Re: Question: Timed out blocks:

Dakidd (dakidd@mindspring.com)
Thu, 15 May 1997 09:56:31 -0400


>On Thu, 15 May 1997, Dakidd wrote:
>
>> Personally, if I had the code to play with, I'd go with the notion of
>> bringing up the UDP connection from scratch and then close it down
>> completely each time I went after a key-block or reported results to the
>> server.
>
>For various other reasons it was necessary to make that change for the
>MacOS client to be able to use multiprocessing, but you guys won't get to
>play with that version for a few days yet.
>
>As a temporary fix, I recommend getting a little program called
>"Mac TCP Watcher" from info-mac and setting it to continuously ping
>www.microsoft.com so your dialup connection appears to be active.

Point the first:
MacTCP Watcher is a great program ... *IF* you're running MacTCP. Under
Open Transport, I've found that launching it is one of the quickest ways
known to man to find yourself looking at the MacsBug screen or a bomb box.
When I was still running MacTCP, I found that it was an OUTSTANDING piece
of software.

Point the second:
I like NOT having to pay commercial rates for my ISP, thanks...
Part of the AUP for my "species" of account here at mindspring is "user
shall not employ programs to defeat the automatic idle timeout mechanisms."

I asked specifically about if/how that affects running DESChall in
conjunction with another program (Eudora was my personal preference) about
4 days ago, and got back the response that using any "program" other than a
live body actually using the connection is a violation.

Moving up to the class of account that would allow continuous connections
would seem to be the easiest way to cure the problem, except for one small
detail: $$$ or lack thereof...

>
>> Is there any provision to manually add a block to the server as having been
>> checked?
>
>There are nearly 10,000 active clients, and my gut feeling is that if
>everyone started requesting that it could become a really labor-intensive
>process for those on the server end, especially if they all sent in screen
>shots. Not to belittle your efforts, but that big 2^26 block is less than
>0.0000002% of the keyspace, so personally I wouldn't sweat over one lost
>block more or less. Rocke's opinion may vary. :)

<nod> I kinda figured that, but at the same time, I also figured "every
little bit helps" :) No biggie, just a rather disappointing loss of several
chunks of keyspace, and the time that was involved in checking them.

Speaking of screen shots... I keep seeing mention of log files and the like
from the non-Mac users. Is that ever going to be an option for us Mac
folks? Aside from a screen-shot, I can't figure out any other way to pass
along dropped results to the server. For "It ain't in this block" results,
that's no real biggie, since somebody else will just pick up the dropped
block at some other time. But what happens in the case of a Mac user
getting the "I found the key!" (or however it's going to be worded) message
and then not being able to get back in touch with the server? Does the
(not-so)lucky person who had the block the key was in get to just sit back
and cuss while somebody else picks up the dropped block, searches it, and
claims the cash and glory?

Dakidd@mindspring.com +------------------------------+
+---------------------------------+ |Do you ever get the feel that |
|I will choose a path that's clear| or |the story's too damn real and |
|I will choose free will -Rush | |in the present tense? -J. Tull|
+---------------------------------+ +------------------------------+