If you read the interview it sounds hopelessly optimistic and naive to imagine that you could implement a multimedia framework compatible with Silverlight as a free software alternative to Flash, that you could have a.NET and C# implementation compatible with Microsoft's, that you could write desktop applications in C#... until you remember that Miguel and his team have an awesome track record of doing all these things.
Actually, in a fresh install you get a popup that warns you when you try to send an unencrypted form.
Yeah, I remember that; everyone turns it off.
I'd like to see a warning only when you enter something that resembles a credit card number or bank account number (basically if the form contents contain more than six digits). Yes, I know phishing sites can work around this, but (provided Javascript permissions are set appropriately) only in bizarre ways that would alert the user something is wrong.
The Quake engine uses OpenGL (or its own software renderer, but I doubt anyone uses that anymore),
Isn't the point of Larabee to change that? With umpteen Pentium-compatible cores, each one beefed up with vector processing instructions, software rendering might become fashionable again.
The question you should ask is why is a website using a self-signed certificate presented to the user as *less safe* than one that is sending all information in the clear?
Point is, if the server indicates a particular Content-type (which could well be based on filename extension, or alternatively on a script header, phase of the moon, whatever) then the client must respect it and not override it with its own guesses.
You're slightly mistaken. The server thinks it's a GIF; the browser figures out that it's actually an applet and starts Java.
This is the same kind of idiotically 'helpful' behaviour that led to lots of exploits with WAV files about a decade ago. Oh I got a WAV file in an email, let's play it because it's just a sound file and that must be safe right? Then the media player saw that despite the filename it was actually an executable, and helpfully decided to run it instead...
Moral: use the Content-type from the server and do not do any kind of guessing based on filename, content or anything else. (With the possible exception of showing an unknown type as plain text.)
File extensions are fine for local filesystems, but they are not the way the web was designed to work. Web pages have a Content-type header.
Sadly, all Bruce P. seems to write these days is Slashdot-grade conspiracy theory.
IE is derived from Mosaic, the original Web browser, open source with a license similar to Apache's.
Er, no, Microsoft had to license it from Spyglass, Inc. Netscape chose to rewrite from scratch because the original Mosaic code was non-free.
Vista's customer-hostile emphasis on digital rights management, often handicapping its own features in suspicion that the user might have illicit content, caused its downfall. IT managers won't stand for that, and thus Microsoft has a lot of code to trash and rewrite before it can make an acceptable server platform.
[citation needed]... or is this just wishful thinking? Ever heard of Windows Server 2008?
Consistency is important but so is reproducibility. If you do an experiment, you need to have the same setup each time (and only vary the one thing you are interested in testing). But you also need to document what you did and why so that others can reproduce it. If you end up writing: we used this particular binary blob, we have no real idea what it does (apart from a few guesses we made by frobbing the CPUID) but it is published by a company called 'Futuremark' which sounds like a cool name, so I guess it's probably a reliable benchmark and stuff... this makes it hard for others to verify the method, which is just as important as reproducing the results.
And yeah, I know, this is just some gaming benchmark or whatever, not a cure for cancer. But if you're gonna do it, why not do it right?
Switching code based on CPUID is fine in a real application but it is a silly thing to do in a benchmark. OK, so to get the best performance you need to run slightly different code for different processors. Fine. But the benchmark can't possibly know ahead of time what the performance characteristics are of every x86 CPU made or to be made in the future, so what it needs to do is run *all of the different variants* and report which is the fastest for this particular CPU, and how long it took. Otherwise you will inevitably get situations like this one where it tries to second-guess what the fastest code is and gets it wrong. Since a benchmark is there to measure performance, it must not be written with any assumptions about performance!
(Even an application is often better off testing which code is fastest rather than doing CPUID. For example see the Linux software RAID code which runs different checksum implementations (or at least used to) at boot time and chooses whichever is fastest on this CPU.)
Why would you even consider running a benchmark program you don't have source code for and cannot compile yourself? (If you are worried about random compiler differences messing up the results, you can check an MD5 sum of the final binary against the published one, but it is important that you can reproduce the binary from source and you can read the source to find out what it does.)
If compilers like ICC cripple their code depending on CPUID, that will just lead all manufacturers to set CPUID to GenuineIntel, just as moronic websites (with help from Microsoft) ensured that all browsers call themselves 'Mozilla'.
Didn't you read the article? The kidney is there in memory of Uwe Thiem, a long-time contributor to KDE who passed away recently after a sudden kidney failure.
I don't get it. Why all the fuss about the desktop background? It is just a background after all, and hidden by any windows you have open.
From observing 'ordinary users' running Windows, they use the desktop background for starting programs which have a shortcut there - because the Start menu is so congested with crap, they don't even look at it and are often incapable of running anything not on the desktop. Because of this most Windows application installers have taken to adding a desktop shortcut as well as a Start menu item. Of course in the long term this 'icon inflation' will make the background itself unusable and we'll have to think of something else. I can't help feeling that just making a usable Start menu would be a better answer.
The second use of the desktop background is because files get saved there by default from your web browser. Again, this seems to be because unsophisticated users have no idea of directories and if it doesn't go on the background, they can't find it. But on Unix everyone has a home directory and I'd expect KDE (or GNOME) to provide easy access to that directory, even for people who aren't aware that any other location exists.
The kind of technically skilled people who used to run Enlightenment probably enjoy having semitransparent widgets flip into shape in 3d on the background, but I don't see what usability advantages that brings. Would it not be simpler to make the background be a background - just blank? There is no difficulty in putting one application window _underneath_ another, so you will see it when the top window is moved or minimized out of the way.
Maybe they do work out all that stuff when they are deciding whether to write a new program. But once you have spent the money and the program is written, all that matters is choosing the price that makes the most money. It doesn't matter if it cost you $1M or $100M to develop the program: your job is to get as much money as you can for it now. How could it be otherwise?
(OK, you might price the program a bit lower to get a toehold in a market with the expectation of selling more in future.)
But basically, you work out what people are willing to pay: for a given price P, the demand D(P) is the number of copies sold. Mostly, D(P) falls as P increases, since people are less likely to buy something for $500 than the same product for $100. In software some people wrongly believe that a high price means high quality, so D might not be sloping downward all the way. Anyway, then you choose the value of P to maximize P * D(P), assuming that marginal costs of sending a CD to someone (or letting them download the program from your site) are pretty close to zero.
and the price reflects the marginal production costs per unit.
So it's free of charge then?
The marginal production costs per unit are almost zero. All they have to do is burn a CD and post it to you. Usually you don't even get a paper manual. There are some costs for support, I admit.
Production costs for software are almost entirely fixed costs, and once paid you can ship one copy or a million copies of the program for pretty much the same cost.
Everyone knows that the moon landings were faked. Clearly this announcement of so-called 'aliens' is just another part of the same government consipiracy. Come _on_ people!
It is completely crazy to pop up all kinds of scary warnings for an encrypted connection that doesn't have a trusted certificate, while at the same time producing no warning at all for a connection that's totally unencrypted! That is the issue here.
Paint.NET? That's on Windows, isn't it? It's a shame that people have to reboot into Windows to run a useful free application like that.
If only there were, I don't know... some kind of compatibility environment so you could run apps like Paint.NET on Linux using only free software? In an ideal world there might even be a port of Paint.NET done by some hard-working free software developer.
The GPL doesn't force you to share your work at all. You can keep it to yourself. It does mean that if you distribute it, those who get it have the right to distribute it further. That is the same as things would be if copyright law didn't exist. The one major difference is that the GPL requires source code be distributed alongside binaries: a copyright-free world would not have that rule.
Why do we still have 'mirrors' in this day and age? It made some sense when Internet users had beards and understood the network topology and how many hops they were away from Kevin Bacon's VAX. In those days you also needed to specify email addresses giving the routing explicitly, as foovax!kremvax!buffy!user. But that long since got replaced with giving everyone a fixed address and letting the computer figure out the boring details of how to route it.
Why not just use Bittorrent to distribute package updates and forget about the whole mirror network? People who spend time setting up and maintaining mirrors can enjoy more free time to hack on more interesting things.
Slashdot Mirror
To quote my favourite font name: \!Andale Mono!
Yeah, I remember that; everyone turns it off.
I'd like to see a warning only when you enter something that resembles a credit card number or bank account number (basically if the form contents contain more than six digits). Yes, I know phishing sites can work around this, but (provided Javascript permissions are set appropriately) only in bizarre ways that would alert the user something is wrong.
Isn't the point of Larabee to change that? With umpteen Pentium-compatible cores, each one beefed up with vector processing instructions, software rendering might become fashionable again.
The question you should ask is why is a website using a self-signed certificate presented to the user as *less safe* than one that is sending all information in the clear?
Read the article: after adding the parity information, the data is interleaved every 512 bytes to withstand losing whole sectors.
Point is, if the server indicates a particular Content-type (which could well be based on filename extension, or alternatively on a script header, phase of the moon, whatever) then the client must respect it and not override it with its own guesses.
Where do you get this from? What legal opinion do you have that says that reprogramming your own phone is illegal?
'Their' product? No, the phone belongs to the person who paid several hundred dollars for it. That's this concept we have called a 'sale'.
This is the same kind of idiotically 'helpful' behaviour that led to lots of exploits with WAV files about a decade ago. Oh I got a WAV file in an email, let's play it because it's just a sound file and that must be safe right? Then the media player saw that despite the filename it was actually an executable, and helpfully decided to run it instead...
Moral: use the Content-type from the server and do not do any kind of guessing based on filename, content or anything else. (With the possible exception of showing an unknown type as plain text.)
File extensions are fine for local filesystems, but they are not the way the web was designed to work. Web pages have a Content-type header.
Sadly, all Bruce P. seems to write these days is Slashdot-grade conspiracy theory.
Er, no, Microsoft had to license it from Spyglass, Inc. Netscape chose to rewrite from scratch because the original Mosaic code was non-free.
[citation needed]... or is this just wishful thinking? Ever heard of Windows Server 2008?
Consistency is important but so is reproducibility. If you do an experiment, you need to have the same setup each time (and only vary the one thing you are interested in testing). But you also need to document what you did and why so that others can reproduce it. If you end up writing: we used this particular binary blob, we have no real idea what it does (apart from a few guesses we made by frobbing the CPUID) but it is published by a company called 'Futuremark' which sounds like a cool name, so I guess it's probably a reliable benchmark and stuff... this makes it hard for others to verify the method, which is just as important as reproducing the results.
And yeah, I know, this is just some gaming benchmark or whatever, not a cure for cancer. But if you're gonna do it, why not do it right?
Would it not work just as well to use your home directory as a temporary store, and have a single button click to show your home directory?
Yes, this is the second use I mentioned; surely the problem could be addressed just as well by making the start menu simple and uncrufted.
Switching code based on CPUID is fine in a real application but it is a silly thing to do in a benchmark. OK, so to get the best performance you need to run slightly different code for different processors. Fine. But the benchmark can't possibly know ahead of time what the performance characteristics are of every x86 CPU made or to be made in the future, so what it needs to do is run *all of the different variants* and report which is the fastest for this particular CPU, and how long it took. Otherwise you will inevitably get situations like this one where it tries to second-guess what the fastest code is and gets it wrong. Since a benchmark is there to measure performance, it must not be written with any assumptions about performance!
(Even an application is often better off testing which code is fastest rather than doing CPUID. For example see the Linux software RAID code which runs different checksum implementations (or at least used to) at boot time and chooses whichever is fastest on this CPU.)
Why would you even consider running a benchmark program you don't have source code for and cannot compile yourself? (If you are worried about random compiler differences messing up the results, you can check an MD5 sum of the final binary against the published one, but it is important that you can reproduce the binary from source and you can read the source to find out what it does.)
If compilers like ICC cripple their code depending on CPUID, that will just lead all manufacturers to set CPUID to GenuineIntel, just as moronic websites (with help from Microsoft) ensured that all browsers call themselves 'Mozilla'.
Didn't you read the article? The kidney is there in memory of Uwe Thiem, a long-time contributor to KDE who passed away recently after a sudden kidney failure.
I don't get it. Why all the fuss about the desktop background? It is just a background after all, and hidden by any windows you have open.
From observing 'ordinary users' running Windows, they use the desktop background for starting programs which have a shortcut there - because the Start menu is so congested with crap, they don't even look at it and are often incapable of running anything not on the desktop. Because of this most Windows application installers have taken to adding a desktop shortcut as well as a Start menu item. Of course in the long term this 'icon inflation' will make the background itself unusable and we'll have to think of something else. I can't help feeling that just making a usable Start menu would be a better answer.
The second use of the desktop background is because files get saved there by default from your web browser. Again, this seems to be because unsophisticated users have no idea of directories and if it doesn't go on the background, they can't find it. But on Unix everyone has a home directory and I'd expect KDE (or GNOME) to provide easy access to that directory, even for people who aren't aware that any other location exists.
The kind of technically skilled people who used to run Enlightenment probably enjoy having semitransparent widgets flip into shape in 3d on the background, but I don't see what usability advantages that brings. Would it not be simpler to make the background be a background - just blank? There is no difficulty in putting one application window _underneath_ another, so you will see it when the top window is moved or minimized out of the way.
Maybe they do work out all that stuff when they are deciding whether to write a new program. But once you have spent the money and the program is written, all that matters is choosing the price that makes the most money. It doesn't matter if it cost you $1M or $100M to develop the program: your job is to get as much money as you can for it now. How could it be otherwise?
(OK, you might price the program a bit lower to get a toehold in a market with the expectation of selling more in future.)
But basically, you work out what people are willing to pay: for a given price P, the demand D(P) is the number of copies sold. Mostly, D(P) falls as P increases, since people are less likely to buy something for $500 than the same product for $100. In software some people wrongly believe that a high price means high quality, so D might not be sloping downward all the way. Anyway, then you choose the value of P to maximize P * D(P), assuming that marginal costs of sending a CD to someone (or letting them download the program from your site) are pretty close to zero.
So it's free of charge then?
The marginal production costs per unit are almost zero. All they have to do is burn a CD and post it to you. Usually you don't even get a paper manual. There are some costs for support, I admit.
Production costs for software are almost entirely fixed costs, and once paid you can ship one copy or a million copies of the program for pretty much the same cost.
Everyone knows that the moon landings were faked. Clearly this announcement of so-called 'aliens' is just another part of the same government consipiracy. Come _on_ people!
Agreed. It shouldn't look any more 'reassuring' than a normal unencrypted connection. But it's crazy to mark it as more dangerous.
It is completely crazy to pop up all kinds of scary warnings for an encrypted connection that doesn't have a trusted certificate, while at the same time producing no warning at all for a connection that's totally unencrypted! That is the issue here.
Paint.NET? That's on Windows, isn't it? It's a shame that people have to reboot into Windows to run a useful free application like that.
If only there were, I don't know... some kind of compatibility environment so you could run apps like Paint.NET on Linux using only free software? In an ideal world there might even be a port of Paint.NET done by some hard-working free software developer.
The GPL doesn't force you to share your work at all. You can keep it to yourself. It does mean that if you distribute it, those who get it have the right to distribute it further. That is the same as things would be if copyright law didn't exist. The one major difference is that the GPL requires source code be distributed alongside binaries: a copyright-free world would not have that rule.
Silly, didn't you know that MakeMaker is DOOMED?
Why do we still have 'mirrors' in this day and age? It made some sense when Internet users had beards and understood the network topology and how many hops they were away from Kevin Bacon's VAX. In those days you also needed to specify email addresses giving the routing explicitly, as foovax!kremvax!buffy!user. But that long since got replaced with giving everyone a fixed address and letting the computer figure out the boring details of how to route it.
Why not just use Bittorrent to distribute package updates and forget about the whole mirror network? People who spend time setting up and maintaining mirrors can enjoy more free time to hack on more interesting things.