HTTP makes provision for caching and caching proxy servers, which does give visiting machines the right to archive and retransmit the works. Of course, there are expiration headers, but there is nothing that says it has to be purged from the cache once it expires.
Besides which, what defines an ISP? I do work for a school that shares an Internet line with a nearby company; the router is in the school, and the company can use the school's cache server and mail relay. Does the school have to log everything? They certainly can't pay for it.
Then again, if the government would provide cash for some upgrades, I'm sure they wouldn't mind.
Hear, hear... but you missed PCI hotplug. (Now that's cool!) Many, many IT people deploy Proliants because, like it or not, "Proliant" is a very respected name in the datacenter.
I think the big win in this article is not Red Hat on Itanium, but rather having Linux available directly from the manufacturer on the Proliant line. There's a lot of people that will feel a lot better now that they can get HP backing both the hardware and the software on their brand new (expensive) server. Plus, it gives us something to bring to management ("HP is offering it, it must be good"), etc.
Bah! Google is capable of producing accurate, concise, and to-the-point results, if you give it an accurate, concise, and to-the-point query. A search for "gladiator historical accuracy", for instance, would fit the bill, don't you think? Look at the results. All of them relate to your search query.
What I'd prefer is the ability to narrow the focus of my searches.
Then focus your search query. Tell Google what you want and it will find it for you.
One of the things that might give Mozilla a fighting change on corporate desktops would be.MSI installers. Yes, they can be built by your local friendly sysadmin, but ready-for-use packages would make rollouts quick and painless for many places. In a Windows 2000 environment, Mozilla can be deployed onto as many computers as desired with a right click, a left click, a double click, and an OK.
You know... I bet someone has already done this, but hasn't made it public.
Good MS admins are very hard to come by, because in my experience if they get good at Microsoft software -- actually understanding how it's supposed to work -- they get sick of it and switch to *nix. As much as people don't want to admit, Windows really is an inferior operating system that has no business in the data center. (Unfortunately, demands of Windows on the desktop may bring it in, but that can be minimized with tools like Samba.)
In any case, I agree with you and I found your disk monkey comment amusing:-)
But your statement about people not wanting to switch off of Novell/Lotus/NT 3.51... I have to ask why should they?
Because supporting Novell, Lotus Notes, and NT 3.51 on the desktop is a nightmare, that's why. I agree in that servers can be left alone (heck, I have a box on a Linux 2.0 kernel that has yet to reboot), but on the desktop it does not make sense to use software that was obselete in 1996.
when the heck are they going to loose the "Owner, Group, Other" file permissions?
It's efficient and not all that hard to use. It's just as flexible as NT 3.5 and Novell 2.x -- what can NT 3 do that *nix file permissions can't?
Have you ever tried to admin a system with over a couple hundred people and had to deal with this. It sucks when compared to Novell and NT.
I administer a network of a couple hundred people. It's my day job. I've had few file permissions issues, and far more issues with NT boxen shooting themselves in the foot. (In the past week, two Win2k print servers having their Print Spooler programs stop responding under zero load for no apparent reason. Of course, it was still "Started", so Win2k didn't give it a swift kick in the pants.)
Besides, if the grass is greener on the Windows side of the fence, why does *nix beat the tar out of Windows in terms of administration? I can reconfigure network cards via SSH, even the one I'm connected through. I can change the hostname without rebooting. I can create and schedule complex tasks with confidence that they will actually run. I can swap out critical system components without creating any downtime. I can partition, format, and allocate new storage space on a file server under full load. Not to mention the fact that the *nix command line is many, many, many times more powerful than the Windows point-and-click interface.
until Solaris 9, the UNIX answer to directory service was LDAP
So, they are officially supporting Linux as a target platform, because they recognize it can have a lower TCO.
Supporters of Linux point to a lower total cost of ownership because they can leverage their UNIX expertise on a free and open operating system running on cost-effective Intel architecture workstations.
Which is true, of course, if they have UNIX expertise in-house. MCSEs are a dime a dozen, but good UNIX admins are quite expensive. If you go the consulting route, you get screwed with huge fees. If you train your personell, you get screwed with long courses and a decent change they simply won't get it.
Then again, if you have a competent staff or a big budget, *nix all the way. I know firsthand that the grass really is greener on the *nix side of the fence, but sometimes that's not feasible for large corporations. (Besides, many corporations are brainless and/or inflexible, and won't switch away from their Novell file servers, Lotus Notes 2.0, and NT 3.51, but they have bigger problems.)
Looking at a solar eclipse isn't any more dangerous than looking at the sun any other day, except that people are more likely to keep staring at the eclipse than the sun on any other day. So, in a way, it is an old wives' tale, but there is some kind of merit behind it.
Curiosity killed the cat, and curiosity can blind a human.
Low-cost for consumers, possibly, but do you have any idea how much a satellite costs? Besides, you still need a modem for uploads (and HTTP requests), so you get screwed with latency from the modem, the 'net, and routing through a satellite.
If you place repeaters everywhere to form a virtual circuit, you run into bandwidth problems with lots of users in said 15-30 mile stretch, because they would share the same "pipe" (as it were). Of course, it would still beat out dialup.
Indeed, AMD has "lost" this round -- Intel has faster chips, better marketing, and lots of power in the pre-built system market (which is very large). Since this last round, however, AMD has gained ground, and is quickly closing the gap.
The only question remains in the 64-bit arena -- are people going to continually extend the x86 instruction set? (Like HTTP is now used for interprocess communication via SOAP, multimedia streaming, and executing SQL.) Windows is an excellent example of the fact that people like backwards compatibility, so unless someone makes an x86 emulator that exposes the Windows APIs for the IA-64, AMD will have a huge lead on Intel.
Re:Re-Inventing the wheel..again !
on
OGRE GPL'ed 3D Engine
·
· Score: 5, Insightful
600 3d engines... yet none of them are "really good"
Uh, you did notice the part saying "last modified on 23.6.2000", didn't you? Two years is a long time; Crystal Space and others have come a very long time since then.
so why can't developers focus and colaborate on making 1 really good engine and make it "perfect" instead of fragmenting into their own little GPL projects that are nothing more IMO than a waste of time ?
3d engines have become a much larger project since two years ago, what with people wanting platform independence, 3d acceleration, and lots of other technically demanding features. The only 3d engines that are getting anywhere have lots of developers coordinating.
I guess time is what programmers waste best
He says smugly, as he pushes the Submit button and millions (if not billions) of lines of code are executed for his message to be posted.
Chess is mostly a bunch of recursive procedures; i.e. if I move here he can move here here or here. So, given that, splitting up the workload wouldn't be too difficult at all -- just tell another node to score this board layout and keep crunching. Excellent application for distributed computing.
My only concern with a distributed chess processing thing is security. What's to stop someone from making their node return random results?
Why write a new port system (Portage) instead of using BSD's version?
In one sentence, because Portage is much better in so many ways. One of the design philosophies of the.ebuild syntax was to make it an analog of what you'd type to install the program manually, thus making Portage very easy to learn and modify to your needs. We also have OpenBSD-style "fake" installs, safe unmerging, system profiles, package masking, a real dependency system, and lots of other good stuff.
It's meaningful because on a portable you want decent quality with as little size as possible. Thus, given the option of using MP3 or WMA, it makes more sense to use WMA because you can get more files in the same space.
Again, on a portable music player, would you rather run WAV or MP3? Of course, you would want to run MP3, because you can still get 6:1 compression over WAV with no noticable loss of quality (on headphones at least). Likewise, I would rather run WMA than MP3, because I can (according to your earlier post) get the same overall quality in half the space of MP3. So, instead of having 2 hours of music, I have 4. In this case, it is meaningful because choosing WMA over MP3 doubled capacity of said portable music player.
As far as comparing codecs goes, perhaps it is not meaningful. But then again, if it produces the same sound in half the size, I would argue even if they are in a different class that WMA is far superior.
Also, for reference, it's an 8-month old dual-1.0 GHz P3 box, that (for my uses) will be quite competitive to your 2.53 GHz P4:-)
So... the fact that WMA provides equivalent audio in half the size doesn't matter?
Also, in my experience at least, WMA at 64 kilobit sounds better than MP3 at 128 kilobit. Remember that CD-Quality is 44.1 kHz, 16 bits per sample, neither of which (being lossy formats) can provide. Music that I listen to often contains a lot of near-random noise in the high end (cymbals, some other percussion), and with MP3 I have to jack up the bitrate to 256 to even come close to WMA at 64.
Look, I don't know about you, but I don't see why "comparing MP3 and WMA at equivalent bitrates and saying one sounds worse" is unfair. WMA can produce the same sound as MP3 using less data -- what makes that a bad comparison?
Anyone else notice that the supported operating systems are "MS Windows 98/2000/Me"? Why would it work under Win2k and not XP? Seems quite strange, seeing as the two are quite similar, and especially because XP is Microsoft's officially-backed "next-generation" home OS.
Well, when you consider that the LCD is two and a half inches wide and running at a resolution of 160x234, 128 MB should be enough for a lot of video. 160x234 isn't exactly huge, you know, and you can get by at a lot lower bitrates than with higher resolution screens. (For instance, 250 kilobit WMV would look decent.)
Granted, it would take a lot of time to crunch all your movies from their original formats to fit this little thing, but you could get a whole lot more video on it by repacking.
Easy -- call Qwest tech support. If you can ignore the periodic "All representatives are busy" and "This call is monitored for quality" notices, you'll have free, portable music on your cell phone for hours at a time.
Slashdot Mirror
Good thing he doesn't work under CrossOver Office...
Stupid paperclip.
HTTP makes provision for caching and caching proxy servers, which does give visiting machines the right to archive and retransmit the works. Of course, there are expiration headers, but there is nothing that says it has to be purged from the cache once it expires.
Are cache servers in violation of copyright?
Besides which, what defines an ISP? I do work for a school that shares an Internet line with a nearby company; the router is in the school, and the company can use the school's cache server and mail relay. Does the school have to log everything? They certainly can't pay for it.
Then again, if the government would provide cash for some upgrades, I'm sure they wouldn't mind.
No, that was an addendum to the above feature list.
Hear, hear... but you missed PCI hotplug. (Now that's cool!) Many, many IT people deploy Proliants because, like it or not, "Proliant" is a very respected name in the datacenter.
I think the big win in this article is not Red Hat on Itanium, but rather having Linux available directly from the manufacturer on the Proliant line. There's a lot of people that will feel a lot better now that they can get HP backing both the hardware and the software on their brand new (expensive) server. Plus, it gives us something to bring to management ("HP is offering it, it must be good"), etc.
Then focus your search query. Tell Google what you want and it will find it for you.
One of the things that might give Mozilla a fighting change on corporate desktops would be .MSI installers. Yes, they can be built by your local friendly sysadmin, but ready-for-use packages would make rollouts quick and painless for many places. In a Windows 2000 environment, Mozilla can be deployed onto as many computers as desired with a right click, a left click, a double click, and an OK.
You know... I bet someone has already done this, but hasn't made it public.
Doesn't everyone read the EULAs?
Good MS admins are very hard to come by, because in my experience if they get good at Microsoft software -- actually understanding how it's supposed to work -- they get sick of it and switch to *nix. As much as people don't want to admit, Windows really is an inferior operating system that has no business in the data center. (Unfortunately, demands of Windows on the desktop may bring it in, but that can be minimized with tools like Samba.)
In any case, I agree with you and I found your disk monkey comment amusing :-)
And I agree.
Because supporting Novell, Lotus Notes, and NT 3.51 on the desktop is a nightmare, that's why. I agree in that servers can be left alone (heck, I have a box on a Linux 2.0 kernel that has yet to reboot), but on the desktop it does not make sense to use software that was obselete in 1996.
It's efficient and not all that hard to use. It's just as flexible as NT 3.5 and Novell 2.x -- what can NT 3 do that *nix file permissions can't?
I administer a network of a couple hundred people. It's my day job. I've had few file permissions issues, and far more issues with NT boxen shooting themselves in the foot. (In the past week, two Win2k print servers having their Print Spooler programs stop responding under zero load for no apparent reason. Of course, it was still "Started", so Win2k didn't give it a swift kick in the pants.)
Besides, if the grass is greener on the Windows side of the fence, why does *nix beat the tar out of Windows in terms of administration? I can reconfigure network cards via SSH, even the one I'm connected through. I can change the hostname without rebooting. I can create and schedule complex tasks with confidence that they will actually run. I can swap out critical system components without creating any downtime. I can partition, format, and allocate new storage space on a file server under full load. Not to mention the fact that the *nix command line is many, many, many times more powerful than the Windows point-and-click interface.
Ever hear of NIS?
So, they are officially supporting Linux as a target platform, because they recognize it can have a lower TCO.
Which is true, of course, if they have UNIX expertise in-house. MCSEs are a dime a dozen, but good UNIX admins are quite expensive. If you go the consulting route, you get screwed with huge fees. If you train your personell, you get screwed with long courses and a decent change they simply won't get it.
Then again, if you have a competent staff or a big budget, *nix all the way. I know firsthand that the grass really is greener on the *nix side of the fence, but sometimes that's not feasible for large corporations. (Besides, many corporations are brainless and/or inflexible, and won't switch away from their Novell file servers, Lotus Notes 2.0, and NT 3.51, but they have bigger problems.)
Looking at a solar eclipse isn't any more dangerous than looking at the sun any other day, except that people are more likely to keep staring at the eclipse than the sun on any other day. So, in a way, it is an old wives' tale, but there is some kind of merit behind it.
Curiosity killed the cat, and curiosity can blind a human.
Low-cost for consumers, possibly, but do you have any idea how much a satellite costs? Besides, you still need a modem for uploads (and HTTP requests), so you get screwed with latency from the modem, the 'net, and routing through a satellite.
If you place repeaters everywhere to form a virtual circuit, you run into bandwidth problems with lots of users in said 15-30 mile stretch, because they would share the same "pipe" (as it were). Of course, it would still beat out dialup.
Indeed, AMD has "lost" this round -- Intel has faster chips, better marketing, and lots of power in the pre-built system market (which is very large). Since this last round, however, AMD has gained ground, and is quickly closing the gap.
The only question remains in the 64-bit arena -- are people going to continually extend the x86 instruction set? (Like HTTP is now used for interprocess communication via SOAP, multimedia streaming, and executing SQL.) Windows is an excellent example of the fact that people like backwards compatibility, so unless someone makes an x86 emulator that exposes the Windows APIs for the IA-64, AMD will have a huge lead on Intel.
Uh, you did notice the part saying "last modified on 23.6.2000", didn't you? Two years is a long time; Crystal Space and others have come a very long time since then.
3d engines have become a much larger project since two years ago, what with people wanting platform independence, 3d acceleration, and lots of other technically demanding features. The only 3d engines that are getting anywhere have lots of developers coordinating.
He says smugly, as he pushes the Submit button and millions (if not billions) of lines of code are executed for his message to be posted.
Chess is mostly a bunch of recursive procedures; i.e. if I move here he can move here here or here. So, given that, splitting up the workload wouldn't be too difficult at all -- just tell another node to score this board layout and keep crunching. Excellent application for distributed computing.
My only concern with a distributed chess processing thing is security. What's to stop someone from making their node return random results?
See the Gentoo Linux FAQ:
Why write a new port system (Portage) instead of using BSD's version?
.ebuild syntax was to make it an analog of what you'd type to install the program manually, thus making Portage very easy to learn and modify to your needs. We also have OpenBSD-style "fake" installs, safe unmerging, system profiles, package masking, a real dependency system, and lots of other good stuff.
In one sentence, because Portage is much better in so many ways. One of the design philosophies of the
It's meaningful because on a portable you want decent quality with as little size as possible. Thus, given the option of using MP3 or WMA, it makes more sense to use WMA because you can get more files in the same space.
:-)
Again, on a portable music player, would you rather run WAV or MP3? Of course, you would want to run MP3, because you can still get 6:1 compression over WAV with no noticable loss of quality (on headphones at least). Likewise, I would rather run WMA than MP3, because I can (according to your earlier post) get the same overall quality in half the space of MP3. So, instead of having 2 hours of music, I have 4. In this case, it is meaningful because choosing WMA over MP3 doubled capacity of said portable music player.
As far as comparing codecs goes, perhaps it is not meaningful. But then again, if it produces the same sound in half the size, I would argue even if they are in a different class that WMA is far superior.
Also, for reference, it's an 8-month old dual-1.0 GHz P3 box, that (for my uses) will be quite competitive to your 2.53 GHz P4
So... the fact that WMA provides equivalent audio in half the size doesn't matter?
Also, in my experience at least, WMA at 64 kilobit sounds better than MP3 at 128 kilobit. Remember that CD-Quality is 44.1 kHz, 16 bits per sample, neither of which (being lossy formats) can provide. Music that I listen to often contains a lot of near-random noise in the high end (cymbals, some other percussion), and with MP3 I have to jack up the bitrate to 256 to even come close to WMA at 64.
Look, I don't know about you, but I don't see why "comparing MP3 and WMA at equivalent bitrates and saying one sounds worse" is unfair. WMA can produce the same sound as MP3 using less data -- what makes that a bad comparison?
"I can see it now: 'Pink raisin found in tub, boy's whereabouts unknown'"
Think about how wrinkly you'd get after watching a movie in it.
Anyone else notice that the supported operating systems are "MS Windows 98/2000/Me"? Why would it work under Win2k and not XP? Seems quite strange, seeing as the two are quite similar, and especially because XP is Microsoft's officially-backed "next-generation" home OS.
Well, when you consider that the LCD is two and a half inches wide and running at a resolution of 160x234, 128 MB should be enough for a lot of video. 160x234 isn't exactly huge, you know, and you can get by at a lot lower bitrates than with higher resolution screens. (For instance, 250 kilobit WMV would look decent.)
Granted, it would take a lot of time to crunch all your movies from their original formats to fit this little thing, but you could get a whole lot more video on it by repacking.
I grammar at good to be!
Easy -- call Qwest tech support. If you can ignore the periodic "All representatives are busy" and "This call is monitored for quality" notices, you'll have free, portable music on your cell phone for hours at a time.
Just be careful not to use up all your minutes.