They mentioned that it can send telemetry as well. Perhaps, if it's bouyant enough, it's simpler to recover from the ocean..and perhaps the water damages it less than hitting the windshield of your car on the freeway.
The irony in this is that in my opinion, the install is the *least* important part of the distribution. What I'm concerned about is post-install usability.
During the install, Joe Blow is probably going to have the Linux geek that convinced him to try it at the controls anyways...if not LUGs are readily availiable for installs...and failing that, anyone who can follow directions can do it basically from scratch. I had a friend install Gentoo, and while he had a few questions, he got it working with minimal help (I mean one or two one line questions, not a hand holding) and no in-person contact.
Back on topic. There's usually goign to be someone more knowledgable around during the installation, so I don't consider it a factor when choosing a distro for myself..or for others.
What *is* exremely important is package managment. Anything that uses RPM is out automatically. Trying to teach someone (even someone of reasonable intelligence with a fair bit of Windows experience) to resolve RPM dependencies..then watching their face turn white as they realize that they have to download 30 other RPMs to install one program should tell you right away it's a bad choice for a newbie. Why RedHat hasn't adopted apt (or some new RPM version) is beyond me. Perhaps RedCarpet will fit the bill.
Again, I'm ranting. I can set up, and configure a friend's system with Debian in not-too-long...probably not much longer than it would take him to install RedHat. But for my time, he gets a usable system in the future. I'll install a bunch of common applications, KDE, and give him some desktop shortcuts...and unless he's a gamer he'll figure it out remarkably quickly. A quick guide on how to open a console and install an application with apt is all that's really needed.
RedHat et. al. really, really need to fix the package nightmare that is RPM before I'll seriously consider using them again. That might be soon, however, seeing how outside developers are now providing suggestions and input into redhat's processes.
In Windows, after approximately a month, the start menu stops auto-sorting new items in alpha order. Then I have to do sort by name whenever I install a new application so I can find it on the menu. This has gone on from Windows 95 to 2000, and I believe XP as well. It starts out fine, and just goes awry for no apparent reason, and it's a real pain in the ass usability wise.
This is bloody slow and annoying. Double clicking takes a marginal extra amount of time, but I do most of my file operations (in gui) by selecting a file and using keyboard shortcuts with my other hand. This is faster than drag/drop when I have multiple large folders open in seperate windows. Waiting 2 seconds for the file to be selected would be hell.
I just feel extremely uncomfortable working in file managers when this is the case. Even dragging and dropping i sometimes end up registering a click and opening apps...just plain clumsy.
a) you probably could've telnetted/sshd into the box, or used a serial terminal and killed the X server, I highly doubt you managed to crash the actual OS.
b) that was a configuration problem, obviously they know exactly what the hardware is, what it can do etc. and obviously again, the pilots aren't going to have access to the configuration files (which are probably in a ROM anyways).
c) I can almost guarantee they wouldn't be using X
The only time I've ever seen a kernel panic was because I filled a hard drive and the kernel couldn't write to it...oh and of course if / isn't mountable at boot time. Other than that, every crash I've seen (with stable kernels) has been a software issue, not an OS issue.
The myth about Windows crashing twice a day is definitely a myth, or perhaps you could call it an exaggeration. I don't crash Windows enough for it to really piss me off, but it does happen from time to time.
That's not been my experience. In college I'd regularly use my discman to listen to tunes during class. With top-line non-rechargable batteries (Duracell Ultra et al.), it would last me from 3-4 weeks (non continuous use obviously). If I ever picked up a set at the student union (they didn't sell name brand batteries) because I didn't have spares with me, they would last much less time, generally a week or two. Perhaps it has to do with the specific batteries and their construction with relation to my cd player, but I suspect, as with most anything else, that you get what you pay for.
Please mod the parent up, buying the cheapest batteries you can find is definitely bad advice. There is even a large quality margin in standard non-rechargable Alkaline batteries, not to mention the more complex rechargables. I always buy Duracell or Energizer alkalines, since they generally last 2-3 times longer than the el-cheapo brand.
As for rechargables, the Alkaline rechargables (Energizer has some, as does a company called PureEnergy) are pretty good for the cost. The chargers are inexpensive, and the batteries don't cost much more than standard alkalines...the downside is they only last for 20-50 charges.
You can run X on a 486 with decent performance, if you don't &*(% the thing up and saddle it with a bunch of useless crap.
The fact remains that Windows is still faster when saddled with all that useless crap, and at least as fast without it. Something needs to change (and it goes deeper than "don't run it, it's slow").
Fonts look far better in pretty much every application on my Gentoo and FreeBSD systems (both were fully updated within the past month).
I'd blame either RedHat, or the slow release-test-upgrade cycle of such a commercial distribution. Font rendering in *ix has definitely matured in the past 6 months or so, get a recent distro and give your eyes a well deserved break;)
Re:Before all the flamers get in.
on
Qt On DirectFB
·
· Score: 1
3D is a whole 'nother ballgame. OpenGL is implemented in X (whereas GTK or QT are not). Just as it would be on a Windows box, each polygon is mapped in the scene, textured, and moved around etc. just as it would be under any other OpenGL application.
The discrepancy is that X primitive libraries (QT, GTK etc.) have to draw each individual X primitive (lines, boxes, splines etc.). As modern GUIs have very few of these primitives, they end up sending basically every pixel one by one. That's a lot slower than just telling X to draw a button here (or to draw a 2x2x3 rectangle here...).
The grandparent was mostly referring to browsers, so I did the same. Sorry if it wasn't clear, but I was referring mostly to Firebird vs. IE than Outlook vs. Mozilla.
IMO Firebird is easily stable enough to deploy in a production environment. Faults are generally less severe, and easier to recover from.
As for the email components, Mozilla Mail as well as Thunderbird (which I use) are both top quality, but rather buggy. I've been hit by the profile bug (or one of them) of which you speak. Luckily, however, my address book is stored in my Palm and my e-mail on my IMAP server. Obviously a bug of that magnitude would prevent it's adoption by anyone depending on them for data storage. That said, I've seen my share of fucked up Outlook mailboxes as well, and they're a heck of a lot more difficult to find/repair/restore than Mozillas.
Mozilla as a browser (Firebird) is ready for the mainstream. There's yet to be a good, open-source e-mail client on Windows though. Perhaps sylpheed-claws will fill that gaping hole once the win32 port is stable. Or maybe the Ximian people will decide to port Evolution (which is, IMHO, the best GUI e-mail client in existence).
You haven't tried recent versions of Firebird or Mozilla, have you. Nearly everyone who says that hasn't used Mozilla since the terrible 4.x series, or earlier even.
Mozilla is now at least as stable as IE, approximate ly as fast, is open source, complies with standards, blocks pop-ups, has tabbed browsing, an extensions system...
If you haven't tried it, and refuse to, you're just ignorant (and closed-minded).
Type about:config in the address bar (at least in Mozilla/Firebird). Find browser.enable_automatic_image_resizing, double click it, and change it to false.
Another one, general.SmoothScroll turns off smooth scrolling.
Two of IEs most annoying 'features', I can't figure out why they were added to Firebird. At least they're easy to turn off.
It was competition until IE 5.0, I'm sorry to say. With that release IE was *way* faster, rendered pages properly, and had a better interface. Sorry to say it, but Netscape 4.x *WAS* trash, it's just that IE4 was worse.
Nowadays of course, Mozilla and Opera are both superior to IE6.
Hmm. I thought Mozilla/Firebird's UI was independent of the OS API. And it seems the behaviour he wants works fine on Firebird 0.6 on my FreeBSD system. Odd.
You're not considering rendering time. Though msn.com may be less than, say, 150KB (which my DSL can download in ~1 sec), the pages take a significantly different amount of time to actually show up. Google is almost instantaneous (due to it's simple design and lack of fluff), while msn.com takes about 3 seconds.
You're missing the entire point of his post. Let me reiterate for you:
Your car hydroplanes - the wheels are suspended in water, thus accelerating above the rate they would be moving at normally. The hypothesized *indicated* speed was 95mph (because the wheels are moving faster but your car is travelling the same speed).
Since your car has an EDR and his does not, it would be determined that *you* were the one speeding, not the speeder himself. You're screwed by your own vehicle.
Now, where this a real case, it would probably be rather obvious that the acceleration experienced was way beyond what was possible for that particular vehicle, thus debunking the data (and probably getting you off). The only issue you might have is if you were hydroplaning for longer than 5 seconds, then the acceleration data wouldn't exist. Deceleration wouldn't be reliable either because it would likely be caused by the accident itself. A sticky situation, to say the least.
I think if they make sure to regulate this properly, it could be a very good thing...but much care will need to be taken to ensure that the data is reliable and not abused (read speeding tickets).
Slashdot Mirror
They mentioned that it can send telemetry as well. Perhaps, if it's bouyant enough, it's simpler to recover from the ocean..and perhaps the water damages it less than hitting the windshield of your car on the freeway.
The irony in this is that in my opinion, the install is the *least* important part of the distribution. What I'm concerned about is post-install usability. During the install, Joe Blow is probably going to have the Linux geek that convinced him to try it at the controls anyways...if not LUGs are readily availiable for installs...and failing that, anyone who can follow directions can do it basically from scratch. I had a friend install Gentoo, and while he had a few questions, he got it working with minimal help (I mean one or two one line questions, not a hand holding) and no in-person contact. Back on topic. There's usually goign to be someone more knowledgable around during the installation, so I don't consider it a factor when choosing a distro for myself..or for others. What *is* exremely important is package managment. Anything that uses RPM is out automatically. Trying to teach someone (even someone of reasonable intelligence with a fair bit of Windows experience) to resolve RPM dependencies..then watching their face turn white as they realize that they have to download 30 other RPMs to install one program should tell you right away it's a bad choice for a newbie. Why RedHat hasn't adopted apt (or some new RPM version) is beyond me. Perhaps RedCarpet will fit the bill. Again, I'm ranting. I can set up, and configure a friend's system with Debian in not-too-long...probably not much longer than it would take him to install RedHat. But for my time, he gets a usable system in the future. I'll install a bunch of common applications, KDE, and give him some desktop shortcuts...and unless he's a gamer he'll figure it out remarkably quickly. A quick guide on how to open a console and install an application with apt is all that's really needed. RedHat et. al. really, really need to fix the package nightmare that is RPM before I'll seriously consider using them again. That might be soon, however, seeing how outside developers are now providing suggestions and input into redhat's processes.
In Windows, after approximately a month, the start menu stops auto-sorting new items in alpha order. Then I have to do sort by name whenever I install a new application so I can find it on the menu. This has gone on from Windows 95 to 2000, and I believe XP as well. It starts out fine, and just goes awry for no apparent reason, and it's a real pain in the ass usability wise.
This is bloody slow and annoying. Double clicking takes a marginal extra amount of time, but I do most of my file operations (in gui) by selecting a file and using keyboard shortcuts with my other hand. This is faster than drag/drop when I have multiple large folders open in seperate windows. Waiting 2 seconds for the file to be selected would be hell.
I just feel extremely uncomfortable working in file managers when this is the case. Even dragging and dropping i sometimes end up registering a click and opening apps...just plain clumsy.
a) you probably could've telnetted/sshd into the box, or used a serial terminal and killed the X server, I highly doubt you managed to crash the actual OS.
b) that was a configuration problem, obviously they know exactly what the hardware is, what it can do etc. and obviously again, the pilots aren't going to have access to the configuration files (which are probably in a ROM anyways).
c) I can almost guarantee they wouldn't be using X
The only time I've ever seen a kernel panic was because I filled a hard drive and the kernel couldn't write to it...oh and of course if / isn't mountable at boot time. Other than that, every crash I've seen (with stable kernels) has been a software issue, not an OS issue.
The myth about Windows crashing twice a day is definitely a myth, or perhaps you could call it an exaggeration. I don't crash Windows enough for it to really piss me off, but it does happen from time to time.
That's not been my experience. In college I'd regularly use my discman to listen to tunes during class. With top-line non-rechargable batteries (Duracell Ultra et al.), it would last me from 3-4 weeks (non continuous use obviously). If I ever picked up a set at the student union (they didn't sell name brand batteries) because I didn't have spares with me, they would last much less time, generally a week or two. Perhaps it has to do with the specific batteries and their construction with relation to my cd player, but I suspect, as with most anything else, that you get what you pay for.
Please mod the parent up, buying the cheapest batteries you can find is definitely bad advice. There is even a large quality margin in standard non-rechargable Alkaline batteries, not to mention the more complex rechargables. I always buy Duracell or Energizer alkalines, since they generally last 2-3 times longer than the el-cheapo brand. As for rechargables, the Alkaline rechargables (Energizer has some, as does a company called PureEnergy) are pretty good for the cost. The chargers are inexpensive, and the batteries don't cost much more than standard alkalines...the downside is they only last for 20-50 charges.
Or you could just use Canadian pennies instead...ours our still pure copper (stingy yanks :P).
I believe the value of the copper itself is greater than that of the penny.
You can run X on a 486 with decent performance, if you don't &*(% the thing up and saddle it with a bunch of useless crap.
The fact remains that Windows is still faster when saddled with all that useless crap, and at least as fast without it. Something needs to change (and it goes deeper than "don't run it, it's slow").
Fonts look far better in pretty much every application on my Gentoo and FreeBSD systems (both were fully updated within the past month).
;)
I'd blame either RedHat, or the slow release-test-upgrade cycle of such a commercial distribution. Font rendering in *ix has definitely matured in the past 6 months or so, get a recent distro and give your eyes a well deserved break
3D is a whole 'nother ballgame. OpenGL is implemented in X (whereas GTK or QT are not). Just as it would be on a Windows box, each polygon is mapped in the scene, textured, and moved around etc. just as it would be under any other OpenGL application.
The discrepancy is that X primitive libraries (QT, GTK etc.) have to draw each individual X primitive (lines, boxes, splines etc.). As modern GUIs have very few of these primitives, they end up sending basically every pixel one by one. That's a lot slower than just telling X to draw a button here (or to draw a 2x2x3 rectangle here...).
The grandparent was mostly referring to browsers, so I did the same. Sorry if it wasn't clear, but I was referring mostly to Firebird vs. IE than Outlook vs. Mozilla.
IMO Firebird is easily stable enough to deploy in a production environment. Faults are generally less severe, and easier to recover from.
As for the email components, Mozilla Mail as well as Thunderbird (which I use) are both top quality, but rather buggy. I've been hit by the profile bug (or one of them) of which you speak. Luckily, however, my address book is stored in my Palm and my e-mail on my IMAP server. Obviously a bug of that magnitude would prevent it's adoption by anyone depending on them for data storage. That said, I've seen my share of fucked up Outlook mailboxes as well, and they're a heck of a lot more difficult to find/repair/restore than Mozillas.
Mozilla as a browser (Firebird) is ready for the mainstream. There's yet to be a good, open-source e-mail client on Windows though. Perhaps sylpheed-claws will fill that gaping hole once the win32 port is stable. Or maybe the Ximian people will decide to port Evolution (which is, IMHO, the best GUI e-mail client in existence).
Don't you mean join the Klub?
You haven't tried recent versions of Firebird or Mozilla, have you. Nearly everyone who says that hasn't used Mozilla since the terrible 4.x series, or earlier even. Mozilla is now at least as stable as IE, approximate ly as fast, is open source, complies with standards, blocks pop-ups, has tabbed browsing, an extensions system... If you haven't tried it, and refuse to, you're just ignorant (and closed-minded).
Type about:config in the address bar (at least in Mozilla/Firebird). Find browser.enable_automatic_image_resizing, double click it, and change it to false.
Another one, general.SmoothScroll turns off smooth scrolling.
Two of IEs most annoying 'features', I can't figure out why they were added to Firebird. At least they're easy to turn off.
My bank simply lists Netscape 4.06 or better. And it works great with Firebird. Hell, they even mention Linux in the supported browsers/OSs:
* Microsoft Internet Explorer (IE) version 4.01 Service Pack 2 (SP2) or higher
* Netscape Navigator version 4.06 or higher
*Beta browsers are not supported
Operating System:
You must have a 32-bit operating system - Windows 95, 98, 2000,NT, ME; MAC 7.5 and higher; Linux 2.2
1.4 == 1.4RC3, it's just a name change. Keep RC3, it's the same build.
It was competition until IE 5.0, I'm sorry to say. With that release IE was *way* faster, rendered pages properly, and had a better interface. Sorry to say it, but Netscape 4.x *WAS* trash, it's just that IE4 was worse.
Nowadays of course, Mozilla and Opera are both superior to IE6.
Hmm. I thought Mozilla/Firebird's UI was independent of the OS API. And it seems the behaviour he wants works fine on Firebird 0.6 on my FreeBSD system. Odd.
You're not considering rendering time. Though msn.com may be less than, say, 150KB (which my DSL can download in ~1 sec), the pages take a significantly different amount of time to actually show up. Google is almost instantaneous (due to it's simple design and lack of fluff), while msn.com takes about 3 seconds.
I went to ASS - Aldergrove Secondary School.
You're missing the entire point of his post. Let me reiterate for you:
Your car hydroplanes - the wheels are suspended in water, thus accelerating above the rate they would be moving at normally. The hypothesized *indicated* speed was 95mph (because the wheels are moving faster but your car is travelling the same speed).
Since your car has an EDR and his does not, it would be determined that *you* were the one speeding, not the speeder himself. You're screwed by your own vehicle.
Now, where this a real case, it would probably be rather obvious that the acceleration experienced was way beyond what was possible for that particular vehicle, thus debunking the data (and probably getting you off). The only issue you might have is if you were hydroplaning for longer than 5 seconds, then the acceleration data wouldn't exist. Deceleration wouldn't be reliable either because it would likely be caused by the accident itself. A sticky situation, to say the least.
I think if they make sure to regulate this properly, it could be a very good thing...but much care will need to be taken to ensure that the data is reliable and not abused (read speeding tickets).
Perhaps not, but they'll instruct the computer to do exactly that ;)
Excuse my blindness. For some reason I didn't see Gecko on the table.
Indeed, but not all of them, sometimes the audio doesn't work.
Ditto, WMP7+ is worse than QT.