From hwyau@epcc.ed.ac.uk Tue Mar 18 17:27:46 1997 Date: Tue, 18 Mar 1997 16:46:49 +0000 (GMT) From: Hon W Yau To: Paul Coddington Subject: Re: financial modeling (fwd) Hello Paul, I've ried to phone you today regarding the matter below, but I guess you must be outside NPAC. So I'll try e-mail... On Mon, 17 Mar 1997, Paul Coddington wrote: > >Did you get tis email? Some mail I sent the same morning from this >machine didn't get through. Actually, it didn't make it across the Atlantic. Must have been discarded out by some site's obsenity filter :) >---------- Forwarded message ---------- >Date: Wed, 5 Mar 1997 13:35:53 -0500 (EST) >From: Paul Coddington >To: Hon Yau >Subject: financial modeling > >Geoffrey says he's happy to have you involved in the project (or projects - > the other one you mentioned is apparently going ahead also), but not until >you finish the HPFA papers! ;-) Man, he drives a hard bargain! Here I was, thinking I was doing *him* a favour!!! Let me say at this point that one may need something in the way of (ahem) `incentive'... As for the HPFA paper (sigh), tell GCF I hope to start work on it this week, beginning with the minor corrections. >Do you know what the status of the code was/is? Geoffrey thinks it was not >incorporated into the Sailfish program since Peter sold the company to >Reuters before the NPAC work was completed. Is this the case? Do you know Basically Geoffrey is right. My understanding is that our code did not receive their total attention because they were either i) busy moving offices, or ii) busy being gobbled up by Reuters. You should double check with Miloje, but I think we can say that we had completed NPAC's side of the work: Sailfish was supposed to go away and evaluate out deliverables and perhaps come back with more suggestions. We even have offered to integrate our code with theirs (as a follow on project). So in the sense of the *project* not being totally complete, Geoffrey has a point, but I do thin we fulfilled our side of the bargain. >what else remained to be done? Do you know where the code and/or data is? Err, basically integration with their package. I've only seen their code as pictures on a VDU, so I'm afraid cannot comment on the enormity of that task. I cannot be much more specific because Miloje did most of the negociations with Sailfish by himself, so I only know what has been written down. And I have a nasty feeling that the initial project specs bears little resemblance to the final list of deliverables. I always got the impression they received more than they had asked for! As for codes, can I ask you to ask Chao-Wei? I've recently mailed him with a tar file on the Sailfish s/w I have, and he should be able to point you to them quicker than me. Let me know if you have problems, of course. >Do you know anything about Miloje's online pricing system using the Web? A little. It was all CGI-forms and bins, with that ugly gnuplot->ps->ps2gif intermediate step for producing pictures. Looking back, one could see that it was screaming for Java(Script). The Monte Carlo engine for generating the statistics (based on the given parameters) is a bit of a black-box: I haven't seen the parallel code for that. Didn't one of your REU students did the parallelisation in '95? >I'm going to send Miloje all these questions also, but gcf says Miloje is >so busy now that getting a decent response may be difficult. I know. That's 'cos they pay him enough :) :) :) Ta ta for now, -Hon. =-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=