Thread View: comp.protocols.ibm
2 messages
2 total messages
Started by Jay Maynard
Mon, 25 Mar 1991 05:18
Re: Pnews comp.protocols.ibm
Author: Jay Maynard
Date: Mon, 25 Mar 1991 05:18
Date: Mon, 25 Mar 1991 05:18
33 lines
2371 bytes
2371 bytes
In article <9103250340.AA23380@lilac.berkeley.edu> BITNIC IBM-NETS List
<IBM-NETS%BITNIC.BITNET@lilac.berkeley.edu> writes:
>Is anyone else out there using Interlink's TCP/IP, is there a list for
>this implementation? Technically we've found it blows IBM's MVS/XA
>implementation out of the water.
We're using it, but it could be better.
1) When they say to stick their 37x2 Ethernet box on a channel by itself,
BELIEVE THEM. It soaks up 45% of the channel, no matter what the traffic going
through it.
2) The documentation is unclear at best, and opaque at worst. I still have
problems with the DNR configuration, and can't seem to do an FTP GET to
save my life.
3) The mailer doesn't do MX records. This is nearly a fatal flaw, especially
when the DNR problems mean that perfectly good addresses with A records
fail at times as well - thus, a manual TSO DNRGET is required to determine
the cause of bounced mail.
4) The API is baroque. They modeled it after the VTAM interface...itself not
exactly a model of simplicity. The thought of writing an NNTP reader client
for it gives me cold chills.
5) The only mail user agent for it appears to be UCLAMAIL. Besides needing
an experienced systems programmer to customize and install, this introduces
the inefficiency of a TSO address space just for users whose only use of the
MVS system is mail - hardly the way to go. I'd like to use something like
MEMO, from Verimation, but that requires applications programming - see 4).
Yes, when it works, it's nice and fast. It's just a pain to deal with.
--
Jay Maynard, EMT-P, K5ZC, PP-ASEL | Never ascribe to malice that which can
jmaynard@thesis1.hsch.utexas.edu | adequately be explained by stupidity.
"You can even run GNUemacs under X-windows without paging if you allow
about 32MB per user." -- Bill Davidsen "Oink!" -- me
Re: Interlink's TCP API
Author: Walter Johnson
Date: Mon, 01 Apr 1991 18:47
Date: Mon, 01 Apr 1991 18:47
22 lines
1736 bytes
1736 bytes
In article <9103250530.AA24715@lilac.berkeley.edu> BITNIC IBM-NETS List <IBM-NETS%BITNIC.BITNET@lilac.berkeley.edu> writes: >4) The [Interlink] API is baroque. They modeled it after the VTAM interface The Interlink API (SNS/API) actually contains three different programming methods. The basic architecture is that defined for a Transport Interface in the OSI Reference Model. One interface provides a library of macro instruc- tions that manipulate data structures (TPLs) which are indeed modeled on VTAM RPLs. A second library provides services similar to the assembler interface, but is for programs written in C. This, too, uses the TPL data structures. Finally, a library of functions implementing a socket interface is provided for applications written in C. With this library, no knowledge of the "VTAM-like" TPL structures is required. In summary, if you like (or at least understand) VTAM, it's there for your use. If you prefer the Berkeley socket model, it too is there: it's your choice. -- Walt Johnson Business Development walt@interlink.com Interlink Computer Sciences sun!ntrlink!walt 47370 Fremont Blvd Phone (415) 657-9800 Fremont, CA 94538
Thread Navigation
This is a paginated view of messages in the thread with full content displayed inline.
Messages are displayed in chronological order, with the original post highlighted in green.
Use pagination controls to navigate through all messages in large threads.
Back to All Threads