PPM Nugget: Huh?! Thats not the workflow the request is supposed to create!

Posted on January 3rd, 2008 by philip.wood.
Categories: Uncategorized.

If a request is set to only use a specific workflow and the first rule of the request creates a workflow of that type that does NOT mean that that workflow will get kicked of if the workflow is disabled!

After enabling it it should work.

[I was using HP Project and Portfolio Management - Version 7.1.0 SP4 UR7 when I wrote this]

(if you are a PPM user who found this - ignore it - it applies to configuring/administrating PPM)

0 comments.

Back to work

Posted on January 2nd, 2008 by philip.wood.
Categories: Uncategorized.

After a brief time of respite - a short holiday - it is back to work tomorrow.

It has been great.

But it is a new year to look forward to …

0 comments.

Database Thinking

Posted on November 10th, 2007 by philip.wood.
Categories: Uncategorized.

Recently I’ve been discussing (arguing) some ideas regarding databases and the relational model with some friends & colleagues:

First I was trying to browbeat Maurice Henry into accepting that the idea of denormalizing a data model for performance considerations is horribly wrong.

Then I was aurguing with Tristan Berg that OO databases were build on very shaky ground due to their lack of theoretical foundations. (And that the relational model is the best thing since sliced bread).

Lastly I’ve been talking to William Cairns about a seperation layer (preferably embedded in the runtime) that would map the conceptual representation of a language/program entity into a representation (memory or/and storage) for implementation.

I’ve started to look at the relational model, relational algebra, relational calculus again. Looking for extensions, redefinitions and tweaks mostly focussing on incorporating OO concepts

0 comments.

New Position

Posted on November 10th, 2007 by philip.wood.
Categories: Uncategorized.

I now have a new position.I’ve left BCX (which bought SiloFx) for EOH Gallium.

I’ve started last week.

This means that I’ve also left the previous client, Eskom, and am now based at VodaWorld working for Vodacom as client.

The new company has a webpage at http://www.eohgallium.co.za/

0 comments.

The wheel is usually re-invented

Posted on October 29th, 2007 by philip.wood.
Categories: Uncategorized.

Browsing tonight I found out about extended set theory, which uses an extension to the normal set theory [element of] concept, basically (as far as I gather) allowing set membership to be qualified by an “index”.

I’ve seen this constructed (rather clumsily usually) out of normal set theory as index sets and/or ordered sets.

Last night/yesterday I was trying to write a small set of c# classes to model this in the old way to express relations in classes as well…

well….

if someone else did it it must be a good idea! :)

TO go even further it looks like the “inventor” has gone ahead and done what I was trying to express to William Cairns as well… about using a formal model to describe data representations in memory to replace the usual clumsy representation model implemented in most runtimes and to extend this to storage systems.

0 comments.

Computer Languages as Notations to clarify and communicate knowledge

Posted on October 29th, 2007 by philip.wood.
Categories: Uncategorized.

One of the things that strikes me from the first few Structure and Interpretation of Computer Programs lectures, is the way that programs are used as vehicles to express and clarify thought, rather that just being used to just express working solutions to defined problems.

Programs expressed in a computer language can be very powerful notations that can express precise or general knowledge in an understandable, yet formally consistent way.

The parsing, compiling, running, etc. of these programs allows the formal verification of the syntax of the ideas expressed, and the actual results from execution to some extent allows the examination of the resulting semantics.
I’ve started writing some short C# programs to examine and define some of the ideas I’m investigating about using the relational model (and possible extensions to it) as a basis for the automated way of the structuring of the in-memory and in-storage representations of data structures and knowledge representations of normal programs and program designs.

0 comments.

Online Lectures

Posted on October 20th, 2007 by philip.wood.
Categories: Uncategorized.

I’ve been investigating functional languages to gain insight and intuition about their suitablity for the automatic/formal determination of what parts of a program can be performed concurrently.

Looking around I’ve stumbled on a collection of video lectures for a course presented by MIT in the 80s called “Structure and Interpretation of Computer Programs” hosted by archive.org at www.archive.org/details/mit_ocw_sicp.

A full set of 20 lectures for the course are available for download.

The course uses a lot of LISP and explains the concepts required.

A newer set of video lectures of a course based on a later revision of the book is also available and focuses on Scheme.

0 comments.

Functional Languages

Posted on October 13th, 2007 by philip.wood.
Categories: Uncategorized.

Reading about an 80 core Intel research processor (link) I’ve started thinking about how to enable programmers to use this kind of capacity. In the future k’s or m’s of cores will become available.

I think that functional languages - due to the ability of the compiler or “runtime” to do an analysis by basically only examining the stack - show promise. The evaluation of a functional language is basically the substitution and reduction of terms (where a term is a “primitive value” or a function involving combinations of other functions and “primitive values”).

These languages do not over-specify the sequence of program steps in the description of the program (like when a loop is used in an imperative language to perform the same operations on a number of elements).

They will also simplify runtime design since they have a strong mathematical basis in  lambda calculus.

0 comments.

Google Apps

Posted on October 12th, 2007 by philip.wood.
Categories: Uncategorized.

I’ve found out that you can have Google apps running on your own domain. So I’m trying to configure that on my site :). This has of course rekindled some interest in the site itself.

0 comments.

Low Activity

Posted on October 12th, 2007 by philip.wood.
Categories: Uncategorized.

There has not been a lot of activity on this blog, or even this site. I’m intending to see if I can change some of it….

0 comments.

Older »