🚀 go-pugleaf

What is great is my first proof of concept worked.  The guts of the work is proofed...searching the tables, resolving the relationships, creating the static values.  What "blows" is the loading of the tables.  My first iteration was...I hope your are seated...search the table, add a new record to the end, then sort it.  (10,000,000 entries, 3 tables, sorting...you do the math on how much wall clock was used.)  Well, the guts logic works, and it by itself is relatively efficient.

I'm going to reverse the table load so I can eliminate the sort.  But there will still be a lot of MVCL's.

After that, I think I'm going to experiment (on my own time) with the balanced tree logic.  I'm thinking macros.  That way I can eliminate the housekeeping overhead of a subprogram.  I have other applications that could benefit from this.

dvn

-----Original Message-----
From: IBM Mainframe Assembler List
[mailto:ASSEMBLER-LIST@LISTSERV.UGA.EDU]On Behalf Of Martin
Sent: Friday, September 30, 2005 1:13 PM
To: ASSEMBLER-LIST@LISTSERV.UGA.EDU
Subject: Re: Shifting memory to the right


Don,

Whatever you end up with- I am highly interested.

I have code that does move data in a dataspace from the lower end to the
upper end- If the areas overlap I do have extra codelength- If you come
up with a nice solution, pls let me know.

The way I do it now is the length of the overlap first to the end and
then the rest

BTW- I did not see the start of this thread- so I am missing parts (so
pls mail me once you are set).

--
Martin
--
XML2PDF - the way to get all features of PDF into your documents
on mainframe or PC systems; more at http://www.pi-sysprog.de