Re: Question about disk interleaves

Martijn van Buul <pino@xxxxxxxx> wrote:

> Regarding the mismatches: that's not necessarily true. It is true that it
> would be rather simple for anyone to get out of the killfile, but that's
> not particulary hard anyway. It kind of depends on which newsreader you
> use; most of them tend to completely ignore the RFC's about how they
> should be generated (822 in particular), and that's when the pain starts.

It wasn't avoiding killfiling I meant (because is is really hard to
killfile someone determined to get around it), but rather that merely
matching on the least three characters of the Message-ID sounds a bit
flakey ;-)

> And which won't really work either, as it would require information that is
> typically no longer available, and/or would make things really slow. For
> starters, you'd have to keep a local archive of message headers, otherwise
> trying to find out whether article X has a parent with author Y becomes
> unfeasable.

I'm not sure what information you're referring to? My client (tin) has
no problems displaying a threaded view of a thread such as this one, and
with dunric@xxxxxxxxx as the originator. As far as I can tell, this
would be all the info required to kill threads/sub-threads based on
author, and it is already there.

That said, I am on a high speed internet connection to the news server,
and I don't know how much info tin pulls each time. However, I also know
it can cache the same information for slower links, so I don't see the
big issue here. I imagine other clients can do the same.

Getting somewhat off-topic now, but all things considered, I'm not sure
I should feel bad about it ;-)


Regards,

Laust
.