KR> Yes. There may be restrictions if you want to use hudson base but
 KR> there is no reason to use that stoneaged format. Hudson have a size
 KR> limit and it stores all messages in one database. If that is corrupt
 KR> then you will loose all mails. Todays formats use one database per
 KR> area and in case of trouble only one area is affected. 

I think it is cool to use stoneaged formats. I can reserve Hudson for inactive or unimportant echoes like BBS_ADS.

 KR> Packets do not need to be inspected/debugged. The bugs are in the
 KR> software that creates the packet so which software do you want to
 KR> debug? 

I want to check that packets contain the right messages and have been routed to the correct outbound directories. How do I check this without inspecting the packet ?

 MJ> Also there are area statistics generators that I would like
 MJ> to use that require message bases.

 KR> Hm, isn't that the rear of the horse? Usually you have an area inside
 KR> a message base and need a statistic tool then. The reverse way i have
 KR> statistic tool and need a message area now is very uncommon. 

It is not uncommon. If I have a hammer, then I know I need a nail. What is the point of having a hammer if not to hammer something.

 KR> Well, it was a long way from Blue Wave to node. What kind of other
 KR> software do you plan to use? I went from offline qwk readers to point
 KR> software and then to node software in point mode before i changed the
 KR> aka to node. 

A journery of a single step begins with a thousand miles.

I was thinking about running Fidonet on a Nintendo.

___ Blue Wave/386 v2.30
--- SBBSecho 3.13-Linux