recently, ive just been reclused. part reason is that I am trying to make progress with my studies. I've seriously been tackling my core chapter in which I attempt to use Zed Notation to present a model through which one can asset, add and validate security within SIP messages. Mike Spivey wrote a Latex style sheet, and then added a fuzz type-checker for Z. My attempts have been to define the primitives and lay out the syntax for describing these messages, thereafter validating them using ztc. While this type checker is uesful it is a pain to force ones own logic to fit into this well formed set notation. There are some advantages to using Z though. In Z, functions are relations and relations are sets, so operators defined for sets also apply to relations and functions. One can put them all together in the same expressions....weird eh?

 

Found this rather critical blog on Singapore, and links to various other outspoken sites. Interesting how these sites seem extreme on the views of Singapores foreign policy, internal politics and harsh measures of control. What caught my attention was the comment on "economic progress and political development" which must be aligned to support continued growth. Here in South Africa we can similarly see (moderate) economic progress lead by a slowly maturing goverment. Economic development is not guranteed by an established goverment, but a country certainly requires modern policies to allow for this growth. Is it true though that the success that "Singapore [has achieved] cannot be realized without the suppression of rights and basic freedoms."? Perhaps various of the African countries should look to the firm governance of Asian countries which have recently strong growth? And further, the lack of contorl in Africa surely gives freedom but with it brings risk.

 

I presented IT Risk on Thursday at an Enterprise Risk Management workshop run by INSSA Multimedia Training (Pty) Ltd, and an interesting statistics (which can be found in the 2007 IT Web Salary Survey) was discussed. Considering what was said above, and the many opportunities in South Africa, it is strange to see that 29% of respondants are still looking to leave because of crime. Considering that job security only makes up 2%, while goverment policies 8%, the 5% who are leaving this year are certainly not seeing enough progress.

 

Currently trying to model how much cost is incurred through soft-dollaring of screens by brokers based on the commission revenue their derrive from the trading flow. I had a look at Orc Trader and found that they use the retail screen provided by The Online Trader. Easy, clear cut model and list prices for market data. Yay! Anyway, trying to get my head around introducing an internal FIX engine to allow various systems including risk management, settlement, allocations, order management, internal matching etc to talk the same language. The trick is not to call a spade a spade since "FIX engine" is taboo. Perhaps I'll call it middleware, order router or something similarly abstract.

 

My last concern at present is this definition of a megabyte!!! To my knowledge it had always been 1024 kilobytes, which in tern are 1024 bytes. Furthermore, the addition of bytes (or kilobytes) over a period should match the sum of all bytes (or kilobytes) over the same period. Thus A + B + C = SUM(A, B, C). It turns out however that mobile network operators here in South Africa dont actually bill on the sum of all traffic, but rather submit a CDR for each bandwidth increment. Further, considering that A != B != C, the CDRs do not accurately calculate back into consumption. This is for two reasons: 1) the bandwidth measured in bytes is converted into megabytes by roundup(bytes / 1024 / 1024, 2), and then further multiplied by the per megabyte charge of roundup(megabytes * 1.05,2). The double rounding certainly never matches the original amounts, let alone sum(bytes / 1024 / 1024) * 1.05. *sigh*