|
|
| Author |
Message |
| 25 new of 82 responses total. |
jp2
|
|
response 25 of 82:
| 
Aug 1 00:08 UTC 2002 |
This response has been erased.
|
bru
|
|
response 26 of 82:
|
Aug 1 13:43 UTC 2002 |
|
jep
|
|
response 27 of 82:
|
Aug 1 16:14 UTC 2002 |
Backtalk is back!
|
valerie
|
|
response 28 of 82:
|
Aug 3 02:31 UTC 2002 |
This response has been erased.
|
jep
|
|
response 29 of 82:
|
Aug 3 02:47 UTC 2002 |
Quite a run of bad luck. It's strange how it comes in waves.
Thanks for getting it back up and running smoothly!
|
tsty
|
|
response 30 of 82:
|
Aug 3 13:49 UTC 2002 |
see? tha's what happens when you get t turned-on donkey turned off.
hurrumpg!
<g>.
thankx to the staff, for sure.
|
jkim
|
|
response 31 of 82:
|
Aug 8 20:00 UTC 2002 |
test
|
janc
|
|
response 32 of 82:
|
Sep 11 04:20 UTC 2002 |
I'm currently working on upgrading Grex's httpd to version 2.0.40. This
is a bigger change than we've made for a while. Like almost all new
software, making this newest Apache work on Grex's old computer is a
challenge.
While I'm testing, there may be brief interuptions to web service, as I
swap in the new server, test something on it, and swap it out again.
Usually these will last less than a minute, and retrying your query will
work.
Unfortunately our development machine broke, so I don't have any good
alternatives to doing this on Grex.
|
janc
|
|
response 33 of 82:
|
Sep 11 07:02 UTC 2002 |
I abandoned the 2.0.40 effort. Write up another piece of softare that can't
easily be installed on SunOS. I installed 1.3.26 instead. This should be
fine. If you notice any changes in the behavior of Grex's web interface, let
me know. There shouldn't be any.
|
mdw
|
|
response 34 of 82:
|
Sep 11 08:27 UTC 2002 |
I was wondering if 2.0.40 is usable. "40" is a pretty big number...
|
gull
|
|
response 35 of 82:
|
Sep 11 14:45 UTC 2002 |
There was an article recently about how very few sites have gone to the
Apache 2.x branch, both because of a lack of module support and because
there just aren't that many advantages to it. The main selling point, IIRC,
is thread support.
2.x is usable but I'm not sure it's worth the trouble yet. I plan to stick
with 1.3.x where I work as long as it's being maintained.
|
janc
|
|
response 36 of 82:
|
Sep 11 19:52 UTC 2002 |
I'm sure 2.0.40 is usable - on Linux, FreeBSD, OpenBSD, Solaris, or
Windows. The Apache Group is pushing it as the version everyone ought
to be using. However, they are somewhat deluded. They haven't written
any documentation for it, hardly. The installation documentation
mentions only the most popular configure options. I had to read the
CHANGELOG to find any hint about how to link in a module not written by
Apache. There document for "how to write your own modules" has not been
updated for 2.0.x. They still distribute the 1.3.x version with 2.0.x,
and there are substantial differences. To port mod_auth_external to it,
we basically had to read source code. So is it surprising that not many
modules have been ported. Why they think they can call a version that
they haven't bothered to document yet "the one folks should be using" is
a mystery to me.
|
janc
|
|
response 37 of 82:
|
Sep 11 20:03 UTC 2002 |
OK, I've installed Backtalk version 1.2.2, the latest and greatest.
I haven't actually gone around testing much yet, but what's new includes:
- The abalone interface flavor has been heavily upgraded. Lots more
commands, including fairwitness commands. Many bug fixes. The goal
is to make it as complete an interface as pistachio, and we're
close. It's missing the preview/spell check function, and not
much else.
Abalone now makes fairly heavy use of Javascript. If you've got it
turned off (not entirely a stupid thing to do), it should still
function OK, though not quite as nicely.
- The papaya interface has been added. This semi-clones M-Net's old
Web-Yapp interface. It was done for the M-Net users who miss Web
Yapp after M-Net changed to Backtalk (apparantly there is exactly
one such person - there's always one). Might as well put it here
to, to make old M-Netters feel welcome. It actually has a couple
good features in it's own right. It's isn't quite as full featured
as some of the other interfaces.
- Lots of internal improvements. Whole subsystems have been
rewritten for better preformance and more flexibility.
|
janc
|
|
response 38 of 82:
|
Sep 11 20:05 UTC 2002 |
Oh, another almost noticable change - reloading after posting an item or
response should no longer attempt to post another copy.
|
janc
|
|
response 39 of 82:
|
Sep 11 20:28 UTC 2002 |
Opps, big bug. The code to list hotlist conferences, which is shared by
pistachio, abalone, and papaya doesn't work. I've temporarily disabled
that function in pistachio, so it's at least usable. Abalone and papaya
will be completely broken until I figure this one out.
|
janc
|
|
response 40 of 82:
|
Sep 11 23:44 UTC 2002 |
OK, all five interfaces should be up and running.
|
keesan
|
|
response 41 of 82:
|
Sep 11 23:54 UTC 2002 |
Jan, now that you have finished this project, any chance you or anyone else
could fix the problem that makes the latest version of lynx display the same
page five times in a row (it 'blinks')? On a 14.4 modem and 486 computer it
is tolerable, but we recently set up someone with an XT and 2400 bps modem
and it takes about 10-15 seconds wait every time you load a new page - it has
to draw it all, painfully slowly, up to 5 times. Longer pages take longer
(one it is all loaded you can page around within a 15 page document okay).
I do, of course, appreciate having the latest version of lynx installed.
SOmeone said it had something to do with curses libraries. ?
|
jhudson
|
|
response 42 of 82:
|
Sep 12 00:09 UTC 2002 |
An idea for swapping the test web server: run it on port 81
|
janc
|
|
response 43 of 82:
|
Sep 12 00:23 UTC 2002 |
That'd probably be useful, but you'd have to change quite a bit of other
stuff in the configuration. The most useful thing would to get the
development machine to boot again, so I can test it in the same
configuration that it would be installed in.
I haven't the faintest idea why lynx is being idiotic.
|
keesan
|
|
response 44 of 82:
|
Sep 12 00:35 UTC 2002 |
It is most likely doing exactly what it was told to do and somewhere it was
told to stutter. It has been like that since it was installed. People with
fast connections and new computers don't seem to notice.
|
other
|
|
response 45 of 82:
|
Sep 12 04:06 UTC 2002 |
I have a request for the next Backtalk development cycle. I think you've
developed a very nice product, and I make extensive use of it, but I have
written my own interface for several features, and it would make things a
little easier if on the "Entrance Page" you included an anchor tag just
before the "Active Items" header for those who select the option of
listing conference activity for their hotlisted conferences on the
entrance page.
That way, I can load the page and jump directly to the active list. (I
do it now with a javascript command that is specific to the size of my
hotlist.)
|
tpryan
|
|
response 46 of 82:
|
Sep 12 19:02 UTC 2002 |
The Backtalk thing I would like is a companion product that
takes usenet messages on my PC and organizes them to look like the
conference, item, response structure we all like.
|
janc
|
|
response 47 of 82:
|
Sep 13 15:49 UTC 2002 |
resp:45: Hmmm...that's a pretty Eric-Bassey-specific feature. It's
trivial to implement, but is it worth sending 10 extra bytes of data to
every backtalk user on the planet when only Eric-Bassey will know
they're there or have any use for them? Possibly, but not definately.
I'd be curious to know more exactly what you are trying to do. Maybe we
can find a solution of more general utility.
resp:46: Hmmm...that's a large and difficult project, and one I'd never
use, since I don't read Usenet. Developing free software you don't want
to use yourself isn't a sane pass-time. You need to find a skilled
programmer who actually would like to have such a thing. Hmmm...maybe
tpryan?
|
other
|
|
response 48 of 82:
|
Sep 14 01:37 UTC 2002 |
Heh, I guess you're right there. Sometimes a little perspective is a
good thing.
I have a small console window which has buttons for each of my hotlist
conferences, and for the entrance page. The entrance page button opens
another small window sized to show the active items part of that page,
and having an anchor at that part of the page would simplify the display
of the current items list in that window. It is a totally trivial thing,
especially since I already wrote a frameset with javascript to scroll the
page by a certain amount when it loads, bu I thought I'd ask anyway...
;)
|
russ
|
|
response 49 of 82:
|
Sep 14 13:05 UTC 2002 |
I dunno, I think I agree somewhat with other's request. I find that
anchors are very under-used in most large pages or complex, and
expenditure of a few static bytes would increase their utility quite
a bit. Compared to even one button image, an anchor is tiny.
|
