[rc5] FYI - Possible PerProxy problem

Marc Sissom msissom at dnaent.com
Mon Nov 10 14:31:19 EST 1997

jpearl at qsi.com wrote:
> Okay - here's the story...2509 proxy problems described...
> Something smelled....

Sounds familiar. To put it simply, the 2509 w32 proxy isn't a
proxy at all. It's a block eater. It even puts out a log of
all the blocks that it has eaten! How dutiful it is. It was
never been announced as a beta, and you made the same mistake
that I did, by assuming that since it was made available, it
must have been at least partially checked. I imagine that it
was run a couple of times and when they found that it did not
crash, they turned it loose upon the rest of us for testing.
Similar to starting the engine and concluding that the car
works having never tested the brakes. Such a gigantic mess as
this ought to be announced. There are probably more unsuspecting
fools out there that are just cruising along thinking that they
are doing something worthwhile as their proxy eats their blocks.

Get the latest(2513 I think). It seems to work. Of course
version 2509 acted like it was working too. You never know.
At least the email addresses that had been completely
swallowed by 2509 have started to show up in the stats since
I installed 2513. So...at least it is no longer eating _all_
of the blocks. Without the source, I'll never be able to
know or sure.

> Anyway - it looks like I've lost about two and a half days worth of
> work at
> about 1 Million keys per second. Tonight I managed to dump my backed
> up
> blocks along with a hundred more or so - my fingers are crossed and

Yeah, it hurts when the proxy reports 4800 'notifications' after
a week's work, yet there's nothing in the stats. Especially it
hurts when you go to look at the mailing list archives and find
that the last post was in August. So you keep on feeding the
proxy more checked blocks figuring that there's just a lag in
the stats...

> I suppose this is a good time to mention how important those stats can
> be
> to those who don't care about them. I'm not bitching about them being
> "down" and I understand the problems - I'm just mentioning that
> without them I'd never have noticed this, thank you Nugget....

Indeed. Having some form of reliable verification of operation
is rather valuable in the face of unreliable software, and no

I think that in the future, I'll wait until build 13 comes out
before I bother downloading anything from d.n. At least by
then the fools that tried to run the first ten versions will
have discovered most of the serious bugs.


To unsubscribe, send email to majordomo at llamas.net with 'unsubscribe rc5' in the body.

More information about the rc5 mailing list