|
2001/10/24 [Academia/GradSchool] UID:22813 Activity:moderate |
10/22 Let's say I didn't apply for NSF grant and I got into grad school. Can I apply again in grad school? \_ Yes, in your first year. http://www.nsf.gov/pubs/2001/nsf01146/nsf01146.htm#elig \_ deadline, btw, is Nov. 7th. \_ thx. Where can I get statistics so that I know what is my likelihood of getting it, and if I should apply? \_ http://www.orau.org/nsf/nsffaq.htm \_ btw they will not accept letters from the Career Center file; they need the prof to use a special NSF form. So forget it :( thanks for your help -original poster |
2001/10/24 [Computer/SW/OS/Solaris] UID:22814 Activity:very high |
10/22 Why are the rc?.d scripts on solaris HARD links? What is wrong with soft links?? How can you tell a hard link anyway? (why doesn't `file` specify a hard link the way it does to a soft link?). Please provide (URL containing) clue. \_ So if you move-aside/rename a file, the /etc/init.d script would still point to the right one? I'm sure this isnt the only reason, but it is one. I also think it's used with some binaries or library files. It's also good cuz you can update one and update all(like when overwriting the file with a newer version). Hard links may be faster too. Just guesses. I don't know the real reason. You should be able to detect hardlinks by looking at 'ls -l' output (a 2 or greater in 2nd column). \_ files that aren't softlinke are hardlinks. \_ You can delete the file the soft link points to and the link breaks. You can't break a hard link. \_ this still does not provide an explanation to why Solaris rc?.d scripts use hard links as opposed to soft links. \_ Give me a good reason why you would do it the other way. \_ maybe some psycho wants to put them on a different filesystem? Can't do hard links across fs's. \_ if you call this a good reason, you're a major fucking twink. \_ I never said it was a _good_ reason, just one possible reason that might make hard links not an option. \_ you obviously have no idea what a hard link is if you're trying to distinguish between that and a "real file" ... \_ i've got a hard link in my pocket. |
2001/10/24 [Computer/HW/Display] UID:22815 Activity:insanely high |
10/24 anybody know of a PCI video card with DVI output? \_ ATI Radeon \_ sorry, I meant for the PC. oh, and specific models would help... \_ http://www.tomshardware.com |
2001/10/24 [Computer/SW/OS/FreeBSD] UID:22816 Activity:high |
10/24 I'm trying to configure my ethernet interface in FreeBSD. ifconfig -a doesn't show an "ed0" device, but it does show a "sis0" that I think might be the equivalent to "ed0". What is sis0, and can I use it as my ethernet interface? \_ man sis \_ the "letters" part of the device name is just the chipset. I don't have an ed0 or an sis0. \_ ed0 would be a NE2000 card or a clone, sis0 would be, eg, a NetGear FA312 ether card. Odds are thats your ethernet. \_ you are probably not tall enough to run bsd. use linux. \_ And RIDE BIKE! too? I think "windows" is the answer if you want to be mean about it. ifconfig on linux and freebsd is essentially the same. How would linux help? \_ Linux ethernet cards are reliably named ethN. |
2001/10/24-25 [Computer/SW/Languages/Misc] UID:22817 Activity:moderate |
10/24 Using Check-point FW1, i have a 1 hour time-out which is very annoying to me. i would like to eliminate it or make it VERY high. Is this a really dumb thing to do, or is it only a slight risk for a lot of benefit? \_ You mean the TCP session idle timeout for NAT sessions or sessions for which FW1 is supposed ot keep state? I don't know about FW1 but IP Filter commonly used on *BSD has this timeout set to a really large value, something like 24 hours or so. The Linux 2.2 ipchains by default sets a very large timeout value too. I don't think this is big security risk but if you have a very large number of idle connections (e.g. thousands..), the software might run out of entries in its session state table. \_ earlier versions of IPFilter had a limit of 1024 state table entries--it's now been fixed. We have the same problem with FW-1; basically, it only comes into play when you have a connection that's idle for over 1 hour. The idiotic thing with this value is that it's a global setting, and can't be done application-specific (ObFW1SucksPlug). I don't believe there's not much of a security risk; part of the reason it was done was to prevent idle connection hijacking. You can get around it by running a simple script that echoes a single character on the target host every x minutes, assuming it's a login session. If it's not (something like database, whatever) just set it to a reasonable level (2-3 hours) if you really need to. -John |
2001/10/24-25 [Computer/SW/Mail] UID:22818 Activity:high |
10/24 Is there some way in pine to instruct it to leave my msgs on the server? Is there some terminal based mail client that will do this? \_ Uh, isn't this the default behavior? My mail's still in /var/mail. \_ d'oh... my mail is no longer in /var/mail. any other suggestions? \_ Do you have a file called mbox in your home directory? \_ ah... perfect, thanks. I did but I renamed it and now pine appears to be using /var/mail. Thanks. What is the explanation for this behavior? \_ Apparently, you had a file called mbox in your home directory. \_ pine is weird. \_ Explanation here: http://www.washington.edu/pine/pine-info/1998.11/msg00060.html |
2001/10/24-25 [Computer/SW/Apps/Media] UID:22819 Activity:moderate |
10/24 If I compress my MP3s using WinAmp and RealAudio, will Archos or others recognize its MP3 format? Iv'e noticed that RealAudio saves more info that WinAmp doesn't pick up. \_ try it! \_ good god, Real's MP3 encoder sucks ass. Why aren't you using LAME? \_ ANYTHING http://real.com makes sucks ass. That POS format/program is worse than WMP. Can't believe one of their VP is now a Senator |
2001/10/24-25 [Computer/SW/Security] UID:22820 Activity:moderate |
10/24 My sshd used to accept connections from machines without matching reverse lookup. Then, all of a sudden, today, it stopped. I changed resolv.conf to use a nameserver with made up ptr records and it works fine, but the question remains, What changed? There is no indication that sshd has been restarted since the machine was 60 days ago. This is on solaris, using (foolishly) F-secure sshd 2.0-2 which is also the same as it has been. (i did stop some services 2 days ago, but nothing that should effect this). \_ You've been hacked. \_ Maybe you have edited the hosts.allow or hosts.deny files and added or removed some rules on those? This would apply if sshd was compiled with support for tcp wrappers. \_ Or it actually _was_ doing reverse lookups and your DNS broke. |