Yes, and this option is, 99.9% of the time, fundamentally BROKEN. POP3 has no concept of message synchronization, so it has to make a blind guess about what's new and what's not. If another client messes with headers in the messages (to update message UIDs, timestamps, etc.), the POP client will frequently get confused - and redownload ALL mail, because it doesn't know what's new and what isn't. I've seen this happen so many times - the client ends up downloading copy after copy of messages the user has already read, then I end up in a discussion like:
user: "your server is broken! what's the matter with it?" me: "Are you using POP?" user: "yes" me: "Did you turn on 'leave messages on server'?" user: "yes!" me: "Don't do that, it doesn't work right." user: "but it was working fine! you broke the server! fix it!" me: "No, the server is _not_ broken, if you want to do that, use IMAP." user: [more complaining, whining about how they don't like the way IMAP works, etc.]
Please DO NOT encourage more people to use this fundamentally broken option! Just because it became a popular feature to provide doesn't mean it works right.
Actually, the phrase is "c'est la vie", which is French, and means, basically, "such is life" or "that's life".
No, I never took French in school. Yes, I know this to be true. Sorry, but it bugs me when people totally butcher phrases like that.
Re:Darl does NOT deserve ANY respect.
on
SCOrched Earth
·
· Score: 1
Oh no, SCO doesn't want to litigate everybody out of business - they just want to litigate until everyone is sufficiently afraid of them that they'll give SCO all their money. "All your $$$ are belong to us. Bitches." Because apparently all OSes are using SCO's "intellectual property", so we should all be whipping out prayer mats and kneeling in the direction of the SCO headquarters in Bumf**k, Utah.
Mine's not quite 3 digits - but pretty damn close. I still remember Taco soliciting people to come and read his site on IRC. Hah. I held off getting an ID until talk started going around about making Anonymous Cowards less visible, then I got one...
Funny, I seem to remember running Wine through qemu on my PowerBook to run sol.exe, just for the hell of it. Wine by itself can't, but that doesn't mean solutions aren't available to the problem.
It's not an emulator by any accepted definition by anyone who knows about computer science. It's a binary loader with a Win32 ABI. It doesn't have to emulate anything on an x86 system. Now if I run Wine on my PowerBook to run a Win32 binary, _then_ I'm emulating. (Emulating an x86, that is!)
Moreover some data on your disk (e.g. the boot sector) need to be copied to the same sector on the target drive. This isn't a "file" so you couldn't do it no matter what filesystem you used. Even under unix you'd have to use something like dd.
Yes, but with UNIX/Linux, you can boot from a CD, remake the filesystems and restore the backup, install your bootloader, and then reboot into the restored system, _exactly_ as it was at the time of the backup. It doesn't depend on "magical" file locations. I've done this myself, so I know it works. I've also had to deal with Windows systems, and I know exactly what he's talking about. You can do "backups" to tape or other media using Windows' "backup" tools, but you can't effectively restore those backups. All they're really good for is retrieving the data files. You'll never be able to restore the bootsector, the pagefile has to be contiguous or else Windows is unhappy, etc., etc.
It seem to me based on the first line of your question that you have unreasonable expectations and a negative opinion of Microsoft.
Yes, and one that's quite muchly justified, considering the fact that I've never seen a successful restore of Windows from tape or CD without disk-imaging software being used, and even then, the software does tend to be moody and flaky.
Amen. My Debian boxes just hum away, doing what I tell them to do, no questions, no bullshit. Update packages periodically, and it works like a charm. I have a PowerBook (Pismo) and a PowerMac 7500, and things work nicely, just like on my x86 systems.
As far as Windows... Windows just eats itself every so often, on every machine I've ever seen in run on. A friend of mine tells me every so often about the huge Unisys 32-way Windows box they have where he works, and more than once, Windows has just eaten itself on that (multi-million dollar) Windows system. They have to pay a Windows "expert" to come in, look at it, and say "eh. Have to just reinstall Windows." Confidence-inspiring, isn't it?
Check your requirements list - I'm pretty sure MacOS X 10.3 completely deprecates _all_ OldWorld systems now - the last officially supported OldWorld systems were the beige G3s, and I believe they are now officially _not_ supported. So you may be stuck with 10.2.x on your beige machine.
The developer of MacOnLinux is supposedly looking at hacking it to run in OS X, so you can run OS 9 and OS X inside MacOS X. There is some preliminary support in its codebase for running a Linux kernel inside the virtual machine, and with appropriate changes to a kernel, it could use a virtual drive, and run a full Linux install, inside the virtual machine. Unfortunately, you can't use an unmodified kernel. You'd have to have support for the MOL block device interface - it doesn't try to pretend it has an IDE interface. MacOS and OS X load drivers at boot (OS 9 loads them out of OF ROMs) to support it, so it's pretty much transparent there, but it might be slightly more involved with Linux.
If you know what you're doing, he would probably appreciate the help. If not, don't overwhelm Samuel with "ooh, I heard you're gonna do this! whenwhenwhenIWANTITNOWGIMMEGIMME!" This is open source, and he's contributing to the community, so be nice to him, he's doing you a favor by trying at all.
Not in the slightest. debian-installer is still being worked on. My guess is this will be yet another effort, like the original Progeny installer (that was used in Progeny Debian) that will be isolated to x86 only. I know that YellowDog made Anaconda work on PowerPC for their distro, though. In any case, I don't expect to see any takeup from the Debian developers of this. Unless it can be made to work on ALL the supported arches with significantly less effort than what debian-installer will take to finish, I can't see why they'd care.
Yes, it makes it easy to _install_ but keep in mind how many architectures Debian supports: x86, Alpha, PowerPC (which covers OldWorld and NewWorld PowerMacs, CHRP and PReP systems, and Amiga PowerUP machines), MIPS, MIPS little-endian, ARM/StrongARM, HP PA-RISC, Motorola 680x0 (which covers 68k Macs, old Suns, Ataris, and Amigas), IA64, SPARC, and S390. How much effort has it already taken to make Anaconda work as a Debian installer for one architecture? (And note those are just the architectures that woody supports, I'm not even counting the experimental ones that are still only in sid.)
Oh yeah... cause you know, I'm reinstalling Debian all the time... wait. That's right, I'm NOT. I last installed Debian on my home desktop about 4 or 5 _years_ ago, and it's never once needed reinstalling. Though debian-installer, once it's done, should improve things significantly, how often do you really see the installer? Seriously, how often?
From what I understood, the Progeny installer went mostly ignored in Debian proper because it only worked on x86, and would take yet more retooling and fighting to make work on the _many_ other architectures that Debian supports (like the PowerBook Pismo I'm typing on now - a PowerPC). The Debian developers wanted a truly architecture-neutral, extensible installer framework, and nobody really provides that. I don't know how much work the YellowDog crew had to put in to make Anaconda work in YDL, but I have no doubts that it was significant. Unfortunately, if you want a product you can package and sell, the "must support ALL arches" requirement seems to be the first to get kicked to the curb.
AOL doesn't care about educating their customers. If they did that, their customers might realize how stupid, patronizing, and generally crap their service is, and go elsewhere. Keep 'em dumb, so you can keep taking their money - It's the Capitalist(TM)(C)(R) Way!
I think you need to go back to school, and take some basic math classes again. Or maybe that's the "new math" I keep hearing about? But then, what do I know...
SCSI is NOT serial. FireWire/IEEE1394/i.Link is serial. SCSI is not, and never has been. If you're going to talk about fair comparisons, then let's be fair, otherwise please stop talking.
How exactly would MS have learned jack about writing a 64-bit OS from NT/AXP? It was run in 32-bit compatibility mode on the Alpha. It was NEVER a 64-bit OS. Which I thought made it a really huge waste of a CPU - although it did run NT rather well (a former employer looked at an Alpha running NT for a project they were working on, and I sat in on the demo).
I'm pretty sure he means "free" as in both meanings - free as in beer, and as in speech. And yes, then public domain code is the "free-est". It costs you nothing, and you can take it and do ANYTHING you want with it - make your own operating system, make a product for sale, whatever. You don't have to abide by any license terms, because once it's in the public domain, it belongs to everyone, basically.
Remember that "free" has multiple meanings, and that a person can use it to mean more than one of those at a time.
BackOrifice was by the Cult of the Dead Cow, not l0pht. They wrote the l0phtcrack password cracker, and probably some other stuff. Don't assign credit (or blame) to the wrong people.
Yeah, I thought that was pretty weak too. You'd think they could manage to have someone do the reporting on this stuff who actually knows something about it. But I suppose, it makes the average Joe feel better if the reporter uses the same sort of terms he would use, because he doesn't feel challenged by the article. It's "We know you think this is bad, we think it's bad too. We're gonna talk about it in a way that makes you feel comfortable, and that makes you believe something is being done." That's today's journalism, kids.
Or you can use something like mPlayer or Xine (it's been ported to Win32, hasn't it?) that demultiplexes QuickTime on its own, and has its own codecs for SVQ1, SVQ3, AAC audio, and all the other formats that QuickTime and whatever other formats use.
Slashdot Mirror
Yes, and this option is, 99.9% of the time, fundamentally BROKEN. POP3 has no concept of message synchronization, so it has to make a blind guess about what's new and what's not. If another client messes with headers in the messages (to update message UIDs, timestamps, etc.), the POP client will frequently get confused - and redownload ALL mail, because it doesn't know what's new and what isn't. I've seen this happen so many times - the client ends up downloading copy after copy of messages the user has already read, then I end up in a discussion like:
user: "your server is broken! what's the matter with it?"
me: "Are you using POP?"
user: "yes"
me: "Did you turn on 'leave messages on server'?"
user: "yes!"
me: "Don't do that, it doesn't work right."
user: "but it was working fine! you broke the server! fix it!"
me: "No, the server is _not_ broken, if you want to do that, use IMAP."
user: [more complaining, whining about how they don't like the way IMAP works, etc.]
Please DO NOT encourage more people to use this fundamentally broken option! Just because it became a popular feature to provide doesn't mean it works right.
Actually, the phrase is "c'est la vie", which is French, and means, basically, "such is life" or "that's life".
No, I never took French in school. Yes, I know this to be true. Sorry, but it bugs me when people totally butcher phrases like that.
Oh no, SCO doesn't want to litigate everybody out of business - they just want to litigate until everyone is sufficiently afraid of them that they'll give SCO all their money. "All your $$$ are belong to us. Bitches." Because apparently all OSes are using SCO's "intellectual property", so we should all be whipping out prayer mats and kneeling in the direction of the SCO headquarters in Bumf**k, Utah.
Sorry Darl, not gonna happen.
Mine's not quite 3 digits - but pretty damn close. I still remember Taco soliciting people to come and read his site on IRC. Hah. I held off getting an ID until talk started going around about making Anonymous Cowards less visible, then I got one...
That was "Chips & Dips", actually.
Or you could use an OTP system like OPIE, and your users can just get a batch of passwords, write them down, and use them up.
Funny, I seem to remember running Wine through qemu on my PowerBook to run sol.exe, just for the hell of it. Wine by itself can't, but that doesn't mean solutions aren't available to the problem.
It's not an emulator by any accepted definition by anyone who knows about computer science. It's a binary loader with a Win32 ABI. It doesn't have to emulate anything on an x86 system. Now if I run Wine on my PowerBook to run a Win32 binary, _then_ I'm emulating. (Emulating an x86, that is!)
Moreover some data on your disk (e.g. the boot sector) need to be copied to the same sector on the target drive. This isn't a "file" so you couldn't do it no matter what filesystem you used. Even under unix you'd have to use something like dd.
Yes, but with UNIX/Linux, you can boot from a CD, remake the filesystems and restore the backup, install your bootloader, and then reboot into the restored system, _exactly_ as it was at the time of the backup. It doesn't depend on "magical" file locations. I've done this myself, so I know it works. I've also had to deal with Windows systems, and I know exactly what he's talking about. You can do "backups" to tape or other media using Windows' "backup" tools, but you can't effectively restore those backups. All they're really good for is retrieving the data files. You'll never be able to restore the bootsector, the pagefile has to be contiguous or else Windows is unhappy, etc., etc.
It seem to me based on the first line of your question that you have unreasonable expectations and a negative opinion of Microsoft.
Yes, and one that's quite muchly justified, considering the fact that I've never seen a successful restore of Windows from tape or CD without disk-imaging software being used, and even then, the software does tend to be moody and flaky.
Apparently the author has never heard of entity references?
Amen. My Debian boxes just hum away, doing what I tell them to do, no questions, no bullshit. Update packages periodically, and it works like a charm. I have a PowerBook (Pismo) and a PowerMac 7500, and things work nicely, just like on my x86 systems.
As far as Windows... Windows just eats itself every so often, on every machine I've ever seen in run on. A friend of mine tells me every so often about the huge Unisys 32-way Windows box they have where he works, and more than once, Windows has just eaten itself on that (multi-million dollar) Windows system. They have to pay a Windows "expert" to come in, look at it, and say "eh. Have to just reinstall Windows." Confidence-inspiring, isn't it?
Check your requirements list - I'm pretty sure MacOS X 10.3 completely deprecates _all_ OldWorld systems now - the last officially supported OldWorld systems were the beige G3s, and I believe they are now officially _not_ supported. So you may be stuck with 10.2.x on your beige machine.
The developer of MacOnLinux is supposedly looking at hacking it to run in OS X, so you can run OS 9 and OS X inside MacOS X. There is some preliminary support in its codebase for running a Linux kernel inside the virtual machine, and with appropriate changes to a kernel, it could use a virtual drive, and run a full Linux install, inside the virtual machine. Unfortunately, you can't use an unmodified kernel. You'd have to have support for the MOL block device interface - it doesn't try to pretend it has an IDE interface. MacOS and OS X load drivers at boot (OS 9 loads them out of OF ROMs) to support it, so it's pretty much transparent there, but it might be slightly more involved with Linux.
If you know what you're doing, he would probably appreciate the help. If not, don't overwhelm Samuel with "ooh, I heard you're gonna do this! whenwhenwhenIWANTITNOWGIMMEGIMME!" This is open source, and he's contributing to the community, so be nice to him, he's doing you a favor by trying at all.
Not in the slightest. debian-installer is still being worked on. My guess is this will be yet another effort, like the original Progeny installer (that was used in Progeny Debian) that will be isolated to x86 only. I know that YellowDog made Anaconda work on PowerPC for their distro, though. In any case, I don't expect to see any takeup from the Debian developers of this. Unless it can be made to work on ALL the supported arches with significantly less effort than what debian-installer will take to finish, I can't see why they'd care.
Yes, it makes it easy to _install_ but keep in mind how many architectures Debian supports: x86, Alpha, PowerPC (which covers OldWorld and NewWorld PowerMacs, CHRP and PReP systems, and Amiga PowerUP machines), MIPS, MIPS little-endian, ARM/StrongARM, HP PA-RISC, Motorola 680x0 (which covers 68k Macs, old Suns, Ataris, and Amigas), IA64, SPARC, and S390. How much effort has it already taken to make Anaconda work as a Debian installer for one architecture? (And note those are just the architectures that woody supports, I'm not even counting the experimental ones that are still only in sid.)
Oh yeah... cause you know, I'm reinstalling Debian all the time... wait. That's right, I'm NOT. I last installed Debian on my home desktop about 4 or 5 _years_ ago, and it's never once needed reinstalling. Though debian-installer, once it's done, should improve things significantly, how often do you really see the installer? Seriously, how often?
From what I understood, the Progeny installer went mostly ignored in Debian proper because it only worked on x86, and would take yet more retooling and fighting to make work on the _many_ other architectures that Debian supports (like the PowerBook Pismo I'm typing on now - a PowerPC). The Debian developers wanted a truly architecture-neutral, extensible installer framework, and nobody really provides that. I don't know how much work the YellowDog crew had to put in to make Anaconda work in YDL, but I have no doubts that it was significant. Unfortunately, if you want a product you can package and sell, the "must support ALL arches" requirement seems to be the first to get kicked to the curb.
AOL doesn't care about educating their customers. If they did that, their customers might realize how stupid, patronizing, and generally crap their service is, and go elsewhere. Keep 'em dumb, so you can keep taking their money - It's the Capitalist(TM)(C)(R) Way!
I think you need to go back to school, and take some basic math classes again. Or maybe that's the "new math" I keep hearing about? But then, what do I know...
SCSI is NOT serial. FireWire/IEEE1394/i.Link is serial. SCSI is not, and never has been. If you're going to talk about fair comparisons, then let's be fair, otherwise please stop talking.
How exactly would MS have learned jack about writing a 64-bit OS from NT/AXP? It was run in 32-bit compatibility mode on the Alpha. It was NEVER a 64-bit OS. Which I thought made it a really huge waste of a CPU - although it did run NT rather well (a former employer looked at an Alpha running NT for a project they were working on, and I sat in on the demo).
Ah, Alpha, we barely knew ye....
I'm pretty sure he means "free" as in both meanings - free as in beer, and as in speech. And yes, then public domain code is the "free-est". It costs you nothing, and you can take it and do ANYTHING you want with it - make your own operating system, make a product for sale, whatever. You don't have to abide by any license terms, because once it's in the public domain, it belongs to everyone, basically.
Remember that "free" has multiple meanings, and that a person can use it to mean more than one of those at a time.
BackOrifice was by the Cult of the Dead Cow, not l0pht. They wrote the l0phtcrack password cracker, and probably some other stuff. Don't assign credit (or blame) to the wrong people.
Yeah, I thought that was pretty weak too. You'd think they could manage to have someone do the reporting on this stuff who actually knows something about it. But I suppose, it makes the average Joe feel better if the reporter uses the same sort of terms he would use, because he doesn't feel challenged by the article. It's "We know you think this is bad, we think it's bad too. We're gonna talk about it in a way that makes you feel comfortable, and that makes you believe something is being done." That's today's journalism, kids.
Or you can use something like mPlayer or Xine (it's been ported to Win32, hasn't it?) that demultiplexes QuickTime on its own, and has its own codecs for SVQ1, SVQ3, AAC audio, and all the other formats that QuickTime and whatever other formats use.