If you can't even begin to vet the source there is no accounting for the bugs and potential back doors. We don't know what the adversaries intentions might be. While it could be profit driven as is the case with most malware it could also be espionage, spying on dissidents, or something else. Of which there maybe no acting on the bug/backdoor. That would make it significantly more difficult to detect in a closed source application.
After that, the next step would be to get someone to provably audit that open source code. We have seen that open source is no guarantee that the eyeballs are actually there. Even some malicious party could distribute something heinous and just get away with it by saying "relax, it's open source".
It's just damn hard to do, as software is so complex these days. Even if you come up with a better idea of how things should be done, you need a good sized team to make it happen. That's why Red Hat takes the pie: it has the resources (armies of paid developers) to make this stuff happen. Well, there's always Kickstarter...;)
Greg Kroah-Hartman is quite relaxed guy, I like him. He's a proof that you can be an elite Linux developer and still be cool. BTW he did an Ask Me Anything session recently.
The author never claims that those don't exist. He says that using proprietary document formats might not make any more sense than using such coffee maker.
A bad deficiency we still are having is that the mail traffic between mail exchange servers is typically unencrypted. Makes e-mail quite easy to spy by the gubbermentz.
Some people fear math too much. Yes, it requires work to master, but it can be done by anyone.
Fully agree.
As far as things like programming are concerned learning underlying theory(in this case mathematics) will save you time, because without it you'll end up reinventing things that were already discovered in middle ages.
Such as? De Morgan's Theorem is the only semi-mathematical concept that I remembering getting benefit from in general programming.
Actually it was probably said just to curb some of the jealousy of some angry people thinking that Bennett has a special position to write these longer opinion pieces. Anyone can submit them.
As the submitted of this article, I recommend you to continue chucking in the articles that you find interesting. You seem to have 17 articles submitted, of which 6 have been published, which is pretty good actually. There's many factors that decide whether your submission gets the front page or not, so there's no need to be too personal about it. I personally keep submitting articles just for fun. Most of my topics come from Twitter these days, it's a good source to pick up interesting stuff.
In practice you needed a Pentium 2 with 64MB RAM to run Windows 98 properly.
This made me think how in the old days, software minimum requirements often described the bare minimum hardware with what the software kinda-sorta could start.:) These days we don't see that as much, but defining requirements is still tricky: for example you can't really meaningfully slap there "2GHz CPU or faster", because the work done per clock cycle has improved tremendously. Describing the GPU requirement is problematic too, because if you say "GTX460 or faster", some people can have hard time weighing how fast a GTX460 exactly was, and what was the performance of various chips that came after that.
Actually more is coming. The Continuum feature will show all apps in windows, but can automatically switch to full screen tablet mode, for example when you undock your device. It kind of tries to give the best of both worlds. AFAIK this feature did not make it yet to the W10TP though.
I guess a good compositor should let a full screen app pass through without getting in the way. If every frame drawn by the game goes also through the desktop compositing pipeline, yeah, it's going to have a toll.
I still suspect that the disk I/O speed would be much a larger bottleneck than the speed of the FS accessing that disk.
If you can't even begin to vet the source there is no accounting for the bugs and potential back doors. We don't know what the adversaries intentions might be. While it could be profit driven as is the case with most malware it could also be espionage, spying on dissidents, or something else. Of which there maybe no acting on the bug/backdoor. That would make it significantly more difficult to detect in a closed source application.
After that, the next step would be to get someone to provably audit that open source code. We have seen that open source is no guarantee that the eyeballs are actually there. Even some malicious party could distribute something heinous and just get away with it by saying "relax, it's open source".
Probably because HFS is an abomination.
The file system cannot make a noticeable performance difference on a basic desktop system. It uses so little resources.
Wohow, you're right! I'll tune my values immediately.
Do you even kernel mailing list?
Linus sends like 1000 emails a month.
Err...you might want to tune that value a bit. The whole LKML gets roughly 1000 messages a month.
It's just damn hard to do, as software is so complex these days. Even if you come up with a better idea of how things should be done, you need a good sized team to make it happen. That's why Red Hat takes the pie: it has the resources (armies of paid developers) to make this stuff happen. Well, there's always Kickstarter... ;)
:D
Greg Kroah-Hartman is quite relaxed guy, I like him. He's a proof that you can be an elite Linux developer and still be cool. BTW he did an Ask Me Anything session recently.
The author never claims that those don't exist. He says that using proprietary document formats might not make any more sense than using such coffee maker.
The Atari 2600 example shows that DRM can be beneficial to the customer too, as the company can better protect the high quality of its ecosystem.
A bad deficiency we still are having is that the mail traffic between mail exchange servers is typically unencrypted. Makes e-mail quite easy to spy by the gubbermentz.
It claimed to be "Öppen Gäst" (open for guests) in the SSID name.
Some people fear math too much. Yes, it requires work to master, but it can be done by anyone.
Fully agree.
As far as things like programming are concerned learning underlying theory(in this case mathematics) will save you time, because without it you'll end up reinventing things that were already discovered in middle ages.
Such as? De Morgan's Theorem is the only semi-mathematical concept that I remembering getting benefit from in general programming.
Actually it was probably said just to curb some of the jealousy of some angry people thinking that Bennett has a special position to write these longer opinion pieces. Anyone can submit them.
As the submitted of this article, I recommend you to continue chucking in the articles that you find interesting. You seem to have 17 articles submitted, of which 6 have been published, which is pretty good actually. There's many factors that decide whether your submission gets the front page or not, so there's no need to be too personal about it. I personally keep submitting articles just for fun. Most of my topics come from Twitter these days, it's a good source to pick up interesting stuff.
What about SimCity 4?
A new version of DirectX would certainly have been classified as a platform update, so you are correct.
Yep, Windows 2000 was fantastic. Windows XP was actually just Windows 2000 with extra garbage added: instability, bloat, malware.
In practice you needed a Pentium 2 with 64MB RAM to run Windows 98 properly.
This made me think how in the old days, software minimum requirements often described the bare minimum hardware with what the software kinda-sorta could start. :) These days we don't see that as much, but defining requirements is still tricky: for example you can't really meaningfully slap there "2GHz CPU or faster", because the work done per clock cycle has improved tremendously. Describing the GPU requirement is problematic too, because if you say "GTX460 or faster", some people can have hard time weighing how fast a GTX460 exactly was, and what was the performance of various chips that came after that.
Actually more is coming. The Continuum feature will show all apps in windows, but can automatically switch to full screen tablet mode, for example when you undock your device. It kind of tries to give the best of both worlds. AFAIK this feature did not make it yet to the W10TP though.
I guess a good compositor should let a full screen app pass through without getting in the way. If every frame drawn by the game goes also through the desktop compositing pipeline, yeah, it's going to have a toll.
Hmm...
The UTF-32 form of a character is a direct representation of its codepoint.
Probably just the day coming to end and it becoming dark.
UTF-32 should make memory allocation more predictable as every character is guaranteed to be 32 bits.