You are not logged in. Login Now
 0-24   25-49   50-74   75-99   100-124   125-149   150-174   175-177   
 
Author Message
25 new of 177 responses total.
ryan1
response 125 of 177: Mark Unseen   Sep 25 23:58 UTC 1996

Ahh, that's why it isn't working for me right now....
BTW, i *REALLY* hope that you allow me to use backtalk
in ALL the confs as soon as possible.  I find it a *LOT* more
convenient to use than picospan.  I will use the BBS a lot more now
that backtalk is installed.
tsty
response 126 of 177: Mark Unseen   Sep 26 07:36 UTC 1996

howse about, when backtalk responds to the massaging included in #124,
that 2-3 conferences get selected for a live experiment for a few months?
janc
response 127 of 177: Mark Unseen   Sep 26 20:09 UTC 1996

Backtalk has been turned on in two conferences for several weeks.
ladyevil
response 128 of 177: Mark Unseen   Sep 29 02:49 UTC 1996

I like this idea less and less, the more I think of backtalk allowwing links
to Sexuality II..
janc
response 129 of 177: Mark Unseen   Sep 29 03:21 UTC 1996

Backtalk was down while I made some improvements to it.  It's up again.  Still
limited to only the Backtalk and Backtalk2 conference.  If you want to try
it, point your web browser to:

   http://www.cyberspace.org/cgi-bin/bt

arthurp
response 130 of 177: Mark Unseen   Sep 29 06:24 UTC 1996

httpd can limit access, can't it?  Or do those can't connect errors happen
in a non-graceful way when a site get's 'swamped'.

I think it would be really neat to make this huge public store of information
called grex available to all those people out there who only understand web
stuff.
robh
response 131 of 177: Mark Unseen   Sep 29 09:26 UTC 1996

Generally, when a site gets overloaded, it just gives an
"Alert! Unable to connect to host" error message, at which point
you can just try again.
popcorn
response 132 of 177: Mark Unseen   Sep 29 17:51 UTC 1996

Re 128: Selena, I'm going to ask for a member vote on it.
davel
response 133 of 177: Mark Unseen   Sep 30 00:18 UTC 1996

Re ##130-131: a difference from what we're used to in Picospan, however, is
that the denial is on a request-by-request basis.  You could be reading an
item & suddenly be unable to connect for the next one, I'd think - and then
be able when you tried again.

If I understood comments by Steve (or STeve or Steve) a year or so ago, at
that time httpd ran with a fairly low priority, or some such, so that web
accesses were apt to be slower than telnet.  Am I remembering right?  And if
so, is that still the case?  (I don't have a good web browser, so I haven't
tried Backtalk - that might have avoided this question.)
janc
response 134 of 177: Mark Unseen   Sep 30 00:27 UTC 1996

Right, denial would be on a request-by-request basis, which could really be
annoying.  For example, you might compose your brilliant response, hit the
POST button, and get a failure to connect.  Of course, you can keep trying
to hit the button.

I don't know anything about httpd priority.

You can use Backtalk from Grex.  Just do:

   !lynx http://www.cyberspace.org/cgi-bin/bt

Be sure to select the vanilla interface when using lynx.  This might even
work for non-members.
srw
response 135 of 177: Mark Unseen   Sep 30 00:46 UTC 1996

Telnet packets are given priority over non-interactive packets. Http packets
are considered non-interactive. this is done in the pppd code in gryps. This
used to be a huge problem for web usage on grex, when we had the MTU set
improperly (too small, I think) and were fragmenting way too many packets.

This problem was resolved a long time ago. maybe a year ago. Anyway the
priority distinctions made by pppd no longer seem to cause any difficulties
as far as I can tell.
davel
response 136 of 177: Mark Unseen   Sep 30 13:22 UTC 1996

(See, I *told* you Steve knew about it.)
dang
response 137 of 177: Mark Unseen   Sep 30 22:02 UTC 1996

If I had to guess about any given staff member saying anything, I'd guess
Steve tooo... Higher chance of being right by sheer numbers. :)
tsty
response 138 of 177: Mark Unseen   Oct 1 06:29 UTC 1996

>neither of the two testing conferences was either large (proportionately)
>or had many participants (proportionatley).
>
>the 'live test' suggested above was for two existing, long term, gotta
>lot of users and bunches of files conferences. a scaling test if you will.
>
>... adn did you fix the bug that i reported with responses being
>listed in duplicates here and there?
>.
janc
response 139 of 177: Mark Unseen   Oct 1 17:00 UTC 1996

Neither of the two conferences are very active, but there was a two-second
period during with 90 responses were entered via Backtalk and 50 responses
were entered via Picospan, all to the same item.  Worked fine.  I'm pretty
convinced Backtalk won't screw up Picospan.

I haven't seen that bug report.  Note that both conferences are *full* of
linked items, including, I think, items linked to other items in the same
conference.  Linked items have lots of interesting issues, so I've been
testing them a lot.  You will see lots of stuff twice.

Yes, Backtalk surely still has bugs.  So does Picospan.  At some point you
have to go forward.  I don't see what we will learn from a partial step.
adbarr
response 140 of 177: Mark Unseen   Oct 2 16:17 UTC 1996

We can only learn by doing. We don't know all and will not approach that goal
until we do something. Bugs is bugs -- big deal -- bugs beget solutions and
growth in capabilities.  Maybe all all the electrons in the universe will lose
spin and fall to (whqt) and your boom box will not work. So we find
alternatives.  Janc -- just do it!
adbarr
response 141 of 177: Mark Unseen   Oct 4 17:40 UTC 1996

I certainly agree with #140 -- well said!
ladyevil
response 142 of 177: Mark Unseen   Oct 7 20:47 UTC 1996

Psst- Arnold.. you're talking to yourself..
janc
response 143 of 177: Mark Unseen   Oct 8 14:53 UTC 1996

I've reconfigured Backtalk with anonymous reading turned off (pending that
membership vote thing).  This necessitated a change in URL that users wanting
to try Backtalk should use.  It is now:

        http://www.cyberspace.org/backtalk.html

Next step, opening it to all conferences.
janc
response 144 of 177: Mark Unseen   Oct 8 18:47 UTC 1996

OK, Backtalk is now fully on line.  You can use it to access any
conference at all.  In fact, I'm posting this response from Backtalk.

It's still got some problems here and there (recompiling to fix one
even as we speak), and there are still many missing features and many
segments of the interface need to be redesigned, but it on the whole I
think it is in pretty good order.  It's still going to be considered
"experimental" for a while, and I'm not yet going to put any pointers
to it on Grex's web page, but I encourage people to try it out and
comment.

Except, gawd, our net connection is slow.
popcorn
response 145 of 177: Mark Unseen   Oct 8 19:38 UTC 1996

Cool!  This is exciting!!
kerouac
response 146 of 177: Mark Unseen   Oct 8 21:20 UTC 1996

I got "requested URL not found"...is it down again or is the
address not http://www.cyberspace.org/cgi-bin/bt?
robh
response 147 of 177: Mark Unseen   Oct 8 21:42 UTC 1996

No, the URL has changed to:

        http://www.cyberspace.org/backtalk.html

which janc posted in Agora, and should probably have posted here too.  >8)
ryan1
response 148 of 177: Mark Unseen   Oct 8 22:51 UTC 1996

janc had already posted it in here
ladyevil
response 149 of 177: Mark Unseen   Oct 8 23:24 UTC 1996

*sighs* and so it starts..
 0-24   25-49   50-74   75-99   100-124   125-149   150-174   175-177   
Response Not Possible: You are Not Logged In
 

- Backtalk version 1.3.30 - Copyright 1996-2006, Jan Wolter and Steve Weiss