|
Grex > Coop13 > #350: Minutes: Grex Board of Directors Meeting. Tuesday, July 25, 2006 | |
|
| Author |
Message |
slynne
|
|
Minutes: Grex Board of Directors Meeting. Tuesday, July 25, 2006
|
Aug 8 22:40 UTC 2006 |
Grex Board of Directors Meeting. Tuesday, July 25, 2006
Board Members in attendance: Mark Conger, Lynne Fremont, Joe Gelinas,
Larry Kestenbaum, John Remmers, and Janc Wolter.
Non Board Members in attendance: Steve Andre (by phone), Jim Diegart,
Sindi Keesan, and Mary Remmers.
TREASURERS REPORT:
In June we took in $271.18 and spent $150. In July we took in $144 and
spent $150. We have 51 members and 41 paid up with the others being in
a grace period. Mark allows people to maintain memberships in a grace
period, especially long term members.
STAFF REPORT:
Remmers installed an IRC client called IRCII. Only members can connect.
It uses a wrapper. Steve made some changes to the kernel. Grex has been
more stable. We are still having users break in.
Grex is currently listed on two spam systems which is an improvement
from what it was before. Steve is working on getting us off of those
systems. Turning off mail to newusers has significantly decreased the
amount of spam sent from Grex. There was one user recently (lullo) who
was caught sending spam. The user attempted to send 110,000 pieces of
mail but only 6000 were actually sent. We need to come up with a policy
for giving new users email. The current policy is that staff will give
newusers email on request. Some suggestions for policy:
1. Limit outbound email to some amount for newusers (e.g 5-20
outbound emails per day)
2. Continue with the status quo of keeping email turned off for
newusers and grant it only by request but with a change to the text of
the newuser program reflecting this policy
3. Grant email to members and former members so that email would
still be mostly free. A person could become a member for one month for
$6 and then allow membership to lapse but still have outbound email
privileges
There was discussion where the board decided that we should move in the
direction of granting everyone email but with limits on outgoing email
of perhaps five per day at first and then an increased limit after a
period of time.
MOTION by Mark Conger: We will continue with the present email policy
until such time as we can implement a policy where users are subject to
an adjustable email per day limit. Seconded by Lynne Fremont. Passed
unanimously.
The modems are still not answering sometimes. They are working now
though.
OLD BUSINESS:
Lynne wants to extend the deadline of the web contest.
MOTION by Mark Conger to give Lynne Fremont authority over web contest.
John seconds. Passes (all in favor, Lynne abstains)
Set date for next meeting:
Tuesday Sept 26th at 8p at John Remmers s House
NEW BUSINESS:
Nharmon would like permission to submit the grex logo to wikipedia
MOTION by Mark Conger to allow grex s logo to be used on wikipedia. Joe
Gelinas seconds. Passed unanimously
Keesan would like a script that would set up Spam Assassin for new
users. Jan says that Spam Assassin takes up a lot of resources so if we
set it up for all users, we would need to make it more efficient. This
would be a good project for someone to work on. Someone does not
necessarily need to have root access to work on that issue so anyone
who wants to work on Spam Assassin is welcome to work on it even if
they are not officially staff.
Meeting adjourned 8:36p
|
| 30 responses total. |
scholar
|
|
response 1 of 30:
|
Aug 9 00:43 UTC 2006 |
Wikipedia still won't let you use the image on Wikipedia, though.
|
cross
|
|
response 2 of 30:
|
Aug 28 04:51 UTC 2006 |
I wonder why people think spamassassin would be so hard on grex's resources?
It seems pretty clear that relatively few people actually *use* grex for
email, and certainly, a global spam daemon would be more efficient than n
invocations for n different users. I'd strongly recommend against modifying
something to make it "more efficient" unless (a) strongly committed to getting
those modifications back into the project's mainline source base, or (b)
measurements clearly indicate a need for such modification. I'm not sure grex
is either.
|
ric
|
|
response 3 of 30:
|
Aug 28 19:32 UTC 2006 |
relatively few people USE grex for email.
That doesn't mean a horrific ton of email doesn't come to Grex accounts.
spamd still spawns processes to scan each mail, even though it's running as
a daemon. running spamd globally will almost certainly be a significant
performance drain.
|
keesan
|
|
response 4 of 30:
|
Aug 28 19:46 UTC 2006 |
Could people be allowed to specify that they do NOT want to receive any mail
at grex and would this save some cpu time? Many people have abandoned their
accounts to the spammers.
|
steve
|
|
response 5 of 30:
|
Aug 28 20:03 UTC 2006 |
I'm not so sure about that Rick. I see lots of people using
email here, and I think its on the rise.
|
cross
|
|
response 6 of 30:
|
Aug 29 03:37 UTC 2006 |
Define "lots". Certainly, many orders of magnitude fewer people than use,
say, hotmail or gmail or yahoo or whatever. I'm not sure that running spamd
would be that noticable on this hardware. Of course, there's also the option
of moving mail to another machine which *does* run spamd. Come to think of
it, this hardware is getting a bit long in the tooth. Maybe it's time to
upgrade to a newer machine, move the users to that, and move mail to this
machine, putting both in rackmount cases. We'd be taking up less room at
provide.net than we presently are.
|
steve
|
|
response 7 of 30:
|
Sep 4 04:20 UTC 2006 |
I've been trying to come up with some numbers on mail usage. I'd say
at least 1000 users are routinely using Grex for mail. It may well be
higher.
|
cross
|
|
response 8 of 30:
|
Sep 4 22:13 UTC 2006 |
I wonder how many resources setting up spamassassin in a daemon configuration
for about 1,000 users would require. I doubt it would be that many.
|
remmers
|
|
response 9 of 30:
|
Sep 5 17:45 UTC 2006 |
Well, we're currently running spamd, the daemonized version of
spamassassin. Also, spamc and procmail are available to users. So
we're certainly in a position to test this.
Yesterday, I set up a procmail/spamc configuration for myself. It's
managed to catch about 90% of my spam, with no false positives, in the
last 24 hours. I'm planning to enter an item in Agora describing how to
do this.
If system performance goes down the tubes then, well, we'll have to
rethink our spam-handling strategy.
|
tod
|
|
response 10 of 30:
|
Sep 5 18:03 UTC 2006 |
I have a .forward pointing to my gmail account and it works beautifully at
detecting spam.
|
nharmon
|
|
response 11 of 30:
|
Sep 5 18:57 UTC 2006 |
What is the difference between allowing people to have POP access to
their e-mail and allowing them to .forward their mail?
|
tod
|
|
response 12 of 30:
|
Sep 5 19:02 UTC 2006 |
re #11
What is the difference between allowing people to have POP access to
their e-mail and allowing them to .forward their mail?
1.Storage locally vs offsite
2.Nodes in use
3.The .forward is available to all (right?)
4.It opens another service for possible exploit
5.POP3 complaints may overburden the overburdened existing staff
|
nharmon
|
|
response 13 of 30:
|
Sep 5 19:10 UTC 2006 |
So in regards to bandwidth use (which seems to be the only reason to
stop POP access)...nothing?
|
tod
|
|
response 14 of 30:
|
Sep 5 19:11 UTC 2006 |
Bandwidth issues could arise from attachments. I've seen POP hang on systems
where people have over a thousand emails some of which have over a Meg of data
file attachments. It really depends on the rules that are put on mailboxes
and the amount of folks accessing POP at once.
|
mcnally
|
|
response 15 of 30:
|
Sep 5 20:03 UTC 2006 |
One of the problems with running a POP or IMAP service is that in
order for such a thing to be useful it more or less implies that we
will set up a corresponding facility to allow clients to *send* mail
from their remote locations.
Securing such a service against e-mail abuse requires FAR more staff
time and attention than just setting up a POP or IMAP service and
staff time is in very short supply as it is.
|
kingjon
|
|
response 16 of 30:
|
Sep 5 20:07 UTC 2006 |
If members were given POP access I would never use it, except perhaps to
alleviate an over-full mailbox during a period in which I had no access to a
SSH or telnet client and was outside the area code. If members were given IMAP
access I would use it only if I had no access to a SSH or telnet client, if at
all (since my preferred mail client that handles IMAP is pine). My other email
accounts that I use these protocols on, I do so only because they don't offer
direct access and there's nothing on those servers for me besides email anyway.
(POP has stopped working for me on the one address I used it for -- GMail -- by
reporting every message as new every time; I use IMAP for my college-provided
mail and much prefer it to webmail because of its strange handling of
mailing-list digests.)
|
tod
|
|
response 17 of 30:
|
Sep 5 20:51 UTC 2006 |
I think POP might work if there was a way to not permit "keep copy on server
side" after its been read.
|
remmers
|
|
response 18 of 30:
|
Sep 6 13:19 UTC 2006 |
I was on the fence about this proposal until I read McNally's #15, which
sounds to me like a strong reason for *not* implementing it.
|
remmers
|
|
response 19 of 30:
|
Sep 6 13:32 UTC 2006 |
(Oops, thought I was reading the proposal item about POP access. Please
mentally transfer #18 and #15 to that discussion.)
|
naftee
|
|
response 20 of 30:
|
Sep 6 20:06 UTC 2006 |
okay, remmers. i've never heard the expression, "on the fence" before.
|
remmers
|
|
response 21 of 30:
|
Sep 8 12:19 UTC 2006 |
It means "undecided".
|
naftee
|
|
response 22 of 30:
|
Sep 8 17:52 UTC 2006 |
fanks.
|
nharmon
|
|
response 23 of 30:
|
Sep 8 19:25 UTC 2006 |
fanks? When did jerryr return to Grex?
|
naftee
|
|
response 24 of 30:
|
Sep 8 23:14 UTC 2006 |
he writes "fanx", nate
|