terry@terminus.UUCP (terry) (02/02/88)
> o I told him about 386 Xenix and that he could probably find a really nice > multi-user accounting package for it (this is the reason why he now needs > a multi-user system: on line order entry concurrent with accounts recievable > and inventory ...). Try Open Systems. It works under BBX from BASIS, Inc. > o I told him about smart multiport cards that will support up to eight > terminals per board. Beware of "smart" cards; SCO, when going from 2.1.3 to 2.2.0 changed the of the clist structs and unless you have right rev drivers and the right ROMs, they can clobber important stuff. > Does the Intel 386 Inboard work? Don't know a thing about it > Will Xenix386 run on it without a SINGLE hitch due to the hardware > configuration? I know that 386 Xenix runs flawlessly on a Compaque 386, Au contriare (or something spelled marginally simiarly); The Compaq 386/20 has some hardware incompatabilities, ESPECIALLY with intelliport boards. To be fair, they have a compatability problem with the 386/20; they aren't fast enough or there is a bus speed problem, or whatever will get me the least number of offended replies. The bottom line is, the bus speed is to fast for your average edge-triggered card. Contrary to popular belief, the 'ol IBM microchannel bus is not level-triggered to speed things up, but to allow slow hardware to keep talking to it. > Does VP/ix work (for Word Perfect and other DOS standards like Lotus > and such)? What happens when you try to run a DOS program that does > direct video RAM writes ??? Kablamo? The whole system goes down or > kernel panic or somthing? Potential nightmare? The VP/ix under the ISC UNIX (Bell Tech, et al) we had ran flight simulator. > Do these smart multiport boards and related drivers work reliably, without > maintainance (Computone, ARnet, Digiboard, Belltech) and constant downtime > of the hardware and the O.S.??? Pre-maintenance is required of at least the Computone, that I know of; it consists of changing the /atx/attype file as follows: from type * 7 E 1 9600 XON FIXPAR NOCHANGE to type * 8 1 9600 It seems that they didn't want the stty command or the ioctl() routine to work with it out of the box; also, binary I/O (read "TERM,XMODEM,UUCP") as lives in the G-protocol, among others, H.A.T.E.S. 7 data bits, for the obvios reason that 7 isn't binary to the programs. KERMIT tends to like it if you are annoyingly careful about parity on the other end... After that, tho, the Computone is a pretty nice board; especially if you have transparent printing in mind. +-----------------------------------------------------------------------------+ | Terry Lambert UUCP: ...!decvax!utah-cs!century!terry | | @ Century Software or : ...utah-cs!uplherc!sp7040!obie!terminus!terry | | SLC, Utah | | | | 'There are monkey boys in the facility. Do not be alarmed; you are secure' | +-----------------------------------------------------------------------------+