197 new of 293 responses total.
I am curious to know if anyone else has been receiving multiple unsolicited e-mail from a "Free Stuff Queens" list at this host. Although I have indicated a willingness to file suit against the owner of the domain, developingdots.com, and also have received notice indicating my address has been unsubscribed, I continue to receive multiple copies of the same unsolicited mail.
This response has been erased.
pine config question ... while i prefer mail to pine (or outlook express, eudora, etc.) there are times when attachments are sent adn i use pine to handle that. so, i created a different directory into which to send email that needs pine's attention. although pine does what i want, it starts with an error message pine bad context no '[' in context: /path/to/inbox from setup (reflected in .pinerc) personal-name = <No Value Set: using "TS Taylor"> user-domain = <Value is Fixed> smtp-server = <No Value Set> nntp-server = <No Value Set> inbox-path = /a/t/s/tsty/pine/INBOX folder-collections = /a/t/s/tsty/pine/ news-collections = <No Value Set> incoming-archive-folders = <No Value Set> pruned-folders = <No Value Set> default-fcc = <No Value Set: using ""> default-saved-msg-folder = <No Value Set: using "saved-messages"> postponed-folder = <No Value Set: using "postponed-msgs"> read-message-folder = <No Value Set> so, what is 'context' wehre is there aupposed to be '[' (the open bracket) and how do i "set values" coreectly to change (i guess) system defaults where 'using' "whatver" is not explicit?
Well, here is what I see in my .pinerc:
} # List of directories where saved-message folders may be. First one is
} # the default for Saves. Example: Main {host1}mail/[], Desktop mail\[]
} # Syntax: optnl-label {optnl-imap-hostname}optnl-directory-path[]
} folder-collections=IMAP Server {g.imap.itd.umich.edu}mail/[],
} IFS home directory mail/[]
}
} # List, only needed if nntp-server not set, or news is on a different host
} # than used for NNTP posting. Examples: News *[] or News *{host3/nntp}[]
} # Syntax: optnl-label *{news-host/protocol}[]
} news-collections=
However, I don't see a way to set the [] in Config/Setup. Could be because
I'm looking at Pine 4.33
TS, look at your .pinerc with less and with Pine's config
Why so slow right now?
The usual culprits are 'load', which can be checked with the "uptime" command (in picospan, !uptime at the nearest prompt), and 'network traffic', which I've not found a way to check from grex (the usual tools are disabled/restricted because of their usefulness to Bad Guys (TM)).
Fixed.
uhhh, re 100, 101 ... that looks like the um pine stuff although there
is some similarity. and , i *have* looked with both less and setup
NumerousTimes (tm) trying to figure this out.
does your .pinerc explicitly state:
# Syntax: optnl-label {optnl-imap-hostname}optnl-directory-path[]
folder-collections=IMAP Server {g.imap.itd.umich.edu}mail/[],
IFS home directory mail/[]
(the first-column braces are indent chars, i presume)
for the record, .pinerc states:
# Path of (local or remote) INBOX, e.g. ={mail.somewhere.edu}inbox
# Normal Unix default is the local INBOX (usually /usr/spool/mail/$USER).
inbox-path=/a/t/s/tsty/pine/INBOX
###################### Collections, Folders, and Files #####################
# List of incoming msg folders besides INBOX, e.g. ={host2}inbox, {host3}inbox
# Syntax: optnl-label {optnl-imap-host-name}folder-path
incoming-folders=
and there are no [] indicated anywhere ???????????????????
Yup; that's a copy-and-paste from my current .pinerc. Here's what shows
on grex:
:r ~/.pinerc
# Updated by Pine(tm) 3.96, copyright 1989-1996 University of Washington.
#
# Pine configuration file -- customize as needed.
{Ellipsis. JLG.}
########################### Essential Parameters ###########################
# Over-rides your full name from Unix password file. Required for PC-Pine.
personal-name=
{Ellipsis. JLG.}
# Path of (local or remote) INBOX, e.g. ={mail.somewhere.edu}inbox
# Normal Unix default is the local INBOX (usually /usr/spool/mail/$USER).
inbox-path=
###################### Collections, Folders, and Files #####################
# List of incoming msg folders besides INBOX, e.g. ={host2}inbox, {host3}inbox
# Syntax: optnl-label {optnl-imap-host-name}folder-path
incoming-folders=
# List of directories where saved-message folders may be. First one is
# the default for Saves. Example: Main {host1}mail/[], Desktop mail\[]
# Syntax: optnl-label {optnl-imap-hostname}optnl-directory-path[]
folder-collections=
{Ellipsis. JLG.}
# List of context and folder pairs, delimited by a space, to be offered for
# pruning each month. For example: {host1}mail/[] mumble
pruned-folders=
# Over-rides default path for sent-mail folder, e.g. =old-mail (using first
# folder collection dir) or ={host2}sent-mail or ="" (to suppress saving).
# Default: sent-mail (Unix) or SENTMAIL.MTX (PC) in default folder collection.
default-fcc=
{Elision of remainder. JLG.}
That's all I know.
yeh, right, tha's how my grex pine STARTED. then i added the explicit stuff ... adn get teh errors. your grex pine has NO customization. was teh first example from um's pine?
Yes.
staff: question: is the protocol gelinas used on um's pine applicable to the pine that grex uses?
Last night and this morning, Grex has been randomly killing my connection after short (five-ten minutes) periods of time. The rest of my internet connection works fine while this is happening, but I get nothing here. Any ideas?
try a traceroute from your starting point to grex; look for dropped packets.
This response has been erased.
Re 112 re 111: This has happened to me intermittently, but way too often, lately. But I was dialed in, and my starting point was just the term server, I think.
Amidst a download of a pile of e-mail, I got this on my tty: Sat Jan 26 13:52:03 This terminal has been idle 15 minutes. If it remains idle for 5 more minutes it will be logged out by the system. Can't Grex recognize that a terminal running something like sz is NOT idle and should NOT have unwanted data dumped on it, let alone be cut off?
i thought that had been fixed when yu were in short pants?
I'm rather puzzled as to how you saw the message, if it was in the middle of a file transfer.
The "down until approximately 1 PM" notice on Grex was still playing at 1:42 PM. If backups are going to take longer or be started later, it would be much friendlier of staff to correct the estimate to be closer to reality.
The problem may be that if it's not known that it'll take longer before the backups are started, changing the estimate would require modifying a filesystem being backed up, which is generally considered a Bad Idea.
Thanks, Scott, for doing that backup!
Re #119: Not necessarily. It's possible to put both multiple filesystems and multiple OSes on a single machine, and even on a single hard drive! A small special purpose OS/filesystem could be dedicated to showing the estimate and running the backup software.
Re #117: It was actually just a bit after downloading over 1/2 megabyte of e-mail and conference text. Regardless, it should not have occurred. Even if the idle-killer checks the tty to see if it's in raw mode and ignores it, one could be logged out immediately after finishing a download if the timing is exactly wrong.
Pl-e--a---s----e-----! -I-- ---n----e-----ed- --h---e----l-----p!- B--r---o----k-----en- --t---h----i-----s -i--s---!----
Grex does not appear to be running the Network Time Protocol daemon, and its clock is coming up on a minute fast.
threr might be an Xnix version as well, but nistime32 (googls search) is an excellent windoze time synchronizer, fwiw. even works on xp!
Re: 110 - Back when I first deployed Pine at UM, I used the IMAP protocol.
I doubt that anything has changed since then. If you set your mailbox to
{mailserver.name.whatever}INBOX, Pine will attempt to use IMAP to connect to
it.
Thanks to whoever fixed the modems (if anyone did) so that I was able to download 43 packets with kermit instead of crashing after packet 12, at normal speed. Are the newer faster modems installed now?
mvpel, whaqt i need to do is point pine to MY CHOICE of inbox files (into which i save special emails tha require pine). as shown above, i hard-coded the INBOX path but still get that sily error about " context ']' " i wo nOT wnat pine to handle everything. mail is JustGreat (tm).
Backtalk is down right now.
I'm pretty sure you can just specify the file pathname instead of INBOX.
Personally, I prefer mutt to pine when I have to send mail with attachments.
re 130 ... omit the filename? the 'directions' dont indicate path-only as i read them. ,
Hi, I'm Twill!
yer mother.
Backtalk just stopped accepting my username and password. It keeps spitting the authorization dialog back at me.
It did that to me too.
Seems to be working now.
Is there any way to protect picospan against denial-of-service attacks, such as mounted by twill and bdh3?
This response has been erased.
Free expression is fine, but one or a few postings of a response is usually sufficient, and the effect of these denial-of-service attacks is to cause inconvenience for other users, which may in fact cause an obstacle, especially to new and less experienced users, and drive them away from the system.
don't leeron's posts have the same effect? :D
Twill's posts weren't exactly a denial-of-service attack, but the program that generated them could easily be used to create one.
This response has been erased.
I was prevented from reading new meaningful responses in reasonable rapidity due to the necessity of having to react to hundreds of items to which the meaningless drivel was attached. I used fixseen, but new or inexperienced users may not know this, and it may also have skipped a response that I would have liked to have read. Doing this was an intentional creation of an obstacle to the convenient use of the system. It is disrespectful to all users.
I agree with Rane. The system is much less usable after a mass posting such as this. There's no way to deal with it which isn't worse than the problem, though. What would be nice is a "fixseen <user> in which all items would be "fixed" if the last response in the item was from that user. Or maybe even: fixseen -n <number> user in which the item would be "fixed" if there was a response from the user in the last <number> responses. But it would probably be hard to implement this, cause a big drain on the system, and wouldn't really solve the problem. I don't think there is a way to solve it.
yea my finger *really* hurts bad from having to press <RETURN> so many goddamn times. took less time than it does to read some peoples' posting here. it twas annoying, but giving it this much attention is annoying too.
This response has been erased.
(what oval said)
(what flem said)
I figure it wasted about 10 minutes total of my time to have to keep hitting the enter key and then wait 5 sec to get to the next unwanted item, per item.
Couldn't you just 'twit filter' the offender?
<g> Remember: security = 1 / (1.072 * convenience)
twill/polytarp's actions were a crapflood attack, a type of DoS attack. (I used to do the same thing as a counter-crapflood against the "last item" crapfloods on M-Net, and oddly enough, some of the people saying it's not a DoS attack here/now said it was there/then....) Re #146: Your connection must be really fast, and you must have a lot of time to spend waiting for the next item to come up. For some people the extra time can be a large fraction of what they have available. Those people are effectively denied service.
I think what you did was, and what they did was. I'm at least consistent.
A solution is Loss of Personality. Not just splatting an
account, but taking the account, locking the password, making the
account only readable to staff/root, redirecting mail to dev/null.
Does not allow account to be re-created. Offender cannot
use old account files easily in new account, mail does not bounce,
it disappears.
So they just create a new account.
So tpryan, if someone did that to you would it mean that you would lose your personality? I like to keep my personality offline.
...in a shoebox with some potpourri and special mementos?
#157 -- i feel so decieved!!!
(on a side note, it's pretty easy to write a shell script to undo the
damage caused by someone flooding a large conference with responses;
regenerating the picospan information is a different story)
resp:158 You looked in my special shoebox didnt you!!!!!
oops.
#!/usr/bin/expect -f
set timeout -1;
set name twill
set host cyberspace.org
set password aqq1#y
proc login {phost user pass} {
spawn ssh $phost -l $user
expect "password:"
send "$pass\n"
return $spawn_id
}
set spawn_id [login $host $name $password];
expect "$ "
send "bbs\n"
set item
while {$item <= 146} {
expect "Ok: "
send "r $item\n"
send "
"
send "r\n"
expect ">"
send "Hi, I'm Twill!\n"
send ".\n"
incr item
}
resp:139 I don't mind free expression. Express yourself all you want, just express yourself where I don't have to look if I don't want to, thanks.
This response has been erased.
Is that an expect script? Seems like massive technological overkill. This
isn't a task requiring intelligence. How about this:
Open some kind of text editor window on your computer. Type
Hi, I'm Twill!
.
Hi, I'm Twill!
.
Hi, I'm Twill!
.
Hi, I'm Twill!
.
Copy those eight lines into your cut/paste buffer, including the newlines
at the end of each.
Telnet to Grex, and enter the bbs. Type
set noedalways
read all nor forceresponse
Hit the paste button repeatedly. Each click will probably post the message
to four items. Shouldn't take more than a few minutes to do the whole
conference. The whole job should take fewer keystrokes than typing the
expect script.
Yeah, but that would have required me to gain knowledge of the commands shown above. While it is certainly more efficient and quick to execute, neither of those were included in my goals. Also, I use SSH!
Besides, you know the sysadmin code. "If you have to do it more than twice, automate it."
For the last couple of days, the response box in Backtalk has been using robot characters. I don't know what the font is, but it's a 60's computer font looking type of character. The comma doesn't descend below the line and looks like a miniature "L" on it's side... I thought it was my computer, but realized this morning it's happening for both my home and work computers, so it's probably not a change I made. I'd guess this is part of the upgrade to 1.1.11, and that it's not intentional. Please fix it! Oh, my eyes...
Hmm...the text entry boxes used to be Courier on my system. Now they appear to be Lucidia Console. Interesting...I had no idea that font could be controlled by anything but browser settings. Lucidia Console isn't any worse looking then Courier, really, but it's an optional Microsoft font and other systems may substitute one that isn't so pleasant. Courier is less of a problem since pretty much every system will have it.
CSS, baby, yo.
Yep. The source of backtalk pages now has a <STYLE> tag with
contents: TEXTAREA {font-family: monospace;}
Is there any way to configure my browser to make it give me a more reasonable choice for the font?
Most browsers have a setting to override a document's specified fonts
with whatever you have set; whether, and how, this works, depends on the
browser.
Interesting. I discovered that there are now a lot of browsers that default to using a proportional font in TEXTAREA boxes. (Mac versions of IE, for example). This has really obnoxious effects, because responses are *displayed* in a proportional font, but the input wrapping is based on a proportional font. Really icky. So I used a style sheet to force TEXTAREAs to use a monospaced font. It sets the font to "monospace" rather than any monospaced fant in particular because I was figuring that that would be whatever monospaced font is prefered by the browser. Obviously Some browsers are doing something different. I shouldn't be surprised. This is easy to change to something different. I'm open to advice what should be used. Courier?
Wingdings.
I never noticed a difference (mac/NS4.7). I use monaco-9pt. as my monospace font, and all the plain text on my backtalk pages is displayed in it. There is a setting in most browsers' options to determine what font and size the browser uses for monospace and proportional defaults. You may have other options, depending on your browser.
Courier would be vastly better than what I'm getting now. I'm using IE 5.5 both at home and work. If there's something simple I can do to set up my browser's posting-box font to be something better, I'd be happy to do it if someone lets me know what can be done. What I'm getting now is hard to live with.
Jep, I don't know the exact menus for IE, but look for "appearance" or "font" settings. There should be settings for regular and monospaced fonts.
Aha. Okay. Here's how to do it in IE: Tools-->Internet Options-->General-->Accessibility Select "Ignore font styles specified on Web pages (Why would this be under Accessibility???) Now to see if the cure is worse than the disease...
Nope, that cures the problem and I haven't seen anything I lost by changing the setting.
It would under accessibility because people with vision problems would want to set a really large font, overriding Web site settings.
Another cure for the problem, if you don't want to dial in or telnet to grex, is to use a non-graphical browser which uses screen fonts. Like lynx.
...One cost being that you lose mouse support. Whatever floats your boat, I guess...
You don't need a mouse to use the conferences, but lynx for DOS does have mouse support (optional).
As everyone knows who is on Grex now, it is running <low voice> rrrreallllly ssslllooooooowwwwww. In addition I received an error message "illegal instruction:core dumped" while trying to send mail in Pine.
telnet session. yup, realllly slow: Respond or pass? ld.so: swap space exhausted for mmap data of /usr/lib/libc.so.1.9.1 Pipe interrupt? Respond or pass? <...later...> Respond or pass? ld.so: swap space exhausted for map heap of /dev/zero Pipe interrupt? Respond or pass? <...later...> Respond or pass? Can't execute "more"! Pipe interrupt? Respond or pass? Can't execute "more"! Pipe interrupt? Respond or pass? !uptime 11:52am up 13 days, 22:04, 20 users, load average: 69.45, 79.94, 76.27 What is up with those swap errors?
People trying to compile big things which we don't allow on Grex.
Which is not allowed: the big compiles, or the things themselves?
The things themselves, so the perps try to bring over their own versions in hope of getting around the "crippled" Grex versions. Except that the Grex versions are not actually crippled, it's the network side that's blocked.
OK. I knew that there were some things not allowed, and properly blocked ;) but your comments opened the possibility that large compiles, themselves, were unwelcome. (I've never written a really large program, so I don't really know what troubles they would cause other users.)
The larger a single file, the larger the symbol table must be to compile it. Breaking the project into many files that are careful about using static whenever possible would greatly reduce the compilation load.
Earlier today I was having trouble logging in with backtalk. Telnet worked fine ... It seems to be fixed now.
Dialing in tonight I got queue-droppings before I could log in. Only one level deep, but that's just *wrong*.
Has anyone else been getting incessant chat requests from smart007 and if so can this account be cancelled? (I will put this login on my don't acceptlist now).
Accounts are not cancelled on Grex. It does no good with an open system. Another account can just be taken out and we are running out of usable UIDs.
I can't imagine Grex cancelling an account because of annoying behavior! None of us would be left except the last-standing staff member.
I don't see that there should even be a problem with smart007, since you can block anyone from chat requests.
I did that, but first I had to waste time dealing with a chat request every 3 second for about 20 minutes while in themiddle of another chat.
This response has been erased.
It takes less than half a minute to add someone to your .nowrite file, and even less if you write a simple shell script to do it.
You would not have had to waste 20 minutes if you had excused yourself from that chat and taken care of the problem.
She would not have to have done anything if an obnoxious user hadn't accosted her. keesan is not the villain here..
but she _is_ the whiner here.
To put this in absolutist perspective, the effort keesan put into complaining about this abuser is far beyond what would have been required for her to eliminate the problem he represented to her.
Sindi, did you tell the person that you did not want to chat right now? If you did and s/he persisted than complaining is appropriate. If you did not, how is the person supposed to know? Granted the requests should have stopped after 2-4, but maybe the person is clueless to how it all works.
This response has been erased.
#205> Yes, but it wouldn't have gotten her nearly the attention. #204> Hear, hear.
Is there some way to send a telegram while I am chatting with someone else? The people I chat with often have trouble understanding English and this particular one would have been offended if I tried to hang up to do so, or change my plan, so I just kept hitting Ctl-L every time the chat request scrambled the screen (every few seconds). (It took about 20 minutes to say goodbye after that, he wanted to chat for just another 30 min, but I finally promised to be online Friday at noon - it is hard to find someone to chat with you in your own language.) After changing my write permissions, I sent a short and polite letter to smart007 explaining what had happened and suggesting that they request a chat only once, and first do a who to find out what the other person is doing at the moment. Did not hear back.
suspend the process.
Cant you shell out of write with a !?
I keep chats and writes off until I am done with other things and are ready to recieve them.
I have had a few chatters get offended when they tried to reach me and I was in Pine (cannot get chat requests while in Pine). One clever guy sent me an email that he was online. Can you shell out of chat?
^Z works with talk. use "fg" to get back to it.
!man tcsh
for more info on fg, ^Z, and all that.
chat/write support ! escapes, and I think it's possible to suspend them too.
Thanks, I will tell the chatter to wait 30 sec next time while I shell.
For the record, I still think the person who invented 'talkdaemon' should be lined up and shot.
Mesg's author, presumably, earned his pardon.
a !mesg N and then a quick restart of your chat <loginid> will stop the problem. that also presumes yu haven't sent a tel to your annoyer already.
Thanks, that sounds much easier than exiting chat and running the change program. I did not know you could even shell from chat.
This response has been erased.
Because only one kernel syscall can run at a time.
Not true. If *that* were so, lots of other things would be slow, and a system call that didn't return would bring everything to a stop.
Actually, only one kernel syscall can run at a time (at least under sunos), but that's a moot point, since long-running system calls typically block on some event and allow other things, including other system calls, to use the CPU. None of this is particularly relevant to why ps et al are slow -- the real reason is because sunos doesn't offer the accellerators that later systems have to access kernel structures. In SunOS, everything is done using "nlist" and "/dev/kmem"; on many more recent systems, there is a kernel ksyms data structure and things in /dev and /procfs that access the kernel ksyms data, and even more usefully, stuff in the kernel that provides access to the process table, per-user kernel data, and other stuff without the bother of going through kmem. This is just one of the things that contributes to kernel bloat in linux, solaris, etc.
Valerie kindly pointed out that there is no need to set up your own twit filter, you can type 'ignore username'. I presume this holds for an individual session. Is there some way to make it permanent for use when someone has trashed all the items in a conference? I. e., can I put the line in some file that runs whenever I use bbs?
The spam rate here seems to be increasing. Is there a central spam filter for incoming mail, for either source or recipient addresses (like Undisclosed-Recipient@cyberspace.org)?
Ask Marcus, who is working on it. I forward my spams to him at UCE, with full headers.
I am unable to connect with lynx to google, altavista, or alltheweb. It says it is trying to connect to 123.45.567.78 (sample numbers) and nothing happens for 60 sec. I Ctl-C to exit.
!finger uce for more information on spam filters on grex. We purposefully don't publish detailed information on the spam logic on grex, because we don't want to educate spammers on how to evade them.
resp:127 :: another lynx trouble report, from party:
stephenl: Does anyone know what is wrong with lynx?
stephenl: It won't connect to any websites. It always goes back to the
localhost.
I wiggled lynx a bit -- I usually don't run it -- and it does seem
to fail to connect to anything.
Apparently gryps bit the dust. Someone is supposed to go reboot it tonight.
this is a tast since im new at this and i dont know how to work it yet lol
Try joining the test conference ("j test"), if that's what you need to do.
re 224 .. ignore username works in party .. does it also work in the real world? this would be interesting.
Heh. TS, is picospan really realer than party? In the real world, ignoring people sometimes works, sometimes not ... <goes off & tries it> As it stands, it does *not* hold for your current session, only for future sessions; but it tells you how to make it take effect for the current session. Someone wrote something to add twit-filter lines to your .cfonce (your bbs startup script). Running it again for someone else will append another set of lines, I think - so the net effect will be to replace any previous twit filter. (You can, however, list all the users you want to ignore on one command.) Valerie or whoever wrote it, nice job. <goes off & looks> Yep, Valerie.
for teh more adventurous (and before applying taliban tactics) there is this missive i have unabashedly plagiarized from 'a player to be named later .. if at all.' "Unfortunately, the answer to "kick him off" is yes, but it wouldn't help. All he would need to do then is run newuser & create himself another account. So, in order to fight that, we'd have to have somebody always watching, and pulling the plug as necessary. It's a big brother cat-and-mouse game. No joy in that. Actually, this is a behavior/socialiazation problem. There are other methods of dealing with such problems that are far more effective. The trick is to pull "orthogonal" thinking on him. You can't just react to him (that plays right into his mindset) and while sometimes it's possible to ignore him, that's not always effective. Things that might work include applying logic to what he says (most crude sayings are truely hilarious when looked at the right way), or applying inventive logic to him. There are some tricky knife edges to watch: he should never know for sure if you are making fun of him or are being deadly serious, and whatever you say should be at least mildly entertaining to others. If you do it right, however, you might find either (a) he undergoes a personality phase shift and becomes an interesting person, or (b) he gets bored and goes elsewhere to get his jolly's. Please do send mail to staff if this doesn't suffice or if things get worse." i believe the plagiarizm is character-for-character .. at least i hope so.
Looks like user tikcloak is launching a new connection to Grex every five minutes. If unchecked, this will be a problem in a couple of hours. The regularity of the logins say to me this is an automatic process. grex.cyberspace.org% finger tikcloak Login: tikcloak Name: TikCloak Directory: /a/t/i/tikcloak Shell: /usr/local/bin/bash On since Fri Mar 1 17:30 (EST) on ttyp3 from 203.162.56.202 On since Fri Mar 1 17:45 (EST) on ttyp5 from 203.162.56.202 On since Fri Mar 1 16:55 (EST) on ttyp7 from 203.162.56.202 On since Fri Mar 1 17:00 (EST) on ttyp8 from 203.162.56.202 On since Fri Mar 1 17:05 (EST) on ttypb from 203.162.56.202 On since Fri Mar 1 17:20 (EST) on ttypd from 203.162.56.202 On since Fri Mar 1 17:50 (EST) on ttype from 203.162.56.202 On since Fri Mar 1 18:05 (EST) on ttypf from 203.162.56.202 On since Fri Mar 1 18:00 (EST) on ttyq2 from 203.162.56.202 On since Fri Mar 1 17:25 (EST) on ttyq3 from 203.162.56.202 On since Fri Mar 1 18:10 (EST) on ttyq7 from 203.162.56.202 On since Fri Mar 1 17:40 (EST) on ttyq9 from 203.162.56.202 On since Fri Mar 1 16:50 (EST) on ttyp0 from 203.162.56.202 On since Fri Mar 1 17:10 (EST) on ttyq0 from 203.162.56.202 On since Fri Mar 1 17:55 (EST) on ttyq4 from 203.162.56.202 On since Fri Mar 1 17:35 (EST) on ttyp9 from 203.162.56.202 On since Fri Mar 1 17:15 (EST) on ttyq1 from 203.162.56.202 No unread mail No Plan.
Further investigation suggests the user began this task early this morning, with only minor outages. However, I do not share Kenneth's pessimism, as I recall Grex has an auto-kill program for multiple logins. At most, it is amusing.
This is quite amusing. There are more efficent ways of bringing grex down if someone really wanted to. (No, I won't demonstrate).
The auto-kill doesn't kick until grex hits 70 remote users.
Still, Tikcloak is a vandal, even if Grex repels this invader. The only purpose of automated logins is to cause trouble.
It could be that they're just amusing themselves with no malicious intent, you know. It still needs to be stopped, but why condemn someone when you don't need to?
Not even a *little* condemnation? Any user would know that repeated logins with no intention of using them will occupy bandwidth and processor time just for personal amusement. If they don't have consideration for others can't a *little* condemnation be offered? Just a *smidgen*?
Very well, Rane, you may condemn them lightly if you wish. :)
If you make a mistake while typing in your login or password, is there any way to delete it, or do you just have to hit return twice and wait for the delay Grex puts in before you log in again?
Backspace works, though it doesn't *look* like it works.
AFAIK ^U will wipe out your entire response. It also may not look like it.
As gull notes, count the ^H characters; enter an appropriate number of them and go on.
I use ^U in that situation.
Cool, thanks. The delay before you get the next login prompt is really really (really) annoying, so I want to avoid it.
(I think it's deliberately annoying, to discourage people from trying to guess passwords by typing things over and over again.)
I doubt it does anything to discourage twits, but it annoys the piss out of legitimate users who make typing mistakes. I'd lose it, personally. If you guys are that worried about people hacking passwords, install a version of passwd that doesn't permit weak ones that people could guess by typing them in. You've already got the limited number of attempts per telnet session, which does a lot more to discourage password hacking than inserting a delay ever will. But knowing you guys, to get it changed I'd have to make a proposal in Coop, get told that my opinion doesn't count because I'm not a paying member, and then touch off a massive flame war about how central the delay is to Grex's culture, and how if it were removed than people who can't type worth shit would log in here more often and bring the tone down, because Grex only wants people who type what they mean the first time. The delay is right up there with confiscating nail files and letting ballpoint pens through checkpoints on the "moronic security measures that sound cool until you think about them for a second" scale. In my personal and professional opinion.
Joe, your opinion doesn't count when you're being this pissy. (Couldn't resist...)
It occurs to me that it would probably not be terribly difficult to look for lots of logins from the same IP address, and just set the router to block that IP address for a while. End of problem.
I sometimes also type my password wrong since I cannot see on the screen what I am typing. Jmsaul appears to be a less than perfect typist. Is there some easy way to change things so we could see the password as we are typing?
You could do it for yourself by turning on local echo temporarily in your terminal program. From a security point of view, it would be a horrible idea to do it globally for everybody.
There is no way in hell that I want to be able to see a password as it is typed in. Even when accessing from my own computer in my own home. That defeats the purpose of a password. If passwords were seeable I would never, ever login anywhere but home, and then when no one else was in the room.
Re #251: Someone needs a hug.
There is nobody in my apartment that I want to hide my password from, and ifI did, they could watch what I was typing. In fact I often watch myself type my password to make sure I get it right. I used to know how to turn Echo on with Procomm (Alt-E) but probably other programs are different. Is there something an individual user can do to make their password appear onscreen every time without changing the echo?
Nope, and there never will be. Never with Grex, and never with anything else into which you have to type a password, unless the author of the software wrote in password functionality as a mere formality without really caring about it.
I second #259, though I do like the more recent convention in some software of displaying asterisks as you type your password. Makes it more obvious that you've actually hit the keys and the software is actually listening.
Re #257: Do I get a wet, sloppy kiss, too?
Re #254: I never claimed to be a perfect typist. I'm a good one, but not
a perfect one. I don't need to see what I'm typing, though,
because I usually notice when I screw up. Making the password
visible for everyone when it's typed is a really bad idea.
Even worse than leaving the annoying-but-somehow-useless-for-
preventing-hacking delay in.
Re #261: Not from me, but perhaps some other Grexer will oblige you.
Tease.
re 260: I often am glad for things that echo asterisks or something. However, if I were logging in from a public place, I'd worry about it. Making it easier for people to know for sure how many characters you're typing is not all that good an idea.
Lotus Notes put up a random number of heiroglyphs, rather than asterisks.
why is this so slow?
What do you mean by "slow"? If picospan ("bbs"), then it's probably because
this is a long-running conversation.
The bad password delay has been a standard part of Unix login programs for a long time. The retry shouldn't be obnoxious unless you've managed to typo your password twice, in which case, it probably won't hurt for you to slowdown and think about what you're doing. The IETF folks want to put a built-in delay of >1second (via a computationally expensive CPU loop) for encrypting K5 passwords via AES. If you want to worry about something, why not worry about that?
What is their excuse?
Re #268: It's obnoxious on one retyping. Try it. (I don't typo it
twice.)
Slow down brute force password cracking.
:)
The problem with ^H, ^U, etc. is buried in /usr/sbin/telnetd.
finding a pattern in spam: http://www.blackant.net/code/oth/random/nlp-spamfilter.php
Backtalk is not working; it stopped working about an hour ago. I can still telnet in, though.
Seems to be working now. My guess is that httpd died, then got restarted.
Looking up www.google.com, guessing (no DNS numbers available?). The internet is not terribly usable with lynx at present.
About three minutes before #277, I noticed that
load average: 12.08, 7.59, 6.00
I could be wrong, of course, but I suspect the two events are related.
"nslookup" is freezing up instead of reporting data.
This response has been erased.
I have been having probs with nslookup freezing up as well.
I'm a lousy typist. I've been known to cut/paste passwords. Of course, you can log in to Backtalk, where most browsers do echo bullets with each keystroke and where the authentication delays are, well, different. I really would like it if login/telnet/etc would accept *both* CONTROL-H and DELETE as backspace characters when entering logins and passwords. It's a pain to remember which key to use where.
Jan, are you not in probably the very best of all positions to make your dream a reality? I've heard UNIX geeks complain before about the lack of concensus on a backspace key. The discussions usually result in everyone throwing up their hands in disgust and frustration.
Re #282: I'd like that too. Very much.
Personally, I find single-character backspacing to be pretty much useless for passwords. I can't see what I've typed, so I'm never sure how many characters to backspace to correct the error that I might have made. It's largely useless for logins too, at least for me -- when I catch a mistake, I find it faster to just kill the whole line (ctrl-U works for this on every system I log into) and start over.
my password has been changed. very mature.
oh it was me who changed the password. very mature.
I am so glad to know that I am not the only person who does that! ;)
i should really write them all down somewhere.
Yep. Post them right here in Agora. Then you'll always know where to look.
:P
Hello none of your games are working in the game menu.
Hmm. I just tried one (boggle), & it worked fine for me.
You have several choices: