Text 10203, 56 rader
Skriven 2007-06-07 20:20:42 av Maurice Kinal (1:140/13)
  Kommentar till text 10194 av mark lewis (1:3634/12)
Ärende: keeps on ticking
========================
Hey mark!

Jun 07 15:21 07, mark lewis wrote to Maurice Kinal:

 ml> ermmm... hint: message base formats are not fidonet technology...

Understood but the idea is the same.  Without something that understands the
format - squish in this case - then it is a useless format.  Originally the
idea was to give a web server that capability and as is it doesn't seeing
squish isn't a format recongized by either servers or clients.  I only
suggested that the server end serve up html rather than squish and that way any
capable client could render messages without any alterations.  Whether the
original be raw pkt's or squish bases is of absolutely no difference and if
most of the users are using web based applications then coverting raw FTN pkt's
to squish would be a complete waste of time seeing the other conversion would
be required anyhow no matter what.  Therefore operation on the raw pkt's into a
suitable storage format that could readily be converted on-the-fly to html via
a cgi or simular method would be more appropriate if cpu, memory, etc. are to
be kept to a minimum and speed of conversion maximized.  Why introduce another
storage format if it can be done by filtering the original raw inbound format
to a format more condusive to any other format?

 ml> are you stuck on the word "runtimes"??

No but it is the best since it covers more territory.

 ml> there are source code 
 ml> libraries available for most all message base formats used by systems 
 ml> in fidonet...

Yep.  At the moment I am using some of those but I doubt I need to replicate it
anywhere else.  If users wanted to use a different format then I'd be inclined
to drop the current strategy.  However as you say, what for?  It works
regardless that I am the only one willing to jump through this particular hoop.

 ml> take the TimED source code and compile it and the libraries it uses 
 ml> right on a mac running OSX...

Don't need to.  If I am going to bother at all, I'd give all this stuff the
heave ho and go to something completely different so that anything is possible.
 I started to look at that idea and already know that it could work if it ever
needed to work.  At the moment it doesn't.

 ml> once compiled, creating and 
 ml> manipulating squish bases is a breeze...

 ml> surely you can run make, right? ;)

Already did that ages ago.  It makes absolutely no difference.  It is and was a
dead end.

Life is good,
Maurice

--- Msged/LNX 6.2.0
 * Origin: Coffin Point BBS (1:140/13)