UFO Home Page

This page is devoted to Unicon and FOrman. For now, the page is to be considered private for Mikhail and Clint's use only. Feel free to bookmark it but please do not put any links to it up without discussing it with us.

Encyclopedia

This encyclopedia is a major piece of unfinished business. How does it relate to our current Case Study?

Phone call, 8/17/06

Our next project after the journal is submitted: Can we do for static analysis the same thing we have done for dynamic analysis? A library (or UFO-like language) for static checking, writing lint-style tools? Working from a syntax tree with semantic attributes, or a flow graph notation.

Categories/Dimensions for a taxonomy of what we look for

statically checkable
dynamically checkable

and

bugs
style rules, and specifications (compare with style rules for Java, C++, SmallTalk)

SuperInterpretation - use a terabyte of storage on a 64-bit machine to remember complete history of execution - who is responsible for each value that we might be looking at during an execution. How to ask the questions, how much will it cost, etc.

Security aspects for the case study - something simple, but useful. Maybe not just variable history, but how values were calculated

Aspects as a special case of our framework. Inserting any aspect you want, without modifying source code. Different types of print statements, combined/composed, etc.

Progress report, 4/15/06

My things-to-do list includes:

combine UFO publications to form a journal-draft

This means: add material from aadebug, softvis, woda to mw paper. A first pass at this is 25 pages coming from mw + aadebug.

select a suitable test case application

Part of this was to remember why we rejected igrep. I ran

     ./go hist.ufo igrep 
     and it seemed to run fine, with reasonable looking output,
     ending with "Total 69307 events detected".

 look into building rigal on amd64 architecture

Progress report towards bringing UFO back online, 4/7/06

Moved all my ufo-related directories into a web area for easier collaboration.
Compared Mikhail's 3/24/06 files against what I had "thought" was current. The goal was to not accidentally lose any features. Of course, most of these diffs simply reflect improvements Mikhail has made.

my generator.rig vs. Mik's generator.rig of 3/24/06 diffs to: generator.rig.diffs
After looking at the diffs, I am almost confident that work has not been lost here, but note that the version I had seemed to have a correction related to PARAM_NAMES. Mikhail, is this your take on it?
my parser.rig vs. Mik's parser.rig of 3/24/06 diffs to: parser.rig.diffs
The diffs towards the end (lines 400-800) look like its possible that the version I had might have a correction or a feature we need to keep, presumably relating to VALUE_AT_BEGIN and/or PARAM.

I reran

     ./go hist.ufo concord 
     in order to tell whether I could still run things.  It ran, ending
     with reasonable looking output, ending with "Total 46512 events detected".