:: 19-Jul-1999 19:11 (Monday) ::

It’s been a day of boring yet productive coding on stats. I’ve
completely re-written the http_auth / cookie authorization code to make
its behavior much friendlier and predictable. Changes include:

o If user enters incorrect password, browser will re-challenge.
This will eliminate the annoying behavior which required you to
close your browser before re-trying.

o The http_auth (login box) will now accept either email/password or
ID#/password for authorization.

o If authentication from a cookie fails, the cookie is now cleared.
This corrects the previous problem which prevented proper use if
you’d saved a cookie for the login credentials of a now-retired
email address.

o Lots of nice fascist logging and tracking added to help identify
abuse.

Also on the docket for today will be to commit these changes to the
team coordinator authentication code.

I also repaired the problem that was preventing team coordinator
passwords from properly mailing. If you’ve had difficulty obtaining your
coordinator’s password, it should work now.

:: 14-Jul-1999 21:17 (Wednesday) ::

fetch@distributed.net and flush@distributed.net
can handle big blocks now, since the client behind
those scripts is build 443 now.

:: 09-Jul-1999 02:20 (Friday) ::

We experienced a lockup on statsbox this afternoon (~21:11 UTC). We’ve
been unable to determine a cause, and since Petr is out of town, there
isn’t anybody local to the box to investigate further. All we can
do is have the best.net staff reboot the box.

It rebooted fine, and there’s nothing too unusual in the logfiles,
so I’m just going through the databases. Running consistency and
sanity checks on the tables, and correcting the infamous sybase
“runaway identity” issue that occurs whenever there’s an abnormal
shutdown of the database.

The consistency check takes a good bit to run, and I’ve just disabled
the web server so the routines don’t have to fight for cpu. Needless
to say, the daily update hasn’t started yet, either. Figure on a late
finish for tonight’s stats run.

:: 08-Jul-1999 03:01 (Thursday) ::

Right after yesterday’s State of the Net was posted
(http://www.distributed.net/SOTN.html), several of you rushed off to
http://www.distributed.net/jobs/ in order to file for the currently
offered positions here at d.net.

Even though the submission cgi was tested shortly before the SOTN
post, it failed to properly record 6 submissions yesterday. The 6
missing submissions were filed on July 7, 1999 between 07:20 and 11:10
UTC. I sincerely apologize for this unexpected problem and would like
to ask the people who submitted their application during that 4 hour
lapse of time following the SOTN post to please take the time to
resubmit their application, so that it can effectively be processed.
2 submissions were from Italy, another 2 from Spain, 1 from the US and
the last one from Europe. If you have any doubts, please consider
reposting your application.

Once again I wish to apologize for that glitch that has now been
fixed, and would like to thank those who have volunteered so quickly
for their enthusiasm. I hope to welcome you aboard shortly!

If you have any questions, email me at jfc@distributed.net

:: 02-Jul-1999 16:56 (Friday) ::

Ok, The article in question appeared in a Swedish Internet World magazine.
Thanks to Rene Mosegaard for the scan, Hans Robinson, Karsten Spang, and
Andreas Ottosson for translations of the article.

Translation is as follows:

———————————————————-

THE WORLD’S BIGGEST BRAIN DRAWS ON YOUR COMPUTER’S IDLE TIME

The more we are together the merrier we will be. And the more computers
working together, the more complex problems we are able to solve. This is
the simple idea behind the technically very complex and smart project
known as Distributed.net – non-profit research project that could also
earn you a bit of money if you are in luck.

It is Distributed.net who succeeded in cracking hard encryption algorithms,
generating attention around the world, demonstrating that many small computers
together can be many thousand times more efficient than a single supercomputer.

It’s easy to join. All you have to do is visit the homesite, download a
small program and start it up. As soon as you are connected to the
Internet – the program automatically contacts Distributed.net. Depending on
the power of your computer you are given a bigger or smaller amount of data
to work on. As soon as the computer has received the task, the actual computing
starts off without needing to be constantly connected. When the task is done,
it is sent back to Distributed.net as soon as you connect to the Internet.

The program can be configured after your liking, either you choose when
tasks should be fetched, or you tell the program to automatically connect as
soon as a task has been done.

Since Distributed.net accepts challenges, in particular from RSA – a
company that develops encryption algorithms and proffers rewards to anyone
who can break their codes – there is even the possibility to make some
money. If your computer is the one to crack the right key, you can presently
get $1000 for the trouble.

The interesting part though is that Distributed.net doesn’t try to
take over your computer completely, but runs silently in the background.
The idle time of your processor is used, that is to say the program runs only
during those times when your computer is doing nothing else. Over the course
of a workday there will be quite few such idle moments. Enough to solve the
most difficult of mathematical problems.

As they say at their website: “go forth and multiply, divide, rotate,
invert, and conquer. May no computer ever be idle again.” NM

——-
Thanks for the assistance!

:: 02-Jul-1999 05:59 (Friday) ::

http://stats.distributed.net/rc5-64/tmember.php3?team=1828 has a new feature now,
it should be pretty obvious just by looking at it ;)

I did change some code that’s used by other parts of the stats system. It looks
ok, but it is possible I broke something, so if something isn’t working now, just
let me know.

:: 01-Jul-1999 19:58 (Thursday) ::

No News lately. Seems people out there have gotten more honest with their
client installs. :)

Just wondering if someone could translate the following perhaps:

http://nodezero.distributed.net/~alde/jun15_01.jpg

:)

–Pete

:: 01-Jul-1999 17:13 (Thursday) ::

After 5 months of laziness, I’ve finally got round to building ARM Linux
and NetBSD 1.3 clients.

I guess they’ll become available sooner or later.