|
Grex > Coop11 > #128: Be a part of a website, write columns (This is NOT Spam) | |
|
| Author |
Message |
| 25 new of 55 responses total. |
gull
|
|
response 14 of 55:
|
Oct 14 03:00 UTC 1999 |
Re #11: Maybe you should explain why you feel this is important? I'm not
sure what inactive conferences are harming.
|
devnull
|
|
response 15 of 55:
|
Oct 14 19:57 UTC 1999 |
I would be very unhappy if the writing conference went away. I suspect it
has some interesting material that I would enjoy reading if I ever find
the time to carefully read it...
|
dpc
|
|
response 16 of 55:
|
Oct 14 20:01 UTC 1999 |
One drawback of lots of inactive conferences is that it gives newbies
the impression that Grex is largely dead.
|
scott
|
|
response 17 of 55:
|
Oct 14 20:36 UTC 1999 |
Perhaps it also gives a sense of history and is a useful reference.
|
i
|
|
response 18 of 55:
|
Oct 14 22:08 UTC 1999 |
Getting more activity into more of grex's conferences sounds much
better to me than reaping the inactive ones.
|
don
|
|
response 19 of 55:
|
Oct 14 22:36 UTC 1999 |
re 14, 16 is largely correct -- I remember getting an account here for the
first time and thinking that grex was a ghost community due to the dead cf's.
They also wasted my time when I posted some stuff in them and kept coming back
to check for responses, until I realized that they were dead. Sure, the
conferences aren't hurting anything -- but there's almost no point in having
them around; we should just tar.gz them so that they'll be available for
anyone trying to catch up on "history".
|
devnull
|
|
response 20 of 55:
|
Oct 14 23:48 UTC 1999 |
Re #19: picospan has a mechanism that lets you say `next', and the system
will automatically check conferences looking for new items until it finds
one
|
krj
|
|
response 21 of 55:
|
Oct 15 03:18 UTC 1999 |
resp:16 :: I've floated a proposal before that we find some way to
segregate the dead conferences so that newbies don't find them
unless they explicitly say: show me the dead conferences.
|
gull
|
|
response 22 of 55:
|
Oct 15 04:03 UTC 1999 |
Only problem I can see with #21 is it pretty well ensures any dead
conferences stay that way.
Could any newbie, having been plopped into Agora, really get the impression
that Grex is a dead, quiet place?
|
remmers
|
|
response 23 of 55:
|
Oct 15 22:09 UTC 1999 |
Being something of a minimalist when it comes to bureaucratic structure
and control, I'm not too fond of the idea of zapping conferences for
inactivity, adminstratively labeling conferences as "dead", or whatnot.
But I like putting power in the hands of the people, so what do folks
think about providing a tool that people could use to find active
conferences, if that's what they want to do? Technically, it's
feasible. Here's one way to do it.
(1) Create a dummy account called "jarjar" or something and have it join
every conference on Grex.
(2) Write a front-end shell script that pipes commands into Picospan
which has jarjar run through all the conferences, doing a "fixseen" on
each one. This will give jarjar a "caught-up" participation file for
every conference.
(3) Write a back-end shell script or program that captures the output
from Picospan produced by running the front-end script. This will
include all the "number of new responses and brand-new items" lines that
Picospan prints when you join conferences. Have the back-end save these
numbers in a file in some reasonable format and throw away everything
else, thus generating a report on activity in every conference.
(4) Running the pipeline
front-end | picospan | back-end
once a week from the jarjar account will produce a report on conference
activity in the last week. This could be done by a cron job. Have a
command that people could run which reads the weekly jarjar file and
displays a report of conference activity, with options to show them the
"five most active conferences", "ten most active conferences", "all
conferences", or whatever. Have a CGI that makes the command avaiable
via the web, for the benefit of Backtalk users.
Comments?
|
lilmo
|
|
response 24 of 55:
|
Oct 15 22:52 UTC 1999 |
You want it to run fixseen every time?
|
gull
|
|
response 25 of 55:
|
Oct 16 04:56 UTC 1999 |
You'd have to, since it's not really reading any items. Otherwise the
response counter doesn't go back to zero.
|
devnull
|
|
response 26 of 55:
|
Oct 16 17:47 UTC 1999 |
Re #23: I think it would also be possible to write a shell or perl script
which does not actually invoke picospan, and get such a thing to DTRT.
Figuring out the details is left as an excercise for the reader.
|
janc
|
|
response 27 of 55:
|
Oct 17 18:25 UTC 1999 |
I think such a program would be nice to have.
Picospan has a feature that would help a bit if more fairwitnesses used
it. It lets fairwitnesses configure which items should be new to new
users. In particular, you can say, for example,
first item, last item, item 17 and all items with postings since two
weeks ago.
So if a new person joins a conference and does "read new," they'll see
the named items, and all the items with recent activity, but not any of
the items that have been abandoned for years. So the deadwood can be
there, but mostly won't be seen.
This only works if the user first joins the conference with Backtalk.
Picospan doesn't know about these settings. It also only works if
conference fairwitnesses set them (last time I looked, there were only a
couple conference on Grex that used this feature).
|
remmers
|
|
response 28 of 55:
|
Oct 18 13:54 UTC 1999 |
Re resp:26 - It would be possible to write a program separate from
Picospan that would do it. But I'd prefer not to, since the program
would have to do the grunt work of detecting new items and responses
that Picospan already does for you. When feasible, I prefer to avoid
re-inventing the wheel. Such a program could be more efficient than the
one I proposed, but since the program would only be run once every few
days, I don't think the extra efficiency is worth the programming
effort.
Disclaimer: I don't have time to write my proposed program right now.
If anybody else wants to give it a try, by all means do.
|
dpc
|
|
response 29 of 55:
|
Oct 18 17:44 UTC 1999 |
I think this proposed program would be really helpful!
|
ryan
|
|
response 30 of 55:
|
Oct 18 17:55 UTC 1999 |
This response has been erased.
|
gull
|
|
response 31 of 55:
|
Oct 18 18:37 UTC 1999 |
It's not ours to tinker with. Notice the copyright and license information
when you run bbs.
It seems to be a common misconception that Grex owns Picospan. Grex
doesn't. Grex just has a license to use it.
|
don
|
|
response 32 of 55:
|
Oct 18 19:10 UTC 1999 |
However, Marcus has the source, and I doubt that the people who owns picospan
are gonna notice if we tinker with it.
|
mta
|
|
response 33 of 55:
|
Oct 18 21:03 UTC 1999 |
Not being noticed doesn't make a thing right.
(What's with this attitude that "it's only wrong if I get caught'. My kids
got over that at age 5.)
|
spooked
|
|
response 34 of 55:
|
Oct 19 00:10 UTC 1999 |
I will write the necessary program(s). I think one will need superuser
privileges to do so, though. I was supposed to probably gain them a while
back, but staff hasn't met in ages to confirm it - this is like 3 or 4 months
ago, when they're supposed to meet monthly.
|
i
|
|
response 35 of 55:
|
Oct 19 00:47 UTC 1999 |
Why would you need 2BUID0 to write a PicoActivityLevelMeter program?
|
mdw
|
|
response 36 of 55:
|
Oct 19 00:50 UTC 1999 |
Guido is at dogwood.nonlineardynamics.com .
|
lilmo
|
|
response 37 of 55:
|
Oct 19 02:21 UTC 1999 |
re 35 & 36: What?
|
scg
|
|
response 38 of 55:
|
Oct 19 05:39 UTC 1999 |
2BUID0 would presumably be, to be UID 0. UID 0 is root. I think the answer
to that question is that you wouldn't, but you would need to be UID 0 to
install it.
dogwood.nonlineardynamics.com doesn't seem to exist.
|