Hello Bob!

Jun 01 20:30 03, Bob Jones wrote to Bo Simonsen:

 >>> By point is why do we need a psedu terminal?

 BJ> I believe you caught why later in my message -- ability to handle all 
 BJ> nodes of maximus exactly the same way..... (wether local, modem or 
 BJ> telnet)....

Oh yeah that's right.

 BJ>> Technically, the (running software)  processes still exist, but they 
 BJ>> need to be killed because the physical connection has been broken or 
 BJ>> disconnected.

 >>> .. and i don't suppose mr. mgetty would like to do that.

 BJ> Possible for (m)getty to handle it if (m)getty is able (or modified) 
 BJ> to send signals to child processes based on changes of hardware 
 BJ> (status) lines....

Else a quick and durty solution would be somekind of monitor that looks if
the process is still detached to at pty.

 >>> Hmm.. ok, i've not been reading the whole Modern Operating Systems 
 >>> (Adrew S Tannenbaum).

 BJ> Mine is from experience, not text books.....

Ok :) I'm studying computer science and we are reading that book. It was
actually also somekind of joke ;)

 >>> else could get a fidonet call :)

 BJ> In older systems, getty was supplied as part of the operating system 
 BJ> routines.  Since mgetty is available with source code, it could be 
 BJ> compiled / ported to other unix based systems....

I guess so.

 BJ>> No!  This should be a configuration item.  Some BBS's change the +++ 
 BJ>> sequence to a differnet character.  And the +++ sequence also 
 BJ>> involves timing of a delay between certain characters.....

 >>> But you're scure about the modem get's the initstring.

 BJ> Modem initialization is only done if max is run in wait for caller 
 BJ> mode, or possibly after max issues a hang up of the modem....

Yes you're right.

 >>> under Linux, i really don't think that the auther of Sqafix would give 
 >>> us the sources.. Did anyone ask him?

 BJ> I don't think anyone has recently.  If Husky handles all of TICK's 
 BJ> functions, then getting Sqafix ported to a Linux / Unix platform 
 BJ> should solve most of the fidonet connectivity problems.  

Why not use the husky tosser, hpt? I really can't see problems about it, i
think squish will be useless with out a areafix..

hpt has smart functions like automatic area creation, forward requests and
lots of other goodies.

But maybe it should be up to people themself.

 BJ> The areafix 
 BJ> solutions I've seen on some of the other linux based platforms have 
 BJ> ignored some of the older fidonet standards resulting in 
 BJ> incompatibilities with raid / allfix / etc....  :(

Maybe crashmail areafix, it sucks very mutch..

 >>> Maximus's msgapi, it didn't wrote the XMSG structure correctly 
 >>> (subject, from, to, ftsc_data and so on). So i borrowed some code from 
 >>> the smapi.

 BJ> Interesting.....

Maybe it's only on unix.. but it works now..

Regards,
Bo

--- Msged/LNX 6.1.2