Re: cross-posted from Scott's HTTPAPI list) I need the SSL version [of HTTPAPIR4] to be backward-compatible with programs expecting the non-SSL version.

On Dec 28, 2011, at 3:54 PM, midrange-l-request@xxxxxxxxxxxx wrote:

It also means that the lousy engineering is in the basic architecture of
the ILE *SRVPGM: "call-by-position" instead of "call-by-name."

Well since I worked for IBM at the time this was done and involved in the design I'm afraid I can't agree with that. The driving force behind ILE was to boost performance in highly modular applications. One of the things that we needed to do was to reduce the path length for calls (hence the introduction of bound calls). With service programs the best way to make that work was by resolving the srvpgm once dynamically and then have it "hand-over" a simple list of pointers that the caller could index into. If you need something closer to the dynamic resolution of *pgms then you can always use the API route to activate the srvpgm and resolve the pointers yourself.

By contrast, the OPM "pseudo-*SRVPGMs" (or maybe "proto-*SRVPGMs"?) I've
worked on, both in QuestView and in Wintouch, use a protocol that passes
an "opcode" parameter to indicate the desired function, making them the
next best thing to call-by-name.

And you can do the same in ILE by treating a PGM as a SRVPGM - but you need to architect it as you do with the API approach.

Even if there had been some overriding reason for call-by-position
instead of call-by-name, given that the OS is famous for fully
transparent self-re-encapsulation of programs through the CISC-RISC and
V5-V6 transitions, I would expect any sane implementation of *PRV to
provide some sort of automatic conversion.

There was as I have already noted. In the interim period hardware price-performance has reached a point where I suspect if the design was being done today they might well go for a name lookup approach. But that is hindsight.

As to the "sanity" - well *PRV was only ever intended to provide signature matches - it was always your responsibility to make sure that the procedure sequence matched.


Jon Paris

www.partner400.com
www.SystemiDeveloper.com