|
|
| Author |
Message |
| 25 new of 194 responses total. |
other
|
|
response 50 of 194:
|
Oct 7 22:27 UTC 1997 |
would staff implement a simple command which would output the time/date of
last backup, for users' info:
prompt> bakdate
Last Grex backup: Monday xx October 1997, 22:15 EDT
(for example)
thanks.
|
scott
|
|
response 51 of 194:
|
Oct 7 22:55 UTC 1997 |
Like the "phones" command? Sounds easy enough. We (staff) really need
to be badgered about doing regular backups.
|
orinoco
|
|
response 52 of 194:
|
Oct 8 00:40 UTC 1997 |
What is the phones command?
|
valerie
|
|
response 53 of 194:
|
Oct 8 03:06 UTC 1997 |
This response has been erased.
|
janc
|
|
response 54 of 194:
|
Oct 8 03:51 UTC 1997 |
The old dates are for /var/spool/mail, /suidbin, and /a. The first two don't
bother me. There is no reason they should *ever* be backed up. But surely
/a has been backed up in the last year.
|
janc
|
|
response 55 of 194:
|
Oct 8 03:53 UTC 1997 |
Or maybe not...is it possible that the October 5 backup was done with a backup
script that predated our changeover to the hierarchical home? Such a script
wouldn't back up /a, because /a didn't exist then.
|
janc
|
|
response 56 of 194:
|
Oct 8 03:54 UTC 1997 |
Nope, Valerie says /a was deliberately not backed up.
|
valerie
|
|
response 57 of 194:
|
Oct 8 15:01 UTC 1997 |
This response has been erased.
|
valerie
|
|
response 58 of 194:
|
Oct 10 19:57 UTC 1997 |
This response has been erased.
|
richard
|
|
response 59 of 194:
|
Oct 10 20:34 UTC 1997 |
#58...so even when Grex gets ISDN connections, it *still* wont seem
particularly faster until when/if Grex gets on the new computer?
*depression*
|
dpc
|
|
response 60 of 194:
|
Oct 10 21:00 UTC 1997 |
Thanx, valerie! You might consider further gradual reductions. Our
load average is now about 25, after your reduction and after we were
off-line to catch up on post-crash mail.
|
i
|
|
response 61 of 194:
|
Oct 11 13:27 UTC 1997 |
Another thanks. (Disclaimer: i normally dial in.) When Grex is really
slow, i spend a lot of my time logged on just waiting for it to respond.
If most other users experience this, then fewer people on the system should
*reduce* the length of the line (since many people can get through their
.cflist, e-mail, or whatever and get back off).
|
valerie
|
|
response 62 of 194:
|
Oct 11 13:45 UTC 1997 |
This response has been erased.
|
dpc
|
|
response 63 of 194:
|
Oct 11 22:09 UTC 1997 |
What M-Net does may (or may not) be of some help. We have 64 ptys.
We reserve the first 16 of them (p0-pf) for people who pay
(patrons/members). The next 16 (q0-qf) are for people who don't pay,
but who telnet in from Michigan. The final 32 (r0-sf) are for those
outside Michigan and who don't pay. These trunk down, of course.
The result is that while the final 32 are almost always jammed tight,
the first 32 are not, so that people who pay and those from inside
Michigan get on with no problem, and the reduced load means that M-Net
is downright sprightly!
Grex doesn't want to make distinctions between payers and
non-payers (members and non-members). However, it *might* be worth
distinguishing between those in Michigan and those who are not.
If Grex reserved the first 32 ptys for those inside Michigan, we
might be able to kiss our lag problems good-bye!
|
tsty
|
|
response 64 of 194:
|
Oct 12 02:32 UTC 1997 |
again... because this goes to someone special...
mucho thankxxx to mdw for the stitch-n-fixit work on /etc/passwd.
like ......./phew! close call, y'all!
|
orinoco
|
|
response 65 of 194:
|
Oct 12 20:17 UTC 1997 |
But do we really want to discriminate thusly against non-Michiganders? I kind
of like having people telnetting in from all over the place. It adds flavour.
|
dpc
|
|
response 66 of 194:
|
Oct 12 23:36 UTC 1997 |
A reasonable distinction is not a bad thing, especially since we're
going to continue to be jammed unless we do something.
|
atticus
|
|
response 67 of 194:
|
Oct 13 02:43 UTC 1997 |
Reads like a line from "Mein Kampf" ;-)
Seriously, I think M-Net can afford that type of distinction because I
guess "M-Net" stands for "Michigan Net". I take grex to be a more global
concept -- a global meeting place. And I would think the real fun comes
from meeting so many people from all around the world.
|
valerie
|
|
response 68 of 194:
|
Oct 13 02:50 UTC 1997 |
This response has been erased.
|
scg
|
|
response 69 of 194:
|
Oct 13 03:27 UTC 1997 |
I don't like the idea of placing that kinds of restrictions on Grex either.
That sort of thing is one of many reasons why I often think the M-Net people
have their heads on backwards.
|
omni
|
|
response 70 of 194:
|
Oct 13 04:11 UTC 1997 |
I'm with scg. I don't like class distinction. People are people, as long
as they have red blood and 1 head, I don't care.
|
krj
|
|
response 71 of 194:
|
Oct 13 05:24 UTC 1997 |
Grex and M-net are like different states in the Federalist model of
American government; they should be free to explore different approches
to managing themselves.
scg in #69: I can't agree that the favoritism policies of M-net
show that "the M-Net people have their heads on backwards."
With these policies in place, M-net has staunched the financial
hemmorage; more importantly, dpc has reported some real gains in
the number of financial contributors over the last year.
M-net has a long tradition of offering better access to its
contributors; Grex has a founding principle of offering the same access
to everyone.
It's been fascinating watching everyone's reaction to the Indian
community that has come to be Grex and M-net users; I include myself
in this, of course! Individually I really enjoy the interactions
with the Indian grexers who chat in party, and especially the ones
who get involved in the conferences. But collectively, I worry
that the Indian community -- and looking beyond India, to when the
Chinese get liberal network access -- has the potential to swamp
Grex's resources: in CPU, network bandwidth, and staff.
I don't believe that approximately 100 financial contributors,
and about 10 staffers, can provide free e-mail service to the world,
forever.
I'm selfish; I'm not willing to contribute funds to a system which gets
so swamped that I can't use it.
But I also know that my perspective is a minority one among Grex
voting members, and I'm not inclined to lead any campaigns here.
The upcoming 128K Internet link and the new fast Sun machine will
buy Grex time for a lot of non-swamped growth.
|
valerie
|
|
response 72 of 194:
|
Oct 13 15:29 UTC 1997 |
This response has been erased.
|
janc
|
|
response 73 of 194:
|
Oct 13 19:44 UTC 1997 |
"M-Net" probably doesn't stand for "Michigan Net". Mike Myers named "M-Net."
He was non-committal when I asked him once what the "M" stood for. But he
originally used the name for a small one-line bbs system running on an Atari
800 back around 1982 or 1983. He obviously didn't expect it to be serving
the University of Michigan or the whole state of Michigan, much less the
whole world. My belief is that it started out as a small personal project -
his own little BBS - so he named it after himself. As it got bigger, the
name seemed a little immodest. It wasn't by any means just his system
anymore, so he let the meaning of the "M" be vague.
|
remmers
|
|
response 74 of 194:
|
Oct 14 00:00 UTC 1997 |
(The name "M-Net" was coined by Jeff Pyonnen, a friend of Mike
Myers.)
|