Re: [leafnode-list] Filtering revisited.

[krasel@xxxxxxxxxxxxxxxxxxxxxxxxxxxx (Cornelius Krasel)]

Why does leafnode have filtering capabilities?

Remember that I started patching leafnode to make it suitable for
my own needs. Now, I have a permanent ethernet connection to the
internet, so short connection times were not the reason to implement
filtering (and in fact, the first version which had filtering included -
which was, IIRC, 1.9 - filtered very slowly). No, the reason is that
I am a reader of the group alt.religion.scientology. This group is
under a constant spam attack with 50000-100000 spam messages posted
to the group per month. I needed a way to filter out these messages
so that they would not clog up space on my hard disk. (My newsspool
is only 100 MB which is more than enough for the groups that I read,
but if the spam gets through for a few consecutive days, it is full.)

While I agree that it would be nice to distribute the code to a
wider audience, I am somewhat reluctant to set up anonymous CVS
since I am not willing to support leafnode versions that are not
really released. I still get a fair number of queries about 1.10b2
which is quite out of date.

I agree with Mark Brown that the slowness of leafnode is due to the
lack of preloading. I have a version of leafnode (I call it 2.0a3)
which contains some not-yet-functional code which is supposed to
preload the connection with a number of commands (which can be modified,
default is a preload of 5 commands) which I would be willing to send to
interested persons via email. There is a second problem with fetchnews
which contributes to its slowness if you get the full grouplist, and
that is a design bug in the way the grouplist is built. I'll probably
address this second problem first since I have already developed some
similar code for handling the overview files which runs quite nicely.

2.0a3 also contains the rnews program. However, this one probably
should undergo a major rewrite so that it is able to use pipes
(at the moment, it creates temporary files because when I wrote it,
I didn't know how to handle pipes in C). As I said, if someone
wants to look at the code, I can send it to him/her.

With regard to Lloyd's proposal, I would prefer a version of Leafnode
in which I could hook cleanfeed which probably comes closest to point
3 on his list. However, so far I have been unable to exactly find out
how cleanfeed works :-)

I apologize that this message does not contain any quotes and is
somewhat long-winded, but I have not been here for the last few
days and didn't want to address every point separately.

--Cornelius.

-- 
/* Cornelius Krasel, U Wuerzburg, Dept. of Pharmacology, Versbacher Str. 9 */
/* D-97078 Wuerzburg, Germany   email: phak004@xxxxxxxxxxxxxxxxxxxxxx  SP4 */
/* "Science is the game we play with God to find out what His rules are."  */

-- 
leafnode-list@xxxxxxxxxxxxxxxxxxxxxxxxxxxx -- mailing list for leafnode
To unsubscribe, send mail with "unsubscribe" in the subject to the list