Try our GCC! Completely ISO-C (ISO Crumpets) compatible. Has many flavors (languages)! However, may not be fully compatible with all CPUs (crumpet processing units).
From what I can tell, this proposes turning OSS into a mild form of adware - randomly displaying credits for developers. Not as bad as, say, "free gambling online" or "eliminate debts now!" in adware like KaZaA (not Lite), but still an annoyance.
The reason I write software is to fulfill a need that I have, not to get my name in the "most famous programmer" slot... I think many programmers are the same.
XPde might have a task manager that could actually kill a process.
Amen to that! I can't count the number of times a locked-up IE or Word window has refused to close on my WinXP workstation. At least under Win98 a repeated Ctrl-Alt-Del / Enter sequence was easy enough to do repeatedly and would eventually kill the process.
Finally, I can eat my electronics when I'm done with them! Even better than regular bread. Why can't we make more things out of this new, amazing "hibread"?
I think I should clarify here - I mean all non-public services. Obviously my public HTTP server doesn't require an username to view my website - same with non-relaying SMTP (ie, destination is local).
I tend to view each machine as an independent entity, untrusting of anything else by default. All access should require authentication - including things like NFS mounts (set up SSH tunneling, or a CIPE tunnel, or something) that work machinemachine (private key auth works well here).
This, combined with a minimal number of open ports (HTTP, SMTP, SSH, IMAP (I know, I should set up IMAP over SSL soon)) and good upkeep of patches, means that possible holes are limited, no matter where the attack comes from. This also has the added bonus that I don't have to worry as much about link-level security (like 802.11b - even though I still encrypt that) because I know that all services are encrypted and authenticated.
I've seen several download managers on Win32 that support continuation and multithreaded downloading, so I guess that means it does. Don't quote me on that though.
but they have to be encoded into text which can waste an extra 40% of bandwidth
Not really... maybe you're thinking of MIME-encoded email (in which attachments are encoded to plain text)? HTTP is perfectly capable of streaming binary content over the TCP connection - it's no slower or faster than any other TCP stream-based download protocol.
I thought these worked because static electricity charges the dust around it to stick to the collectors. Isn't static electricity bad for components?
Yes, static is bad for components. That's why you shouldn't stick your components inside your Ionic Breeze purifier. A grounded PC case? In that case, no danger at all.
The first thing I'd recommend is for everyone here who has an eBay account to give the user a negative review containing the body of the copyleft license and how (s)he's violating it.
Two things here - first, I think there's a limit on comment length (80 or 120 chars, maybe?). Also, I don't think you can leave anonymous comments - careful lest he retaliate and mar your account's feedback.
When I first got the motherboard I had problems with some type of power management (DPMS I think) and X - the monitor would turn off after a timeout, and the machine would lock hard. I fixed that one though.
I did read somewhere that APM was incompatible with SMP though. I've never tried to suspend/set idle timeouts/whatever with APM, so I don't know.
It's an ECS D6VAA, with dual PIII-800s (socket 370s). It was the cheapest SMP board I could find in Dec 2001:-) Curiously, it's perfectly stable under Linux w/o the nVidia drivers (uptime of 50+ days), and it's fine under WinXP Pro with the latest nVidia drivers. At one time I got it to be a _little_ more stable by disabling AGP in the nVidia driver options (Option "NvAgp" "0") but it still crashed.
Why would anyone not use the nVidia drivers under Linux
They were unstable for me when I still used them, version 2880 I think (a while ago), with XFree86 4.1.0 (Debian woody) and SMP system. Complete system lockups, usually after a few hours. Not fun when the machine was also my main server (before I got a separate one)...
They may be better now, I haven't used them for a while...
I had problems too when I ran FreeBSD on my desktop, through a 4-port KVM. I figured out that it had to do with what the psm(4) driver detected at boot time, and it wouldn't detect my wheel mouse as a wheel mouse through the KVM. It works fine in Linux though - because the Linux/dev/psaux doesn't interpret the data stream, whereas the FreeBSD driver does. Not sure of a fix - I just didn't use my KVM for my mouse:-)
Slashdot Mirror
GCC: GNU/Crumpets (Compatible)
Try our GCC! Completely ISO-C (ISO Crumpets) compatible. Has many flavors (languages)! However, may not be fully compatible with all CPUs (crumpet processing units).
from the gnu/crumpets dept.
I didn't know crumpets were POSIX compatible...
From what I can tell, this proposes turning OSS into a mild form of adware - randomly displaying credits for developers. Not as bad as, say, "free gambling online" or "eliminate debts now!" in adware like KaZaA (not Lite), but still an annoyance.
The reason I write software is to fulfill a need that I have, not to get my name in the "most famous programmer" slot... I think many programmers are the same.
XPde might have a task manager that could actually kill a process.
Amen to that! I can't count the number of times a locked-up IE or Word window has refused to close on my WinXP workstation. At least under Win98 a repeated Ctrl-Alt-Del / Enter sequence was easy enough to do repeatedly and would eventually kill the process.
Finally, I can eat my electronics when I'm done with them! Even better than regular bread. Why can't we make more things out of this new, amazing "hibread"?
All access should require authentication
I think I should clarify here - I mean all non-public services. Obviously my public HTTP server doesn't require an username to view my website - same with non-relaying SMTP (ie, destination is local).
I tend to view each machine as an independent entity, untrusting of anything else by default. All access should require authentication - including things like NFS mounts (set up SSH tunneling, or a CIPE tunnel, or something) that work machinemachine (private key auth works well here).
This, combined with a minimal number of open ports (HTTP, SMTP, SSH, IMAP (I know, I should set up IMAP over SSL soon)) and good upkeep of patches, means that possible holes are limited, no matter where the attack comes from. This also has the added bonus that I don't have to worry as much about link-level security (like 802.11b - even though I still encrypt that) because I know that all services are encrypted and authenticated.
Does HTTP support file continuation now?
I've seen several download managers on Win32 that support continuation and multithreaded downloading, so I guess that means it does. Don't quote me on that though.
but they have to be encoded into text which can waste an extra 40% of bandwidth
Not really... maybe you're thinking of MIME-encoded email (in which attachments are encoded to plain text)? HTTP is perfectly capable of streaming binary content over the TCP connection - it's no slower or faster than any other TCP stream-based download protocol.
I thought these worked because static electricity charges the dust around it to stick to the collectors. Isn't static electricity bad for components?
Yes, static is bad for components. That's why you shouldn't stick your components inside your Ionic Breeze purifier. A grounded PC case? In that case, no danger at all.
The first thing I'd recommend is for everyone here who has an eBay account to give the user a negative review containing the body of the copyleft license and how (s)he's violating it.
Two things here - first, I think there's a limit on comment length (80 or 120 chars, maybe?). Also, I don't think you can leave anonymous comments - careful lest he retaliate and mar your account's feedback.
Standard T1 is a 1.5kbps connection,
You mean 1.5Mbps, right? 1.5kbps is 1.5 kilobits per second, about 1/40 of a 56k dialup.
and a SDSL T1 is 1.5kps up and then same speed down. i.e. 3kbps combined
actually, 1.5Mbps both ways _is_ a T1. Both SDSL and T1 lines are full-duplex, so your "SDSL T1" is the same speed as a T1.
Assume an average 3-minute 128kbps MP3 - about 3 MB. 3 MB * 652000 = 1956000 MB. About 2 TERABYTES.
Did this guy have a 20-disk RAID in his box, or am I missing something?
When I first got the motherboard I had problems with some type of power management (DPMS I think) and X - the monitor would turn off after a timeout, and the machine would lock hard. I fixed that one though.
I did read somewhere that APM was incompatible with SMP though. I've never tried to suspend/set idle timeouts/whatever with APM, so I don't know.
It's an ECS D6VAA, with dual PIII-800s (socket 370s). It was the cheapest SMP board I could find in Dec 2001 :-) Curiously, it's perfectly stable under Linux w/o the nVidia drivers (uptime of 50+ days), and it's fine under WinXP Pro with the latest nVidia drivers. At one time I got it to be a _little_ more stable by disabling AGP in the nVidia driver options (Option "NvAgp" "0") but it still crashed.
Why would anyone not use the nVidia drivers under Linux
They were unstable for me when I still used them, version 2880 I think (a while ago), with XFree86 4.1.0 (Debian woody) and SMP system. Complete system lockups, usually after a few hours. Not fun when the machine was also my main server (before I got a separate one)...
They may be better now, I haven't used them for a while...
I had problems too when I ran FreeBSD on my desktop, through a 4-port KVM. I figured out that it had to do with what the psm(4) driver detected at boot time, and it wouldn't detect my wheel mouse as a wheel mouse through the KVM. It works fine in Linux though - because the Linux /dev/psaux doesn't interpret the data stream, whereas the FreeBSD driver does. Not sure of a fix - I just didn't use my KVM for my mouse :-)