:: 10-Nov-1998 00:40 (Tuesday) ::

Public web statistics are now available:
http://www.distributed.net/webstats/

:: 09-Nov-1998 16:43 (Monday) ::

Just to clarify a few issues:

o Yeah, stats are still wonky. Yeah, I’m working on it. My wife’s
not too happy about it, but I am committed to getting stats back
online as soon as possible.

o The state of the stats server has no impact on the continued operation
of the keymaster/keyserver network. All work being performed now is
logged and retained, and there is no reason to “hold” your blocks nor
will you need to re-submit any blocks for credit.

:: 09-Nov-1998 09:24 (Monday) ::

Well, I’ve finally moved house! The relevance of this is that I now live
closer to work (where I do most of my coding/client building) so I’ll be
able to hang around and do that after work, rather than waste hours each
day driving.

I’ve had a few requests for RISC OS personal proxies, which I’ll hopefully
be able to fulfill – however my priorities at the moment look like this:
1. v2.7102.419 release (RISC OS)
2. v2.7102.419 release (ARMLinux + RiscBSD)
3. x86 card support (RISC OS)
4. personal proxy (RISC OS)

:: 08-Nov-1998 22:06 (Sunday) ::

I’ve been a busy beaver lately – too busy in fact to allow myself to be
frustrated by the packet loss problems I’ve been having.

The biggest changes I have implemented in the last days have been:
a) I drop-kicked the maximum number of CPUs/threads limit.
b) the client only runs one buffer per thread.
c) the client now displays the number of 2^28 blocks in the buffer files.

:: 08-Nov-1998 16:13 (Sunday) ::

There are obviously still some problems. I have an inbox full of
people who have noticed discrepancies in their stats as reported
yesterday so I’ve pulled the server offline and will be continuing to
hammer away on things. I’m guessing it’s still a space issue so I will
start by looking at ways to reduce the drive footprint of the
intermediate stats data that’s generated as a result of the daily
processing.

More details to follow as I make progress…

:: 07-Nov-1998 20:17 (Saturday) ::

For MacOS users without 8.5 or Apple’s beta themes, I present to you the
MacOS Client in its full glory, under four Appearances.

http://nodezero.distributed.net/~vetere/macos.html

:: 07-Nov-1998 18:44 (Saturday) ::

I’ve found /a/ problem, if not /the/ problem plaguing stats. It would
appear that the email index attached to the main stats table (tot_team:
[email],[team],[date],[blocks]) went wonky on us and grew to an obscene
size. I stripped off the index last night and rebuilt it and the
rebuilt index weighs in at less than 1/4 the size of the original
index. Anyway, that certainly went a long way towards freeing up space
in the rc564 database and I felt comfortable re-running the 5-Nov and
6-Nov stats.

The stats run began at about 09:00 and is not too far from wrapping up
as I type this. As luck would have it, I’ll be afk when it wraps so
in my infinite optimism I’ve just flagged stats to come back online
when they wrap without a chance for me to audit the numbers. I’m
sure if there are problems, my inbox will tell me when I get back
this afternoon.

So, keep an eye out for stats to come back online, and be gentle on me
if the numbers aren’t right.

Also, as a teaser, there’s some big big statsbox news in the near
future.

:: 07-Nov-1998 11:09 (Saturday) ::

The only way tcsh “rocks” is when the rocks are attached to its feet
in the deepest part of a very deep lake. (Linus Torvalds)

:: 06-Nov-1998 23:13 (Friday) ::

Well, here’s a frustrating one for you.

It appears that in the midst of doing the stats run, mssql decides
that it’s out of space. Now, I’m not entirely convinced that it’s
really running out of space, but the fact that it thinks it is is
sufficient for the statsrun to get hosed.

I’m running a consistency check on the database right now, which
takes a good bit to run. Basically, it’ll be a while before stats
are online again. Rest assured, I’m working on it and stats will be
back just as soon as it’s possible.

:: 06-Nov-1998 13:46 (Friday) ::

*grumble*

Thanks for your patience.