application, you get this and much other functionality basically for free. This is how it should be.
Finding keyboard shortcuts by pressing the keys is like trying to find a CLI command in the shell without being able to search the manual pages. Besides, if these programs won't display the keyboard shortcuts in the menu's, they probably won't be displayed on this keyboard either.
What this keyboard might be good at is for displaying special characters after pressing an alternative alt key. Maybe. And possibly games (which would have to be written for it).
It says Netscape 6.2 as *minimum* and the forms run quite well on Firefox. There are enough examples that are truly screwed, why not point to them?
Yes, it's called Static RAM (not synchronous RAM) and it, err, does that.
This is what smart card IC's do although it is more scrambling than actual encryption (at least most of the time, many manufacturers don't release the specs for their implementation).
Bah, that's the same as saying that you only need assembler to be able to create any application. Sure, you can make a demo program, and with enough time you can do pretty astounding things with it, but it breaks down for any modern application development. Of course you can argue that students need some basic CLI handling and C++ classes, but lets not carried away and stop showing them the applications that most programmers/designers use.
I used UltraEdit + the JDK for the start of my career, and back then you didn't have the IDE functionality that we're relying on now. Then came along Eclipse 2.1 and I've never looked back. I presume it will be the case for most C++ engineers when they switch to Visual Studio or Eclipse for their needs. There are just way too many advantages.
It took a long time to start up on my company PC as well. But this was a McAfee infected PC, which had enabled on access scanning on read operations. And it scanned in ZIP files as well. This seems to include all jars and javadoc archives. Disabling McAfee for the java folder and zip files removed the long wait, and the minute long waits at 100% CPU whilst I was typing (Eclipse displays javadoc for methods that you are trying to use in real time).
It now also does not delete > 600 MB zipped backup files with a single infected file in it somewhere (in my "INFECTED, DO NOT EXECUTE" folder that was backed up as well). Gods, I hate McAfee.
"And their are two pitfalls"
OMG, I really have to stop reading Slashdot. Did I just type "OMG"? Darn, it might be too late...
The thing we need is encryption in the CPU. VIA and Sun already provide processors with build in encryption (and several embedded processors can do the same). It is really time for AMD and Intel to provide support for symmetric primitives, if possible with a well designed, documented and compatible instructions and registers (the one provided by VIA has some drawbacks, I think). A hardware key store that can be used by these instructions might also be an interesting option, although some people would no doubt argue against that (since it might enable much stronger DRM).
Of course, it would be important for the applications to be able to make use of this interface. The most standardized way is to use a PKCS#11 interface, which would in turn use some driver or device. Applications could then use either the low level stuff, the PKCS#11 interface (a dynamic library) or, of course, openssl. TrueCrypt would probably opt for the low level interface. This would probably require some rework since they've opted to program their own version of AES. And their are two pitfalls: a crypto processor on the CPU means that you need to stream all the data to the CPU, and hard disk encryption normally uses a specific AES mode of operation that might be a bit harder to support.
I believe AMD is currently looking into ways of embedding crypto into their processors. It's really weird this has not happened yet since, in my opinion, it is one of the easier ways to speed up current computers. It might also do wonders for their server sales (https/ssl/tls).
Wish list:
- streamed md5, sha-1 & 2 + whirlpool
- streamed 3des & aes (allowing for different operational modes)
- dh, rsa & ecc (Montgomery multiplier)
- true RNG
- keystore
While they're at it, why not add a few commonly used codecs (within encryption) such as base-64 or even LZH? Check out the numbers posted on the VIA and Sun sites to have an idea of what a HW capable CPU might be capable of. These numbers are always best case scenarios, but these CPU's can saturate a harddisk and a gigabit network interface, even at the same time.
That's a lot of pictures, but this one is the most telling one:
:) ).
http://rides.webshots.com/photo/1292364026025974737Jrroac
Sometimes the scale of this planet amazes me. But maybe that's because I'm from the Netherlands, which is one of the most populated places in the world. It can be pretty hard to go somewhere and not meet someone (especially in Amsterdam, where I live
Breaking news: Weird sparking coming from basements all over the US.
People have reported that sparking and loud screams have been reported all over the US. It has been thought that a posting on Slashdot caused the sparking when thousands of nerds drooled over their laptop computers. In other unrelated news, the amount of rugged laptop sales has skyrocketed.
I'm a bit afraid that I will be loosing most of the benefits of a dual CPU (better GUI response) when applications are going to start using all those cores. Even now, if the applications are run from the same disk things get really messy. Of course, flash SSD's may alleviate some of that last problem.
I've had it with my video screwing up when running an IO or CPU sensitive task. The problem is that prioritizing tasks - for both CPU and IO - on current O/S does not really work, and it is starting to hurt.
They don't have to enforce this, they will just have to rely on the fact that watching horror movies does not generate any side effects. How are you going to tell if those Chinese people are watching horror movies or not? If watching horror movies generate no visible side effect, why would you ban them? Well, publicity of course.
In my opinion, the whole idea that you would download the thing is idiotic. I mean, somebody - probably unauthenticated at that time - sends me a text file, and my HTML/XML implementation is downloading files from the URL's in the text file? It could point to a huge XML file and my implementation would happily start to download the thing. It could even point to some child porn for all that matter. It would be much better if the default would be to lookup a file in the filesystem (cached from disk) for the specific URI (a hashtable comes to mind).
I've even seem this problem with XML signature in the Java libraries. The default implementations will even be stupid enough to download the XML schema from the internet! And then the problem becomes that you cannot specify which algorithms etc. the file should adhere to. These are all things you, as a developer, should think about, according to the builders of the libraries.
As long as the libraries and documentation doesn't take care of these major issues, don't expect the developers to think about them, that's all I'm saying. Anyway, why would we rewrite those handlers over and over? That's the idea of libraries, isn't it?
"It can also be used for free publicity: require a "sponsored by [company]" tag somewhere in exchange for releasing the code."
:)
Then again, if it is crappy coded, it might backfire. And if you stick your name to it you might be harassed a bit by users of the code. But I could imagine that it might be possible to get this done, and I imagine it could be worth it for the company concerned. It would also matter if the employee is important within the company, and if he's proven he can handle such responsibility (the coder should *really* ask himself this question first).
I agree with another poster is that starting of by writing letters is not the best thing to do. Try to get hold of the person responsible at a birthday celebration or other informal meeting would be the best place to start things off.
As for the GP, it seems this is exactly the right place to get advice (if only informally) of lawyers. Unfortunately I'm not one, otherwise it would be three in a row
"I can pay the wages of a gas pump attendant for 10 years for the price of one gas pumping robot."
Yes, and the pump attendant will be there 24/7? So you might need about 4 pump attendants, and hope they don't get ill. Not that I am necessarily in favor of robots like these, but there seems to be something wrong with your calculations. And, come to think of it, your wages.
Great!
I'm waiting for several years now for a system that is completely silent, uses very low power and does not heat my room. And can be used and accessed all the time. And of course, one that does not make the performance penalties that VIA makes in their current EPIA offerengs (otherwise I would be there).
Fortunately this seem to be going to happen in the very near future. Chipsets and CPU's are partially powering down where ever possible, and with a flash SSD's there is no spin-up or (loud) rattling when an indexing service turns on. With voltage scaling like this, there is no need for a separate low power CPU within such a system. Screens are already powering up very quickly and OLED screens are on the horizon as well.
Who needs all those problems with hibernate and suspend if your system is on all the time, while staying below a few volts?
Good, I'm switching off my main computer now, I cannot sleep with the fans and hard disk rattling away. I'll have to read the replies tomorrow after powering up for a few minutes (sleep & hibernate are broken due to some peripherals).
Because although the items in the list are quite interesting, the list isn't. It's simply a list of items that the author encountered and thought of as interesting enough to put in the list. So enjoy the stories in the list and purge the list itself from memory. Especially the mention of the ark of course, since there is very little evidence that the thing existed in the first place, let alone that it had such special capabilities.
Come to think of it, the vials containing fluid blood of St Januarius in an Italian church were more of a gadget, since they definitely contained interesting chemical compounds (but, as expected, probably not blood).
If you look at the screen shot you see two different threads running in parallel, so I guess we are currently out of luck. It would be great if they could make them run in sequence indeed. Nothing breaks down hard disk speeds and responsiveness like two simultaneous copies. Of course, you might want to check which devices are being used by a copy. If you have a copy reading/writing /dev/hda and another copy reading/writing /dev/hdb, you might still want to run those in parallel. But even if they cannot do that, pretty please give me a choice to do those copies in sequence.
From the fortune cookies file (always handy to have the fortune db accessible):
Hanlon's Razor:
Never attribute to malice that which is adequately explained
by stupidity.
I got the title, here it is:
Helpful, isn't it?
Sleepyhead, we were talking WAN instead of LAN, not WLAN. Strangely enough, it might not matter. VPN is not just point to point, it can be used to create a secure virtual LAN as well. Actually, that's in the name, isn't it.