suggestion for bionet feed discussion

[Reinhard Doelz]

Colleagues, 
rereading the ongoing debate on the bionet feed hickups please remember 
the following peculiarities of _bio_netters. 

     - apart from some few sites who can afford own staff, biologists 
              are usually 'guests' on central computing infrastructure. 
              This applies, in particular, to resource-intensive features
              like newsgroup moderation by hand (if , for some reason, 
              INN rejected creation, or ANU-NEWS never received a control
              message, etc.) Those who do manage their own site might get 
              annoyed by several peculiarities of USENET NEWS - I remember 
              the trauma well as I first installed C-NEWS and had to read
              bourne shells to find out how it worked. Conclusion: It might
              be that less experienced persons are getting trapped more 
              easily than others.  

     - traffic is, depending on the group, sporadic. We need to realize that
              'expiring' newsgroups might get 'forgotten'. 

     - unlike 'comp' the 'bionet' feed was considered to be 'second priority'
              at some sites - eg at our site, the central NEWS server never 
              wanted our feed of the 'updates' groups.  

     - there are two major sites relaying mail traffic; at Daresbury and 
              net.bio.net. Messages passing from and to there might take 
              longer due to the travelling of eMail. 

I watch the bionet.software.www spread with some interest and realized that, 
as we are the 'originator', it very much depends on the way how NEWS decides
to go. We exchange 'bionet' with several sites. Let us consider Daresbury 
and our central University NEWS server. The UK folks are usually first, and 
we get excellent timings from US-channeled bionet. If, however, we happen 
to feed 'YOGI' (our central system), the poor bionet message travels through
lots of 'slow' sites, and takes considerably longer to get in the US than 
we achieve to post it (up to three days for a single post seem possible). 

I must admit that bionet.software.www was NOT correctly created in automatic 
fashion here, and I had to create it manually. Checking the logs is a 
job which might be odd but a 'good' news administrator has to do this 
occasionally, and why not create the missing groups manually if you identify 
those easily from log files? All we need is a news administrator who is 
_interested_ in bionet. This seems to be the major problem! 

There was once the 'inpaths.c' program from Brian Reid, which measures the 
sites of traffic which participate most in relaying news. Daresbury UK here 
in the good old Europe is indeed a prime relay for us - and the UK is extreme-
ly helpful to help things get settled. I am convinced that the net.bio.net 
staff will also be interested in keeping the ball rolling. So why don't you 
check the paths occasionally and find out where the 'slow' sites are to get 
a better feed around? After all, NEWS is a _net_ and not a peer-to-peer 
tree which is statically implemented. Yes, I know, more work to be done than
a 'normal' hierarchy. However, we have created an own top-level hierarchy 
'embnet' and went through all this in our community - once you meet special
interests, be prepared to invest in special efforts. And it seems to me that 
biology is worth a try. 


Regards
Reinhard 
 
-- 
 R.Doelz         Klingelbergstr.70| Tel. x41 61 267 2247  Fax x41 61 267 2078|
 Biocomputing        CH 4056 Basel| electronic Mail    doelz at ubaclu.unibas.ch|
 Biozentrum der Universitaet Basel|-------------- Switzerland ---------------|
<a href=http://beta.embnet.unibas.ch/>EMBnet Switzerland:info at ch.embnet.org</a>