It's certainly the most likely browser to be tested against; I agree with that. On the mobile market it might still play second fiddle to Opera, but for desktops Firefox is a likely baseline.
Future web-apps will be written for a mishmash of platforms; (X)HTML+JS, XUL+JS, pure JS (see Objective-J), Prism, AIR, Silverlight, Flash... All of these will be more of less compatible with whatever you're running. Prism and XUL-based apps will never run on anything but Firefox. Silverlight apps will always run best under Windows. Flash will always lag behind on Linux. And so on.
There's too many big players who want their slice of the pie. I'd say that own-runtime technologies like AIR, Silverlight and Flash will fight for the crown with (X)HTML + Javascript. And of course every browser will have its own rendering standards which will mean that we still have to be careful about which parts of CSS we use.
While Tamarin is the ActionScript interpreter for Flash, it also will be the JavaScript interpreter for a future version for Firefox. So we can also expect C in JavaScript.
That's because you're not 2.0 enough. Everything gets better when run inside a browser. If you have Firefox, you can type chrome://browser/content/browser.xul into the URL^H^H^HAwesome bar. That will run Firefox inside Firefox, exponentially increasing the awesomeness. I tried it and I came when I reached the third layer. That's how awesome running stuff inside a browser is.
I really hope Firefox 4 will have full DirectX 12 support so I can play all my games inside it as well. Also, the next CPU enhancement (like SSE or 3DNow!) will most probably be something that specifically speeds up browsers. Because running stuff in browsers is that awesome and everyone will do it in the furure, even BIOS vendors.
Unfortunately bstring doesn't do Unicode. While using strings in C/C++ is mostly annoying, using non-English strings in C/C++ is mostly like pulling your own teeth. Latin-1 can only take you so far (and doesn't always work).
Actually, why do we bother with operating systems? We just build a SOC that runs Gecko and have the OS be an instance of Firefox, loaded from an EEPROM. Since everything is better when executed on a server we can save money by cutting down on the client hardware - after all, a web browser only needs that much power; a 1 GHz singlecore with 512 MiB of RAM should suffice. The telcos just need to stop skimping on bandwidth and give everyone 10 Gbps down/100 Mbps up and we can even play video games.
You want a mobile graphics chip that has a TDP of 300 W? Good luck trying to cool that beast.
Either Larrabee is about five times more powerful than anything AMD and Nvidia have to offer (unlikely) or Intel needs to somehow reduce the TDP of that thing by a factor of five to ten. Otherwise I don't really see the appeal.
Heck, I can build a gaming system with decent performance that sucks 300 Watts in total. Apparently Intel longs back for the days of the Pentium 4, when they were a manufacturer of space heaters.
Actually, agnosticism comes in two flavors, as well: Weak agnosticism says that we merely don't have any compelling arguments and strong agnosticism says that there can be no compelling arguments. Weak agnosticism and weak atheism are very similar and there's only a very slight difference between a sceptic weak agnostic and a weak atheist.
It's not as easy as saying "all X believe/don't believe". Some atheists believe, some don't. Same goes for agnostics.
And then you have ignosticism, an agnosticism derivate which postulates that the question whether God exists doesn't matter at all. Ignosticism has an opinion ("it doesn't matter"), but it doesn't neccessarily have a belief - uness you count "it doesn't matter" as a belief, in which case it does.
Actually, it does in all cases.
In the end it all comes down to whether someone makes an assumption without having hard data to back it up. Everyone falls under this - even weak agnostics can't claim without assumption that there is no proof unless they have evaluated every proof so far and ignostics can't claim without assumption that it doesn't matter because it could matter.
Everyone believes. As soon as you make an assumption, that's a belief. Since we don't have definitive, widely accepted proof in this area so far, everyone operate on nothing but belief, even the guys who say they don't care about the topic. Arguing over whether one believes or not is pointless and so is being offended by the notion.
I have an unusual name and thus most of my results turned out to be relevant - embarassingly, my top results are bug reports to several high-profile projects (like the Linux kernel) that I ultimately retracted when I found out I had a sneaky hardware fault.
Confusingly, an amateur soccer player goes by the same name. I hate soccer. It would be weird if I showed up to an interview and the interviewer started asking me about recent games of some obscure club my namesake has played for...
However, my direct Google results do mainly paint me as an Open Source enthusiast. I can live with that.
That way you make one mistake and you end up contacting an existing but unrelated server. With an SHA-1 hash any mistaken number is virtually guaranteed to send you to a non-existing host. That way my scheme has error detection built in.
Is there a particular advantage in using acronyms instead of subdomains? I'd assume that dns1.solaris.production.foo would be a more memorable way to describe the first solaris-based DNS server in production than SPDNS00. Is there something besides "that happens to be what we use" or "we started using that naming scheme when the computers couldn't yet handle subdomains"?
How about using an SHA-1 hash of an incrementing counter? The first box is 356a192b7913b04c54574d18c28d46e6395428ab.company.internal, the second one is da4b9237bacccdf19c0760cab7aec4a8359010b0.company.internal etc. The mapping between counter values and machines is stored in an Excel spreadsheet, printed out and stored in the server room.
That way you get a unique naming scheme that's both logical, understandable (you can convert the host name into its counter value through a simple rainbow table) and reasonably safe from hash collisions.
...and also the first post I see where there is more than one name for each machine. It does make sense, though - so much sense that I wonder why most people don't seem to do it. It does allow for transparent moving of duties between machines, which sounds very useful to me. But then again IANAsysadmin.
Note: Any sysadmins who do have sound reasons for using exactly one name per machine are encouraged to share them. I'm not an admin, but I'm also not stupid enough to pass up information.
I don't want to start a holy war here, but what is the deal with KDE 4 preview performance? I've been sitting here at my freelance gig in front of an Opteron box running KDE 4.1 for about 20 minutes now while it attempts to create a thumbnail for a 17 Meg file in one folder on the hard drive. 20 minutes. At home, on my Pentium Pro 200 running KDE 3.5, which by all standards should be a lot slower than this box, the same operation would take about 2 minutes. If that.
In addition, during this file transfer, Konqueror will not work. And everything else has ground to a halt. Even KWrite is straining to keep up as I type this.
I won't bore you with the laundry list of other problems that I've encountered while working on KDE 4, but suffice it to say there have been many, not the least of which is I've never seen a KDE 4 box that has run faster than its KDE 3.5 counterpart, despite the KDE 4's faster library architecture. My 486/66 with 8 megs of RAM runs faster than this 3000 MHz machine at times. From a productivity standpoint, I don't get how people can claim that KDE 4 is a superior desktop environment.
KDE 4 addicts, flame me if you'd like, but I'd rather hear some intelligent reasons why anyone would choose to use KDE 4 over other faster, cheaper, more stable desktop environments.
Adding an option to do something that the program should be able to figure out is a bad idea. So in that regard we should be trying to minimize option dialog clutter by making programs smarter.
Just don't overdo it. Few things are as aggravating as "smart" software that isn't - and software that autoconfigures itself for the most common use case when I want to use it differently falls under that category.
In general, I like the idea of smart programs only when you can disable the "smart" logic. Otherwise you risk ending up with software like Word/OOo Writer that can't be used without sitting down and learning all about its "smart" features beforehand or risk the program "helpfully" destroying the formatting or even contents of your text.
Apparently we don't get as many fraudsters over here. Also, wire transfers are usually free (unless the seller sits in another country) and fast. I trust them more then PayPal and they still are the default payment method on German eBay. Virtually nobody takes CCs (probably because CCs aren't that popular in Germany) and PayPal is nowhere near universal.
If I get scammed I can get the scammer's eBay account blocked and I can go to the police and have them look into things. Granted, that doesn't guarantee that I get my money back but so far I didn't have any problems.
Now if PayPal becomes a real bank and stops their random account suspensions and atrociuos-even-for-German-standards customer support I might start actually trusting them.
Where's the problem with accepting wire transfers? In Germany payment per wire transfer is always understood as an advance payment - the seller will not send the article until he has the full payment on his account statement (usually one to three business days after the buyer has issued the transfer). The risk lies entirely with the buyer.
Are American wire transfers unsafe or extremely slow or something?
Slashdot Mirror
It's certainly the most likely browser to be tested against; I agree with that. On the mobile market it might still play second fiddle to Opera, but for desktops Firefox is a likely baseline.
Future web-apps will be written for a mishmash of platforms; (X)HTML+JS, XUL+JS, pure JS (see Objective-J), Prism, AIR, Silverlight, Flash... All of these will be more of less compatible with whatever you're running. Prism and XUL-based apps will never run on anything but Firefox. Silverlight apps will always run best under Windows. Flash will always lag behind on Linux. And so on.
There's too many big players who want their slice of the pie. I'd say that own-runtime technologies like AIR, Silverlight and Flash will fight for the crown with (X)HTML + Javascript. And of course every browser will have its own rendering standards which will mean that we still have to be careful about which parts of CSS we use.
While Tamarin is the ActionScript interpreter for Flash, it also will be the JavaScript interpreter for a future version for Firefox. So we can also expect C in JavaScript.
Ah, you're talking about obfuscation! Okay, Perl does win that one, hands-down.
That's because you're not 2.0 enough. Everything gets better when run inside a browser. If you have Firefox, you can type chrome://browser/content/browser.xul into the URL^H^H^HAwesome bar. That will run Firefox inside Firefox, exponentially increasing the awesomeness. I tried it and I came when I reached the third layer. That's how awesome running stuff inside a browser is.
I really hope Firefox 4 will have full DirectX 12 support so I can play all my games inside it as well. Also, the next CPU enhancement (like SSE or 3DNow!) will most probably be something that specifically speeds up browsers. Because running stuff in browsers is that awesome and everyone will do it in the furure, even BIOS vendors.
Unfortunately bstring doesn't do Unicode. While using strings in C/C++ is mostly annoying, using non-English strings in C/C++ is mostly like pulling your own teeth. Latin-1 can only take you so far (and doesn't always work).
Actually, why do we bother with operating systems? We just build a SOC that runs Gecko and have the OS be an instance of Firefox, loaded from an EEPROM. Since everything is better when executed on a server we can save money by cutting down on the client hardware - after all, a web browser only needs that much power; a 1 GHz singlecore with 512 MiB of RAM should suffice. The telcos just need to stop skimping on bandwidth and give everyone 10 Gbps down/100 Mbps up and we can even play video games.
They will also incorporate DWIM, use Google to perfectly translate themselves into any language the user wishes and debug themselves.
You want a mobile graphics chip that has a TDP of 300 W? Good luck trying to cool that beast.
Either Larrabee is about five times more powerful than anything AMD and Nvidia have to offer (unlikely) or Intel needs to somehow reduce the TDP of that thing by a factor of five to ten. Otherwise I don't really see the appeal.
Heck, I can build a gaming system with decent performance that sucks 300 Watts in total. Apparently Intel longs back for the days of the Pentium 4, when they were a manufacturer of space heaters.
Look at my middle initial. Not many people have a middle name beginning with "_".
Wow. That actually sounds pretty awesome. I didn't know you could mount that kind of defense against something that happens in your brain.
Actually, agnosticism comes in two flavors, as well: Weak agnosticism says that we merely don't have any compelling arguments and strong agnosticism says that there can be no compelling arguments. Weak agnosticism and weak atheism are very similar and there's only a very slight difference between a sceptic weak agnostic and a weak atheist.
It's not as easy as saying "all X believe/don't believe". Some atheists believe, some don't. Same goes for agnostics.
And then you have ignosticism, an agnosticism derivate which postulates that the question whether God exists doesn't matter at all. Ignosticism has an opinion ("it doesn't matter"), but it doesn't neccessarily have a belief - uness you count "it doesn't matter" as a belief, in which case it does.
Actually, it does in all cases.
In the end it all comes down to whether someone makes an assumption without having hard data to back it up. Everyone falls under this - even weak agnostics can't claim without assumption that there is no proof unless they have evaluated every proof so far and ignostics can't claim without assumption that it doesn't matter because it could matter.
Everyone believes. As soon as you make an assumption, that's a belief. Since we don't have definitive, widely accepted proof in this area so far, everyone operate on nothing but belief, even the guys who say they don't care about the topic. Arguing over whether one believes or not is pointless and so is being offended by the notion.
I have an unusual name and thus most of my results turned out to be relevant - embarassingly, my top results are bug reports to several high-profile projects (like the Linux kernel) that I ultimately retracted when I found out I had a sneaky hardware fault.
Confusingly, an amateur soccer player goes by the same name. I hate soccer. It would be weird if I showed up to an interview and the interviewer started asking me about recent games of some obscure club my namesake has played for...
However, my direct Google results do mainly paint me as an Open Source enthusiast. I can live with that.
That way you make one mistake and you end up contacting an existing but unrelated server. With an SHA-1 hash any mistaken number is virtually guaranteed to send you to a non-existing host. That way my scheme has error detection built in.
Is there a particular advantage in using acronyms instead of subdomains? I'd assume that dns1.solaris.production.foo would be a more memorable way to describe the first solaris-based DNS server in production than SPDNS00. Is there something besides "that happens to be what we use" or "we started using that naming scheme when the computers couldn't yet handle subdomains"?
How about using an SHA-1 hash of an incrementing counter? The first box is 356a192b7913b04c54574d18c28d46e6395428ab.company.internal, the second one is da4b9237bacccdf19c0760cab7aec4a8359010b0.company.internal etc. The mapping between counter values and machines is stored in an Excel spreadsheet, printed out and stored in the server room.
That way you get a unique naming scheme that's both logical, understandable (you can convert the host name into its counter value through a simple rainbow table) and reasonably safe from hash collisions.
But its locale is set to de-DE! A system that speaks German can't be a bad system.
...and also the first post I see where there is more than one name for each machine. It does make sense, though - so much sense that I wonder why most people don't seem to do it. It does allow for transparent moving of duties between machines, which sounds very useful to me. But then again IANAsysadmin.
Note: Any sysadmins who do have sound reasons for using exactly one name per machine are encouraged to share them. I'm not an admin, but I'm also not stupid enough to pass up information.
I don't want to start a holy war here, but what is the deal with KDE 4 preview performance? I've been sitting here at my freelance gig in front of an Opteron box running KDE 4.1 for about 20 minutes now while it attempts to create a thumbnail for a 17 Meg file in one folder on the hard drive. 20 minutes. At home, on my Pentium Pro 200 running KDE 3.5, which by all standards should be a lot slower than this box, the same operation would take about 2 minutes. If that.
In addition, during this file transfer, Konqueror will not work. And everything else has ground to a halt. Even KWrite is straining to keep up as I type this.
I won't bore you with the laundry list of other problems that I've encountered while working on KDE 4, but suffice it to say there have been many, not the least of which is I've never seen a KDE 4 box that has run faster than its KDE 3.5 counterpart, despite the KDE 4's faster library architecture. My 486/66 with 8 megs of RAM runs faster than this 3000 MHz machine at times. From a productivity standpoint, I don't get how people can claim that KDE 4 is a superior desktop environment.
KDE 4 addicts, flame me if you'd like, but I'd rather hear some intelligent reasons why anyone would choose to use KDE 4 over other faster, cheaper, more stable desktop environments.
Just don't overdo it. Few things are as aggravating as "smart" software that isn't - and software that autoconfigures itself for the most common use case when I want to use it differently falls under that category.
In general, I like the idea of smart programs only when you can disable the "smart" logic. Otherwise you risk ending up with software like Word/OOo Writer that can't be used without sitting down and learning all about its "smart" features beforehand or risk the program "helpfully" destroying the formatting or even contents of your text.
No, the word "beta" means "production-grade software in some way related to the internet".
What did he leave and why is it right now? What if what he left was actually wrong, like Roland Piquepaille porn?
Apparently we don't get as many fraudsters over here. Also, wire transfers are usually free (unless the seller sits in another country) and fast. I trust them more then PayPal and they still are the default payment method on German eBay. Virtually nobody takes CCs (probably because CCs aren't that popular in Germany) and PayPal is nowhere near universal.
If I get scammed I can get the scammer's eBay account blocked and I can go to the police and have them look into things. Granted, that doesn't guarantee that I get my money back but so far I didn't have any problems.
Now if PayPal becomes a real bank and stops their random account suspensions and atrociuos-even-for-German-standards customer support I might start actually trusting them.
Where's the problem with accepting wire transfers? In Germany payment per wire transfer is always understood as an advance payment - the seller will not send the article until he has the full payment on his account statement (usually one to three business days after the buyer has issued the transfer). The risk lies entirely with the buyer.
Are American wire transfers unsafe or extremely slow or something?