320 new of 384 responses total.
Oh, in case you haven't heard it mentioned, Grex staff does not advise storing any irreplacable data on Grex.
Regardless of all discussion issues, OUTSTANDING JOB!!! for merge-restoring old e-mail with new. Re: e-mail inboxes, I guess that forewarned is forearmed (forlegged? ;-). This should probably be pointed out prominently in newuser and elsewhere, if not already. (i.e. GREX MAKES NO GUARANTEES ABOUT STABLE LONGEVITY OF E-MAIL or some such)
Okay, earlier this evening, we were able to restore all the mail from the old disk and preprend to any mail you may have received since the disk trouble. So while we make no guarantees, we certainly do make the effort. :) Thanks again to Jan, who was up most of the night Sunday. I got the easy shift tonight restoring the mail spool. It's only a little after 1am and I'm home already.
thanks kip
Once again I give a well done to the fixers and keepers of grex!
Well done!
Thank you, Jan and Kip for pulling night shift to keep mail flowing. Very generous of you both.
Um. WOW. Kip, that is amazing, besides being an awful lot of work. Thank you very much.
Staff on Grex is awesome! Thanks folks for carrying enough about us to put in that kind of effort.
Indeed, I'm very impressed that Kip fixed the disk, and that Jan and Kip spent so much time on this problem. THanks very much!
Re resp:52: Neither your mail spool nor your home directory are guaranteed to be backed up on a regular basis. If you care about the contents, copy them to your own computer. I'm lax about doing that too, but when I do I usually just tar everything up and then FTP it to my home computer. Re resp:67: Many thanks, kip, for resurrecting the disk. :>
Yeah, staff here are awesome. Staying up all night for email! Wow. That is very much appreciated.
This response has been erased.
Do I understand correctly that it is only the INBOX that is not backed up, but that other e-mail folders in ~/mail are?
This response has been erased.
Yes, but is it backed-up?
This response has been erased.
re #78: On most Unix systems, the basic storage unit for backups is
the disk partition. When staff backs up the system, they probably use
the Unix dump command to back up individual disk partitions.
Your home directory, which contains ~/mail, is on one partition (e.g. /a)
and incoming mail (your inbox) is on another (/var/spool/mail)
To complicate matters further, the most usual unit of catastrophic disk
failure is an entire hard drive, which can contain several partitions.
Grex has six hard drives mounted at the current time, containing the
following filesystems:
sd0: /
/usr
/bbs
/x
/oldbbs
sd2: /rootbak
/suidbin
/oldvar
/tmp
/s
sd4: /c
sd6: /var
/usr/local
sd7: /d
/var/spool/mail
sd11: /a
I'm assuming this partitioning scheme (which doesn't make a whole lot
of sense) grew up as the result of historical necessity and "if it ain't
broke, don't fix it.." I likewise assume that the partitioning on
nextgrex will be a little more organized. In any case, that's a topic
for another conversation..
re #80: it's best to presume that backups are not predictable or frequent, but occasional. No matter what the backup policy on any system, I *always* advise people to do their own backups of information they consider important. (and I don't just mean this policy to apply to volunteer-run systems like Grex. take charge of your own data everywhere if it's important to you.)
Getting past the technicalities, have I lost all the love-emails from my stalker?
Question, is Veritas/Vinum available for OpenBSD, and are we gfoing to use it?
This response has been erased.
are we still using the (repaired) dying disk for email or has it been replaced? Was it really dying or just in need of fixing. It is lovely to get back all my mail through Jan 1. I should go through it and copy down the parts I want to save, on paper. I wrote someone who probably sent me mail since then to try again.
Sounds like the electronics failed, not the disk, and that the electronics have been replaced and the disk is back in service.
The failed disk was replaced. After repair, it was mounted long enough to copy the data from it. It does not seem to be mounted now.
This response has been erased.
Re resp:77: No one is stopping you from copying your email to somewhere safe. Your mailbox is simply a plain text file. You can easily transfer it to your home computer if you want to.
RE 84> Oh goodie. Forward them so I can spend valuable time deleting them one at a time.
re 86 Isn't the SENT folder a dumb Outlook Express idea?
Certainly Outlook and I assume Outlook Express *have* a Sent Items folder. So do other e-mail systems (Juno, hotmail, yahoo). I don't consider it a dumb idea. At least with Outlook, you have the ability to specify (and set as an option) whether or not sent messages get saved in the Sent Items folder.
Whoops, I was thinking of the outbox. Mybad.
lol
This response has been erased.
Re resp:95: An 'outbox' (or similar structure) is necessary for any mail client that has an offline mode. Where else would you queue up sent messages that can't be transmitted yet? Re resp:97: I think a better argument, in Grex's case, is there's no convenient way to back up the mail spool often enough to matter. People aren't really *supposed* to let old mail linger there anyway, so any data in the mail spool ought to be transient. It'd be like backing up /tmp.
Re 97> Eh, no biggie. Just this person who sent me long emails and said I was second on his list to marry (yeah, I was flattered :P). Petered off when he said he couldn't continue corresponding if I didn't answer. I'm not complaining. :)
re 98 Ahh, I see.
This response has been erased.
re #101: > The legal argument was the best I heard > for keeping it seperate, but like I said, I don't buy it. You might really be surprised how time consuming it can be satisfying a subpoena, summons to produce records, or other type of court process. > The argument then shifts to one of space and quotas, but on > nextgrex we'll be using normal UFS quotas anyway, which would negate > that problem for the most part. Then the argument shifts to not > wanting to use Unix quotas for mail, since the mail still has to > be transfered to grex and bounced; it can't just be rejected during > the SMTP transaction. However, looking at the logs, one sees that > on any given day, only 5% of the mail grex deals with is rejected > due to the target mailbox being over quota. It doesn't strike me > as worth it to augment the SMTP server with code to deal with quotas > just to save on 5% of the mail we go through in one day (particularly > given the amount of effort that goes into maintaining the result!). Among its many other virtues, the postfix MTA has a very simple-to-use feature to limit mailbox size (it's called something painfully obvious like inbox_size_limit or something like that..) Are we sticking with sendmail on NextGrex or could we move to something just as capable but far easier to manage?
This response has been erased.
No subpoena ever written can get blood from a stone: You cannot produce what you do not have. If you don't back up the mail spool, you don't have back ups of the mail spool to produce.
Unless the mail software really does enforce mail quotas, we really don't want mail files in the users' space. Because someone sends me spam, I've exceeded my disk quota?
I'm not sure, but I think that /newdum is the script that is used to do backups on Grex. It does not appear to back up the mail spool. It also does not appear to back up /s, which is where my home directory is.
This response has been erased.
We haven't backed up the mail spool in many years.
Can this file please be mae world-readable? -rw------- 1 cfadm software 87783 Oct 29 2000 /bbs/errorlog1.gz
mae==made
I have not received any mail except for 2 spams since Jan 1. I wrote three friends asking them to send mail to keesan@grex.org. I sent myself two mails just now to grex.org and cyberspace.org (but not grex.cyberspace.org). They have not arrived. Are other people getting non-spam mail from anywhere? I usually get 5-10 emails a day from real people.
I've been receiving mail from correspondents elsewhere normally all week.
Yeah. /bbs/errorlog's world readable. Why shouldn't the previous version be the same?
I've been getting all kinds of mail myself. No one I usually hear from has failed to write, and no one else has reported problems.
I still have not received the two mails I sent myself yesterday, or replies from three friends, one of whom phoned and said all his mails to me were bouncing. I will try writing @grex.cyberspace.org. So far no mail since Jan 1 except 2 spams. I normally use @grex.org. Mike and Colleen, to what address are your mails arriving? grex, cyberspace, or grex.cyberspace?
What is the error message in the rejection notices? Knowing that would help diagnose the problem. If you like, ask him to send the rejections to you at another address, then you can paste in the text from the rejection notice.
Once I had set up my procmail wrongly , and it was rejecting all messages. Just thought you might want to check that too.
Re 113 re 109 (talking to himself?): Why should the error log be world readable in the first place?
I've been getting mail normally from a couple mailing lists I'm on, ebay, and any number of spammers. I use "cyberspace.org" as the host part of my address. Maybe you should temporarily turn off procmail and make sure it's not deleting your mail for you.
At the risk of indulging in an exercise in futility, I'd like to not ethat I keep getting items coming up which allegedly have new responses, but where new responses are no-whee to be seen, and all i get is the title of the item and the author.
This response has been erased.
Thanks for the explanation, but that might need a little more explaining! ;-)
Valerie wrote a script before she left that allows one to scribble and expurgate one's responses en masse. Apparently that causes the items where one's done so to come up new.
Ah, so that's what's happening. In that case, apologies to you, twila, for the harsh tone in that other item.
This response has been erased.
Well, well.
I should say, too, that I didn't plan on doing this in the current agora. I am not leaving Grex, but was upset by all of the hoopla this week.
Not surprising. :-(
This happens because when someone changes an item, it is considered as something new. Valerie ran a program that erased her responses to a number of items and that counts as a change, thus all those items come up as having new responses.
Ah. thanks.
High load averages! ast pid: 17012; load averages: 21.74, 20.82, 17.61 11:18:15 298 processes: 269 sleeping, 28 running, 1 stopped CPU states: % user, % nice, % system, % idle, % spin Memory: 234M available, 232M in use, 2544K free, 7076K locked PID USERNAME PRI NICE SIZE RES STATE TIME WCPU CPU COMMAND 17006 nobody 62 0 340K 548K run/0 0:01 28.95% 6.64% backtalk 17010 naftee 59 0 3364K 2856K run/1 0:01 31.14% 5.86% top 256 root 78 0 12K 8K run/3 109:54 11.31% 5.86% update 17013 root 1 0 76K 352K sleep 0:02 35.08% 5.08% ftpd
You can do 'fixseen' for all those conferences where Valerie deleted her responses. I will check my procmail filter to see if I did something stupid, but normally I would expect procmail to just send anything it caught to the bulk folder, not delete it. I will ask for a copy of the bounced mails. And try sending myself mail to grex from somewhere else.
re #115: my mail is mostly addressed to "mcnally@cyberspace.org" I sent a test message to your account yesterday. Presumably you haven't received it, but I got no bounce message. I'd highly recommend turning off your procmail filter for a while and testing whether you get mail without it..
all incoming mail is ignoring my .procmailrc file and sending spam into my /var/spool/e/p/eprom directory....please fix procmail
Why are ancient item headings and responses appearing in many other conferences without new responses being entered? Here is a brief selection from the consumer cf: > Item 90: Water Filters > Entered by Valerie Mates (valerie) on Tue, Nov 4, 1997 (23:27): > <expurgated & scribbled> > > 1 new of 22 responses total. > #22 of 22: by Rane Curl (rcurl) on Sat, Feb 28, 1998 (13:36): > They add chlorine *and* ammonia. Chloramine itself is a highly explosive > liquid. The chloramine is formed in solution (no, its not explosive in > solution). > Respond or pass? > Item 92: Advertising and public services > Entered by C. Keesan (keesan) on Thu, Mar 5, 1998 (17:12): > Respond or pass? > Item 98: Looking for a plumber > Entered by Valerie Mates (valerie) on Tue, Oct 27, 1998 (22:10): > <expurgated & scribbled> > Respond or pass?
It's the bitch's fault.
It has caused me a considerable waste of time to run fixseen in a large number of cfs that I watch.
This response has been erased.
Rane: Valerie decided to leave Grex, and to delete all the responses she's ever made on Grex. So every item she ever responded to shows up as new.
Great... (that's a contradiction, so it is false but humerous, although the extra dots are required to convey the sense in which it is said).
(except that it's not really funny.)
re139
*sigh*
Re 140: so it relates to your elbow, Rane?
re 141 "humerous" is not funny either.
...some people sure don't have senses of humer...
Backtalk doesn't allow a fw to delete posts from items, even if the post was by the fw himself. (FWs can delete complete items though)
How do I turn off procmail? Today I got two spams address to @grex.org but nothing from real people including McNally who sent me the text mail.
One way is just to rename the .forward file in your home directory. For instance, type "!mv .forward .forward.old".
I think you can turn off procmail by removing that line in .forward which refers to procmail or simple solution is to just rename .forward to a different name for some time. like mv .forward bkp.forward.
148 slipped in.
I put a # in front of the line in .forward to disable it - is that right? I since today I got two more spame to @grex.org I just sent myself one more test mail to keesan@grex.org. The friend who got back bounced mail said it was bounced on Jan 3 with a 'host name lookup failure for cyberspace.org'. When was our domain name problem fixed? Perhaps the spammers have some way to get through that real people don't know about? I have not had any bounced mail when I mailed to my grex account from another account in the last two days.
I *think* sendmail respects # as a comment character in .forward, but I have encountered other mailers in the past that didn't. I think it is better to move the .forward to a different file name if you aren't using it right now. I've observed no domain problems for the last few days now, not even in asia (for some reason, dns on servers I frequent in Japan and Hong Kong took a bit longer than the US based ones I use to notice the dns corrections).
After I put # in front of the line in .forward, and also after I renamed it .forward.bak, I got both my test mails and also four of the spams that would normally have gone into the bulk or other spam folders. Anyone want to take a look at my .forward.bak (or .forward) file to see what stupid thing I might have done to it last time I added a few filters? Thanks all. But four spams slipped by whatever filter was causing the problem. I have saved two of them in case they are helpful. Do I need to make my .forward file readable somehow (chmod +x .forward)?
I got my grex mail fixed just in time. Today I got a bounced mail at my temporary address which I sent to someone using a service that uses spamcop's blackhole list. I would greatly appreciate someone helping me to fix my spam filter here so my box does not fill up with trash.
Sindi - that error message your friend got sounds like it was during the period when a lot of people were reporting domain lookup problems on both grex.org and cyberspace.org. I still don't know what caused that - it was only cyberspace.org that changed registrars.
named on grex was running but not, apparently, working. I restarted just before the machine went down for the mail problem, IIRC. Apparently, we dropped out of puck.nether.net's config files, while named wasn't responding.
IRC
I am getting mail okay now and have started a new .procmailrc file and am sending myself a test mail every time I add a filter but I would still appreciate people telling me what I did wrong with .procmailrc.old so I won't do it again. How do I filter (with one filter): viagra, v-i-a-g-r-a, v!iagra, v1agra without filtering all my mail to a bulk folder?
I'm not certain, but I believe it was just a manner of poor luck and timing that the DNS issues occured about the same time and were the real cause of your mail issues.
The DNS problems were in addition to my mail filter problems. I started getting mail yesterday a minute after I removed my filter.
Hmmm, okay then. May I suggest you add something like this to your filter file near the top: LOGFILE=/a/k/e/keesan/mail/procmail.log Or whereever you would like to point that file. That will give you a file that records the From:, date, Subject: and eventual file or destination for every email procmail handles. Very useful for tracking these types of issues down.
Thanks, but I think I would fill up my disk quota quickly if I logged all my spams. Jeremy found my problem - I added five filters without a * before ^Subject. I restored my old .procmailrc and will test that i can send myself mail every time I change it.
That's true, procmail averages about 150 characters per log entry for me at my home domain and last month I had a little over 16,000 emails, so the whole month's procmail log was just under 2.5 million bytes. That would chew up the quota here pretty fast.
*if* you left it on..
I have a very long filter set up and all four of the spams that arrived in the last two hours were of the XOCPT she swims bananas type with five lines of nonsense words before the HTML. I have not found any way to filter them - any ideas?
Dialing in to 761-3000 several times in the past few days, i've gotten NO CARRIER disconnects as soon as the modems are done with their training tones. Dialing in to 761-3451 gets me in normally when this happens.
This response has been erased.
Someone hogged up a lot of space. I'm looking into it, but I don't have a lot of time right now.
Hey, punk, that doesn't help Grex.
tail: /var/log/messages: Permission denied Fix this please.
That's intentional.
I'll bite. What is in /var/log/messages that you want to see?
ACN"T YOU SEE I"M COLOUREd??
re 172 Someone killed my login processes last night while I was in party. I want to know who did it.
last pid: 15042; load averages: 10.59, 15.37, 14.54
13:35:16
220 processes: 213 sleeping, 3 running, 4 zombie
CPU states: 28.2% user, 0.0% nice, 18.7% system, 31.6% idle, 21.5% spin
Memory: 233M available, 151M in use, 82M free, 7992K locked
PID USERNAME PRI NICE SIZE RES STATE TIME WCPU CPU COMMAND
5058 root 1 0 17M 16M sleep 325:47 11.34% 11.33% named
15031 k0i 1 0 924K 464K sleep 0:00 17.03% 3.91% sendmail
15029 k0i 1 0 924K 464K sleep 0:00 15.33% 3.52% sendmail
15020 k0i 1 0 924K 464K sleep 0:00 10.22% 3.12% sendmail
15024 k0i 1 0 924K 464K sleep 0:00 8.95% 2.73% sendmail
15022 k0i 1 0 924K 464K sleep 0:00 8.95% 2.73% sendmail
15018 k0i 1 0 924K 464K sleep 0:00 8.02% 2.73% sendmail
256 root 15 0 12K 8K sleep 591:38 3.73% 2.34% update
15027 krj 3 0 364K 472K sleep 0:00 7.27% 1.95% vi
15041 root 40 0 904K 460K run/1 0:00 30.77% 1.56% sendmail
14995 k0i 1 0 924K 464K sleep 0:00 3.02% 1.56% sendmail
14990 k0i 1 0 924K 464K sleep 0:00 2.76% 1.56% sendmail
15034 root 1 0 904K 444K sleep 0:00 8.30% 1.56% sendmail
15009 k0i 1 0 924K 464K sleep 0:00 2.69% 1.17% sendmail
Looks like Valerie's back to her old tricks.
This response has been erased.
re 174 /var/log/messages doesn't contain information that would say who killed a login process. Or at least it doesn't to my eyes. I'll defer to other staff with more experience.
This response has been erased.
re #174: I recently had an evening where I kept getting logged out abruptly, several times while in party. Did you just leap to the conclusion that you were being persecuted by a vengeful root user? My assumption was that it was likely that network problems were causing my session to drop out.. I'm not sure how you'd differentiate between the two scenarios in a typical login session, so perhaps you should at least consider other causes..
I've been having this, too - after a while, if i don't type anything, my session just gets dropped. Is that the problem you've ben having, Mike?
This response has been erased.
I already have an item about this in coop. maybe oldcoop.
re 179 actually that would be /var/log/sulog.
I get Pine sessions killed with various signals from time to time. I don't know why. I'd been assuming some hardware glitch.
I for one am glad that the messages file is closed to all but roots.
re 177 All the crappy e-mail bombers are written with Italian instructions. re 180 Both ryan and jimj were in party at the same time; it's more than enough to get worried. re 184 It is /var/log/messages on FreeBSD, maybe it's different on SunOS. re 186 I'm glad you're just a silly old goose.
It is different here. Use of the su command is logged in /var/log/sulog, not /var/log/messages
Naftee you are paranoid. And now for my paranoid issue. I logged in to find that I supposedly sent junk crap mail to people and it was bounced back. I find this more than just strange but odd since I never send mail on grex anymore and all my mail from grex should be forwarding which is forwarding again. so... there, and who ever the unethical jerk who did this, f- you! f- you and your large a-hole!
Jiffer quit being mean.
It's probably a virus, choosing an e-mail address from the infected machine's address book.
Occasionally, if I am disconnected from grex for being idle to long but log in just after having been disconnected, I will get forcibly signed-off again. Unfortunately, I don't have enough debugging evidence to really nail down the cause but I assume there may be a bug in the idle login killer or robocop that causes it to mistake my fresh loging for a valid target process to kill.
s/loging/login session
RE#189 -- Yep, your address was harvested from someone elses machine that has been infected with a virus. It looks like you sent mail, but you really didn't. It's a huge pain in the ass, but there is nothing you can do about it.
Drats, that is a bit embarressing. Naftee I am always mean.
I frequently get spam with the "From:" address forged to look like it came from somebody I know. As Chris says, the most likely explanation is that both addresses were lifted by a virus from the address book on some third party's infected (Windows) machine.
Is this the proper item in which to report an e-mail "difficulty"?
Probably as good as any. What's up?
For I'm not sure how long now an e-quaintance who receives my grex e-mail OK cannot get mail delivered [back] to grex. After a few days, he gets a bounce message that looks like this: From: "Mail Delivery System" <MAILER-DAEMON@out1.mx.nwbl.wi.voyager.net> To: <auser@voyager.net> Sent: Monday, December 29, 2003 11:34 AM Subject: Undelivered Mail Returned to Sender This is the Postfix program at host out1.mx.nwbl.wi.voyager.net. I'm sorry to have to inform you that the message returned below could not be delivered to one or more destinations. For further assistance, please send mail to <postmaster> If you do so, please include this problem report. You can delete your own text from the message returned below. The Postfix program <auser@cyberspace.org>: connect to grex.cyberspace.org[216.93.104.34]: read timeout I guess I should contact the voyager.net postmaster (my e-quaintance isn't computer savvy), but before I do, does the grex "side" have anything to comment on about this?
My _guess_ is that they've set a too-short timeout on the voyager end.
Hmm...
Jan 18 02:08:37 grex sendmail[19179]: NOQUEUE: SYSERR: putoutmsg
(out0.mx.nwbl.wi.voyager.net): error on output channel
sending "220 ESMTP spoken here": Connection refused by
out0.mx.nwbl.wi.voyager.net
The connection is supposed to be something like:
} telnet grex.cyberspace.org 25
} Trying 216.93.104.34...
} Connected to grex.cyberspace.org.
} Escape character is '^]'.
} 220-grex.cyberspace.org Sendmail 8.6.13/8.6.12 ready at Mon, 19 Jan 2004
13:57:13 -0500
} 220 ESMTP spoken here
Apparently, that second line never gets back to out?.mx.nwbl.wi.voyager.net,
and then things time-out/break.
Does that mean that the grex side is "failing" in some way, or is it still something on the voyager.net side. BTW, I'm very sure that my e-quaintance would have mentioned if this problem ocurred with other systems he tries to send e-mail to...
I think it's on the voyager side: grex is trying to send something, but voyager doesn't accept it.
This response has been erased.
why is procmail acting funny again?
(Voyager is rejecting the message with a "time out" error, but grex is reporting "connection refused.")
re 200 IS THAT THE MAILLOG?
re#204 nevermind...i'm stupid...I turned on the verbose logging and figured it out.
re 123 & 139 ... "valerie left grex" ????????? surely you jest!
Heh, I guess you havent logged on a while.
Maybe he doesn't read coop.
/pops some popcorn and puts "lost weekend" into the vcr
Use PicoSPAN, not backtalk, when mass-forgetting items.
OKay, one simple and possibly pwertinent question. Is it possible to do a
Fixseen on an individual Item instead of a whole ocnference? Seems somebody
has offered us the benefit of Greek classics, and sooner or later, somebody
will come along and do it as some sort of implicit censorship to ANY item,
or author of an item they might dislike.
No, I do not know Scripting, much less C. maybe the idea is of use,
in any case.
"fixseen" or "seen" defaults to all items in a conference if no arguments
are given.
If you want to "fix" a single item, just add that item's number as an
argument. For example:
fix 1
seen 1
fixseen 1
all do the same thing, and mark all the responses in item 1 as
already read.
Can you fixseen 63-end?
Yes, you can, though I think if you want to just deal with agora and the repeated items, fixseen 100-171 would do the trick.
This response has been erased.
By the way, there's still lots of space left on the device: /dev/sd0e 706783 446639 189466 70% /bbs
re 216 'r 100-171 pass' also works. Or just 'forget 100-171'. That too.
Read since jan 23 pass in the conferences that I check regularly generated a 2+ megabyte file. No way I can spare the connect time to download it, let alone the real-time to read it. This is more than an order of magnitude greater than the usual traffic. What's going on here?
What are you talking about?
I just did a fixseen on agora. However it took many minutes to "fix" all the damage.
It took a couple of seconds to run the command that I posted.
re 209-210 ...uhhhh, correct. family situations have absorbed grex-time, much to my dismay. it does seem that growing pains (on grex) have led to privacy controvesies. a looooong, looong,looong time ago a person whom i quicky learned to abhor did, however, provide one valuable chunk of advice; if you are not willing to see your text on the front page of the new york times tomorrow, dont' type it tonight. that protects a lot of people, actually, but confounds the misiion of both grex and the m-b0x. catharsis and conversation initially reserved to a small group of 'known' intellects having been spawned to the vulgate masses creates a nearly untenable conflict for the existence of the forum. constructing a higher civilizatin necesitates brick walls somewhere. a friend, a real friend, sent me some confrernce pointers which reinforce al of the above. tankxx, (xxx). ---more later ---
> Grex locked up because /usr/local filled up. What is that [partition] used for, and how did it fill up?
It's used for various lcoal (i.e., machine/host specific) configurations and programs. It filled up with core files had been saving, but we didn't really need. There are also some log files on that partition, I think. It was probably writing to one of them that immediately filled the disk partition. Removing the core files solved the problem.
what is it about the difference between logging in by regular telnet versus ssh that allows mail by telnet but says thre is no mail by ssh. i can invoke the -f flag and my mail file by ssh though. curious behaviour methinkxx.
If your .login file does not include the line
source /usr/local/etc/global.login
you can add the lines
#ssh is setting $MAIL wrong, so let's re-set $MAIL here: -vm 1/30/01
setenv MAIL `/usr/local/bin/maildir $USER`
Re resp:227: sshd does not set the MAIL environment variable properly. Resp:228 is correct if your shell is csh or tcsh, which is what tsty seems to be using. For those of you who use bash, you probably want this in your .profile: MAIL=`/usr/local/bin/maildir $USER` ; export MAIL
(Note that those are backticks, not apostrophes.)
hmmm, will try teh #228 fix .. thankxx to gelinas AND -vm 1/30/01
well done!! for ssh and tcsh.
What might it mean if a foreign host trying to ftp into grex were shown the following, after the correct password were accepted: 230->>>NO PSYBNC<<< >>>NO EGGDROP<<< NO NO NO!!!!! Won't run here! 421 Service not available, remote server has closed connection Login failed. Note that the same foreign host has no problem telnetting into grex - that is happening right now.
You're not trying to ftp in as 'anonymous', are you?
No - I enter my grex username & password. It even knows if I use an incorrect password. Of course, all I "know" about what the foreign host is doing w.r.t. ftp is what I see on the screen. Since I assume that inbound ftp to grex works in general, I suspect it must be something about how the foreign host is connecting etc. to grex. What does the 421 condition truly mean?
Hmm...it worked for me, but Grex's FTP daemon seems really sluggish. I wonder if your connection is timing out somehow.
Could be. I just tried connecting, twice. First time it timed out with a 421 error. Second time, I made it in and got to the ftp> prompt.
I just logged in and found 72 newresponse items in agora, but they are mostly false - what shows are just item headings. Is there another scribbler at work now?
Tod appears to have decided to join the ranks of the unrecorded.
(for the 3rd time - moron)
Note that my 421 was immediate - no timeout associated.
Just as disruptive as farting in every item.
Just like your mom.
*mommyfarts*
FYI, I still get 421 every time I ftp-connect to grex from the only foreign host from which I would do that.
[Jura:~] gelinas% ftp grex.cyberspace.org. Connected to grex.cyberspace.org. 421 Service not available, remote server timed out. Connection closed ftp> ftp> pwd Not connected. ftp> open grex.cyberspace.org. Connected to grex.cyberspace.org. 421 Service not available, remote server timed out. Connection closed ftp>
Why are you on staff if you don't know how to fix that?
Just now, it seems I am unable to ftp-connect (I waited for a while with no response, then broke out via Ctrl-C). In the past, I successfully ftp-connected with grex, but after logging in, I was immediately given a 421, but immediately, with no mention of a timeout.
I just ftp'd to my Grex account using the client Fetch. It took quite a long time to get through, but it did eventually.
ssh, tonight was sure snarfed ... didn't even get a login pormpt.
Looks like the problem with FTP might have been caused by a problem with named, which has now been restarted.
ttys tied up .... unnamed ttypa 10:56am 6 6:02 33 /bin/ld -e start -dc -dp -o conf kjutas ttypb 10:45am 17 6 5 lynx www.text.plusgsm.pl markis ttypc 10:56am 8 7 1 -bash crapula ttypf 11:00am 8 3 /usr/local/grex-scripts/.inet_re crapula ttyq3 10:55am 6 4 /usr/local/grex-scripts/.inet_re crapula ttyq8 10:51am 9 5 /usr/local/grex-scripts/.inet_re crapula ttyqa 10:54am 6 4 /usr/local/grex-scripts/.inet_re crapula ttyr5 10:53am 8 4 /usr/local/grex-scripts/.inet_re crapula ttyr8 10:58am 5 2 /usr/local/grex-scripts/.inet_re crapula ttyre 10:54am 5 4 /usr/local/grex-scripts/.inet_re crapula ttyrf 10:59am 7 3 /usr/local/grex-scripts/.inet_re crapula ttys7 10:52am 7 4 /usr/local/grex-scripts/.inet_re crapula ttyt3 11:04am 7 2 /usr/local/grex-scripts/.inet_re chetansk ttyt6 10:42am 33 28 lynx pronstar ttyt8 9:12am 17 8 /usr/local/bin/party_ crapula ttyu5 10:56am 10 4 /usr/local/grex-scripts/.inet_re crapula ttyu8 11:05am 7 2 -bash durg ttyu9 9:55am 14 16 11 lynx crapula ttyua 11:01am 9 4 /usr/local/grex-scripts/.inet_re newuser ttyp0 10Feb04 8days 5:18 3:56 pine newuser ttyp5 8:59am 41 3 w >
crapulatastique!
GreX is slower than a peanut-butter river and there are high load_averages.
IT"S RUNNLING LIKE MOLASSES!
This response has been erased.
IT WASN'T ME SOMEONE LOCKED MY ACCOUNT I'M INNOCENT WAS'NT EVEN LOGGED ON !last -2 naftee
Re: #251: Seeing that, I tried to ftp again from my remote host. I got just a little further, but after grex displayed this: 230->>>NO PSYBNC<<< >>>NO EGGDROP<<< NO NO NO!!!!! Won't run here! 230- 230-You must keep file transfers below 100 kilobytes >>>per day<<<. 230- 230-DO NOT BRING THESE FILES TO THIS SERVER: 230-NO eggdrop, NO psybnc or any bouncer, NO IRC servers/clients/bots, NO emech, 230-NO MUDs, NO mp3s, NO jpegs, NO gifs, NO audio files, NO PTLink, NO BitchX, 230-NO Windows or Linux programs (won't run here - Grex runs SunOS), NO Unreal, it "froze" - I had to use Ctrl-C to break out. Now I'm back to getting the immediate 421 after login.
NO NO NO!!!
I don't know what the problem is, albaugh; I just tried it with no problem.
The problem is, you fuckers stole my account.
What fraction of incoming e-mail on grex is spam? It must be a very large fraction, given what I observe on a couple of accounts. Is this a good utilization of server capacity?
This response has been erased.
Isn't this a concern to Grex staff/board? It would seem that Grex is providing "service" primarily to spammers rather than to users. I don't see indications that this bothers anyone very much (unless there is something in coop about it).
Marcus has added anti-spam features to sendmail. When he gets time, or at other impetus not clear to me, he looks at the mail forwarded to uce and tries to "sharpen the teeth" of his spam-filter. There has been some discussion but, so far as I know, no progress on the mail system for the new grex machine.
I was forwarding all of this spam I've been getting to uce, but I saw no change (well, actually a great increase) in the amount of spam received, so I quit. It is quite time consuming to forward the spam, and with no evident improvement, I just delete it now. If it is really true that 90% of the e-mail received here is spam, I would suggest some drastic action, like everyone changing their e-mail address, or do it wholesale by (say) adding "6" to the end of every account. But then everyone would have to inform their correspondents.... I HATE SPAM.
I h8 u
I h8 fags.
The feeling is most likely mutual. Or it would be, if cigarettes had feelings.
Do you have terrifically bad reading comprehension, or are you purposely misinterpreting the fact that "fag" in that context undeniably refers to homosexuals?
The latter, obviously.
Why would you do that? It's not very clever?
This response has been erased.
send it 2 uce, newB.
I get around 40 spams a day on Grex, thanks to several of my addresses being on the web. I, for one, am pretty annoyed by it, but I don't know what we can do. Of course, I don't know how many more I'd be getting without Marcus's spam filter, but my sense is still that his approach isn't working. Can we allow users to filter their own spam on the new machine?
I get hundreds of spam messages per week. I often delete important real email by mistake while deleting the spam. It is very irritating.
This response has been erased.
Yes, 90+ percent of the e-mail I receive at grex is SPAM. I haven't yet made the effort to set up procmail rules to try to filter out the obvious ones...
You can try my .procmailrc (and .forward) - I filter out about 2/3 of my spam that way. Maybe more, as I am sending some larger categories to /dev/null and just saving stuff that might also be real mail (hi, hello, test....).
This response has been erased.
Only add names of people you want mail from at the beginning if you find that mail from them is ending up in your spam folder(s). I filter into different folders out of curiosity - you could filter it all to /bulk/ instead. I was filtering 'hi' 'hello' and 'test' because of a recent virus that no longer appears to be around (the SCO virus) so you should probably remove those filters or you might catch a lot of friends in them. If you tend to get mail from your friends on the subject of medications, or refinancing, remove those filters too. When something slips through my filters, I look at the header to see what I might filter on - sometimes something simple like X-Authentication or X-Mailer that nobody else uses. Half the Nigeria spams get through - I might add 'Good Morning'.
This response has been erased.
When you set up a filter, it automatically creates folders when it catches spam for them. You can mkdir to create them yourself in ~/mail. You can also save mail to folders and they will be created. To see the different folders, type L while in Pine. I delete my folders after looking in them and finding nothing but spam.
This response has been erased.
Congratulations on your first three filtered spams! Today I added a filter for X-MSMail-Priority:.* when I got two spams with that in the header (type H to see the header). Trying to sell vkiagra.
Which twit accounts are currently active? I would like to fit all the twits onto one line of my twit filter. Is the polytarp account, or the dah or asddsa or naftee or dah account still open?
None of those accounts are still active.
Neither are the boltwitz account(s) . You might want to add the coopcf account, since I plan on adding some more shit with that account.
You can find that out using the 'finger' command. For example, "!finger -m asddsa" yields "no such user".
That's not foolproof. The 'naftee' account still exists but I can't use it.
re 289 Not anymore!
Since *all* of my e-mail is spam, I set up a rule to bounce it all.
This response has been erased.
Glad to hear it - how many do you catch every day?
Jim finally succeeded in attaching a file to webmail he was sending with
opera, and mailed it to me. It is plain text. I got the email without the
attachment but with the following explanation:
--------
WARNING: This e-mail has been altered by MIMEDefang. Following this
paragraph are indications of the actual changes made. For more
information about your site's MIMEDefang policy, contact
Frisco Roque/Rex Roof <admins@localhost>. For more information about
MIMEDefang, see:
http://www.roaringpenguin.com/mimedefang/enduser.php3
A non-multipart attachment named feb29.eml was removed from this document as it
constituted a security hazard. If you require this document, please contact
the sender and arrange an alternate means of receiving it.
[ Part 2: "Attached Text" ]
this has one attachmeent of text. feb29.eml from /D/kermit/
Would it have gone through if it had been .txt instead? Some filters are on the extensions.
That is what I told Jim. I have no idea what .eml meant to him. email? I recall sending .slk files which Windows interpreted as Excel when they were slovak. We can't use smtp mail until he figures out the smtp server for WCC -anyone know that? He wants to compose offline, so he ended up composing the mail at home, uploading it to grex, importing it into the message body, sending it to himself at WCC, going to webmail there, and forwarding it. He sent me a test mail and it got defanged. He wants the From address to be at wcc - is there some way to set Pine to show a reply-to address or a From address other than grex?
No, it's not possible to reset the "From:" field in Pine on grex.
Is there some way to set a reply-to address with Pine? I also tried a program which does stmp/pop mail but WCC wanted a domain name and password which differ from his email login and password and they don't make public (at least online) what these might be. The handout also mentions nothing about pop mail at WCC.
Yes, the commands are:
M ; Main menu
S ; Setup
C ; Configure
Then page down to "customized-hdrs" and
A ; Add value
At the prompt, enter
Reply-to:
(Be sure to include the colon).
<RETURN>
to accept the cahnge, the
E ; Exit config
Y ; to commit the changes
When compoosing a message, use
<CTRL>R
to bring up the "Rich Headers", which will include the "Reply-to:" line.
I typed Reply-To: instead of Reply-to: but that seems to have worked. At least it shows up in the received header. Will the recipient's email program ask if they want to reply to the reply-to rather than the from address?
It's supposed to, Sindi.
The ".eml" format is how Microsoft usually does attachments. It's proprietary, of course (those bastards!) but there's an option somewhere to use plain-text or some other simple format.
".eml" has also been used as a virus vector.
This time Jim tried to email me as attachment sax.x and it disappeared en route with no error message. I suggested he mail things with grex reply-to and give up on webmail for writing mail. WCC also sent us server info - smtp.wwwnet.org pop mail uses stu.wccnet.org (like webmail does). I wonder if Jim named his file .eml or if WCC's mail program did that. He said he had to use Opera to attach it but I think it was linux Opera. Lynx would not do attachments to webmail and told us so.
Jim finally got it right and mailed me an attachment, at about the same time that I figured out that to use non-webmail at wcc the regular login is jdeigert@stu.wccnet.org and the email login is plain jdeigert. I have smtp mail working now with 2 DOS mail programs, but WCC does not like the email password (which worked for regular login and for webmail). No big deal, he can use webmail to read his mail but it is a pain to write mail with it.
This response has been erased.
I have changed my filter to send a few things directly to /dev/null (last line of the filter) instead of ....../bulk - viagra, via-gra, viakgra, xanax, cialis, vicodin. Anything addressed to webmaster and me goes to /dev/null. Have fun with your spam net.
system is slow, and loads are high. 4:29pm up 27 days, 18:18, 50 users, load average: 28.93, 29.96, 27.39
That's normal!
/tmp has a lot of files called "newuser.xxxxx" there is a lot of junk in /var/tmp There seems to be no script to successfully clean up everything.
There is a script, and it does a pretty good job. But thanks for pointing out that it may overlook a few things.
newuser ttyp6 22Feb04 9days 7 2 /usr/local/bin/bash newuser ttyp0 10Feb04 20days 11:39 10:17 pine long emails?
??? How did somebody get the name "newuser" ?
re 309 Normal? sure... This looks normal: 6:04pm up 30 days, 19:54, 42 users, load average: 5.55, 5.65, 5.74
Only 15 people "on" according to "finger", but this is probably due to something refusing to allocate pty's. Also, Grex is excruciatingly slow. Maybe someone can clear this problem?
It's the return of the cold weather. Grex moves slower in colder weather. 1:10pm up 33 days, 14:59, 52 users, load average: 29.58, 29.24, 24.87
Fixed; it's coming back down to normal: 3:02pm up 33 days, 16:51, 60 users, load average: 4.86, 9.67, 15.73
wow much peppier
HAWT
We dialed 761-5041 and got screenfuls of gibberish and beeping scrolling by, followed by a disconnect. Did this three times. Tried 761-3000 and got BUSY, then tried again andgot through okay. Modem going haywire? Would this indicate that hte last modem in the queue is the culprit. Antoher grex user reporting things scrolling by and gave us her computer to fix. It is not broken. She said grex also stopped letting her type anything (disconnect?).
Grex was down/unaviaolble between the hours of 12:45 and 15:20 GMT (approximately). At one point it was inaccessible by tel;net but could be pinged, then later it bgecame unpingable also. What happened?
I think /tmp filled up, but I'm not really sure. I came in to find that a couple of the drives wouldn't fsck automatically and had to be fsck'd manually. I couldn't log on at the console and had to power-cycle the machine, which probably contributed to the fsck problem. (Dirty shut-downs are never nice.)
You're right there.
Simetmoes O raeyll wohs O cuodl tpye batter.
This response has been erased.
Looks like I fatally fucked up item 320 in the test conference!! SOrry guysd! please fix
:(
No, I don't think you do anything to the item. It looks fine to me. Especially in light of its first response.
O, well, that would mean ryan is fucking the system up, which is prefectly fine with me. But it still doesn't solve the problem!!
I will be plain: there is nothing wrong with the item in the test
conference. The author of the item used an executable as the text of
the item. It takes a while for picospan to process the file because of
its size, but that's all.
Someone (rational? salad? I forget which) has now entered a very similar
item in agora. (I don't remember the number. The commands
browse 310-328
and
browse 330-last
should find it.
The item is not similar at all.
Respond or pass? Got error 13 (Permission denied) in opening item file
An item was taken off-line for a while. It has now been put back, with the text modified.
You mean you PURPOSELY vandalised the bbs.
This response has been erased.
This response has been erased.
Jerkoff is posting /etc/passwd all over the place. You had the privilege of seeing the Man in Motion.
haha
Item 331 gave me an error when I tried browsing it, because the file is gone. Typing "forget 331" fixed it so browse doesn't choke.
I don't know who removed 331, nor do I know how it was removed. I jury-rigged something. I'm kind of surprised it worked. :)
anyone know what/why there are loooooong delays when going from one item to the next after the rfp prompt? and also in starting to read a specific item frm the ok: prompt. i mean times like a minute or more, not just a net-lag few seconds.
My *guess* is that it has to do with certain text being entered into items and then removed by staff. The removed text is quite large.
What hardware problem might be responsible if we are unable to do Y and have it print mail from within pine, on two computers? On a third computer it all works. We tried two printers (which worked elsewhere), two cables, two computers, one with 2 different modems on 2 comports (2 and 3), two of grex'es phone numbers (3000 and 5041). The printer works otherwise and will PrtScr mail. Is it possible to print (on your home computer) files that are in your home directory and if so how, short of downloading them first? These two computers are ones we put together for a friend who likes to print all the jokes people send her forwarded in email. Also when she forwarded mail from grex (pine, f) the attachments arrived as large strings of 7-bit characters instead of images, tho the receiving mail program (webmail) identified them at attachments before she tried to read the mail. When I sent her mail directly with an attached image that worked (she saw the image). WE are using Kermit. Once we had a 14.4 bps modem that would not, with kermit, print from pine. But we tried 3 modems tonight. (I did not want to take apart my computer to try the one that was in it.)
I print text files from my grex directory wby selecting the text and then using Print Selection (Apple-P).
Our keyboards don't have any Apple-P key. I discovered that the problem is kermit will do prYnt from c:\kermit but not d:\kermit and now we have to talk her through making a new directory, copying files to it from the old one, and editing a batch file to change d: to c:. Is there a way to print directly with a non-Mac computer from one's home directory at grex? I think you can print from lynx directly.
Are they the same version of kermit? I'm *really* surprised that kermit would care which drive it was on for printing. It seems more likely that there is a difference in the terminal-support settings. Pine uses "attached to ANSI" for its printing. (I don't use kermit, so I can't offer suggestions on checking its settings.) I think lynx uses the same print mechanism Pine uses. If one works, the other should. "Print Selection" is a feature of the terminal software. I seem to recall seeing it in ProComm or ProComm Plus. Again, I don't use kermit, so I can't comment on its features.
Re #345: For printing Grex text on your local printer, you might try
the "pcprint" command, e.g.
pcprint FILENAME (from a shell prompt)
or read pass 100 | 'pcprint (in Picospan, to print an item)
This might or might not work, depending on how your terminal software
is set up.
There is a new vote in progress on a proposal to amend the bylaw article on member votes. Voting runs through March 23. Type vote at a shell prompt or !vote at any other prompt to read the proposal and/or cast a ballot. See item 122 in Coop for discussion.
I should add that since this is a bylaw amendment, at least 75% of members voting must vote in favor in order for it to pass.
I have identical versions of kermit in c: and d: and the one in c: prints and that in d: does not. Three computers will print from c:, three won't from d:, on four printers, various cables.
My latest linux comes with microcom, a little dialing program smaller than minicom, which when I try to use it to log into grex, immediately after I give my login, does Password: Incorrect password. It appears to be sending some string which gets interpreted as password, and does not leave me a millisecond to enter my own password. When I try this at mnet it dutifully pauses for me to enter my password, twice. Other people say it works for them. What might be happening at grex to cause problems with password giving?
This response has been erased.
I don't understand the question in #352. I don't see any strings being produced before I am told Incorrect password - this appears immediately after the request for Password:, with no time to enter my password.
This response has been erased.
I prefer 6 bit
I can try this and get back in ten minutes. This all works properly at mnet for some reason, just not at grex. Works for other people elsewhere.
Isn't m-net down?
Ctrl-M made no difference. I wonder if the grex modems are a problem. I will try without hardware flow control next.
M-Net isn't responding to my calls. Is this a known problem? Is someone working on it?
(I was referring to the fact that keesan had said she had tested something on m-net)
I officially give up on microcom. I will learn to compile the latest version of kermit since the precompiled one for SW40 complains about floating points and the one for SW3.5 does not do FTP, it is too old. make linux KFLAGS=-DNOFLOAT after unpacking the source code.
We just rebooted.
and who is 'we'? % ping -c 3 arbornet.org PING arbornet.org (209.142.209.161): 56 data bytes --- arbornet.org ping statistics --- 3 packets transmitted, 0 packets received, 100% packet loss
(I think the "we" is grex, which was taken off-line last night, as described in the motd.)
re 364 ... 363 was ref to 359.
Thanks kip and gelinas !
Hmm. Has ttyp0 been fixed? it seems to be working properly. Contrast this with testerday, when in an effort not to get a borken terminak I opened several tty sessions simultaneously, intending to close the ones down htat complained "server refused to allocate pty" once I got a working one. It took me four goes before I got one I could type into.
Has been awfully quiet around here the last couple of days.
Is that a problem grexstaff can fix?
No but we have a twitstaff on call.
The message "server refused to allocate pty" usually indicates a telnet queue, which SSH doesn't deal with very well. Opening several different connections at once is not going to make it better.
ahhh... thanks.
Is that a problem related to SSH in general or just the implementation htat GREX uses?
BTW, Grex hasn't rolled over yet.
Speaking of ssh: psftp: no hostname specified; use "open host.name" to connect psftp> open grex.cyberspace.org login as: drew Sent username "drew" drew@grex.cyberspace.org's password: sftp-server: Command not found. Fatal: unable to initialise SFTP: could not connect psftp>
I always get an obscene message when I try to ssh into grex. Plus it is unbearably slow compared to m-net.
Is this some new virus? And if it is some local vandal, how are grexers supposed to be reading this anyway? From management@grex.org Sat Mar 20 20:47:10 2004 Date: Sat, 20 Mar 2004 20:27:48 -0500 From: management@grex.org To: keesan@grex.org Subject: E-mail account disabling warning. [Part 1, Text/HTML 12 lines] [Unable to print this part] [Part 2, Application/OCTET-STREAM (Name: "Text.pif") 28KB] [Unable to print this part]
Yes, it's a virus.
ssh works for me, sftp does not.
sftp-server is a seperate program which is enabled as a subsystem in your sshd_config (for openssh, at least). "sftp-server: command not found" implies, i believe, that the subsystem directive is there in grex's configuration, but the program itself is not. scp should suit your needs, though.
sftp-server is actually run in the command shell by your ssh client, when you connect. This seems to happen *before* your environment is set up, too, which means there is no path to search. So your ssh client has to have, in its configuration, the proper full path to where sftp-server is *on the remote system*. I ran into this on FreeBSD. WinSCP's default for where sftp-server is located is not the same as where FreeBSD installs it.
same thing. it's not there.
So where is sftp-server on Grex?
It looks like there isn't one. Presumably our version of sshd is too old to have come with it.
You have several choices: