<!DOCTYPE HTML PUBLIC "-//W3C//DTD HTML 3.2//EN">
<HTML>
<HEAD>
<META HTTP-EQUIV="Content-Type" CONTENT="text/html; charset=iso-8859-1">
<META NAME="Generator" CONTENT="MS Exchange Server version 6.5.7652.24">
<TITLE>RE: itaps-parallel Phone conf today</TITLE>
</HEAD>
<BODY>
<!-- Converted from text/plain format -->
<P><FONT SIZE=2>Sorry, I must have copied the wrong time into my calendar. My notes have 1:30P.<BR>
<BR>
Vitus<BR>
<BR>
-----Original Message-----<BR>
From: owner-itaps-parallel@mcs.anl.gov on behalf of Devine, Karen D.<BR>
Sent: Mon 1/7/2008 11:32 AM<BR>
To: itaps-parallel@mcs.anl.gov<BR>
Subject: itaps-parallel Phone conf today<BR>
<BR>
<BR>
My calendar shows that we have a phone conference scheduled for today at<BR>
1:30pm PST. Please let me know if you expect to miss this meeting (or if<BR>
you think my calendar is incorrect).<BR>
<BR>
There has been lots of good discussion since our last meeting. Please take<BR>
time to review it before the phone conference.<BR>
<BR>
Below are topics for discussion.<BR>
<BR>
- Tim asked that we clarify the questions we are addressing. He listed the<BR>
following:<BR>
<BR>
1) Do we want to do booleans on query results below the interface level?<BR>
An example query would be "What are all regions in some set in a<BR>
particular part?"<BR>
<BR>
2) How do we express and query sets spread across processors?<BR>
<BR>
3) How do we refer to parts and partitions in the interface, and which<BR>
functions do we use to get information about them?<BR>
<BR>
So far, our efforts have been addressing (3) and, more recently, (1). With<BR>
respect to (3), we have agreed to provide helper-functions that expose<BR>
partition and part information through the interface. These<BR>
helper-functions are natural for an application to use and allow the<BR>
greatest flexibility for the implementations. With respect to (1), we<BR>
identified the need to answer queries such as the example Tim provided.<BR>
Several options were proposed to answer these queries (see below). Tim<BR>
noted that similar capability could be useful for intersections of multiple<BR>
entity sets. We have not deeply explored question (2) yet, and unless there<BR>
is a reason to do so immediately, I'd like to resolve (3) and (1) before we<BR>
start into (2).<BR>
<BR>
- I'd like to try to reach agreement on a solution to question (1) above.<BR>
Our original idea of overloading the mesh instance argument is insufficient<BR>
for answering these queries, especially in a multiplexing environment. Carl<BR>
summarized three options; I'd like to discuss the pros/cons of each.<BR>
<BR>
> a) Adding args to about 10 functions, for both serial and parallel.<BR>
> Requires: universal agreement about what a null handle is, or else<BR>
> definition of an implementation-specific global variable that apps can<BR>
> use to find out what the null handle is for that impl. Also, for every<BR>
> call, the function has to identify which args it was passed and dispatch<BR>
> accordingly (creating more functions internal to the interface,<BR>
> probably, but not user-visible). Changes to serial code.<BR>
><BR>
> b) Leaving those 10 functions alone in serial, and adding parallel<BR>
> versions. Requires: adding probably one function for each of the<BR>
> originals in the parallel interface. My guess from the current document<BR>
> on parallel is that we'll be looking at >100 functions already, so this<BR>
> may or may not be considered significant. Serial code left untouched.<BR>
><BR>
> c) Adding some magic to squish three args into one in parallel.<BR>
> Requires: A function to squish (and unsquish, internally) the args;<BR>
> therefore, extra effort for the app and impl for each of those 10<BR>
> functions every time they're called (even in serial, probably).<BR>
> Depending on how the squishing is done, risk of handle collision.<BR>
> Changes to serial code.<BR>
<BR>
- Finally, I'd like to get clarity on how decisions are made in ITAPS.<BR>
When the iMesh interface was defined, for example, was unanimous agreement<BR>
needed on a function before it was accepted? Or was a vote taken (as I<BR>
believe CCA does)? <BR>
<BR>
<BR>
<BR>
<BR>
</FONT>
</P>
</BODY>
</HTML>