|
PureBytes Links
Trading Reference Links
|
<x-html><!DOCTYPE HTML PUBLIC "-//W3C//DTD W3 HTML//EN">
<HTML>
<HEAD>
<META content=text/html;charset=iso-8859-1 http-equiv=Content-Type>
<META content='"MSHTML 4.72.2106.6"' name=GENERATOR>
</HEAD>
<BODY bgColor=#ffffff>
<DIV><FONT color=#000000 size=2>Please add my name to the list of customers
strongly urging this be included.</FONT></DIV></BODY></HTML>
</x-html>From ???@??? Tue Aug 04 13:46:29 1998
Received: from freeze.metastock.com (204.246.137.5)
by mail02.rapidsite.net (RS ver 0.3) with SMTP id 4880
for <neal@xxxxxxxxxxxxx>; Tue, 4 Aug 1998 15:13:35 -0400 (EDT)
Received: (from majordom@xxxxxxxxx)
by freeze.metastock.com (8.8.5/8.8.5) id KAA19049
for metastock-outgoing; Tue, 4 Aug 1998 10:46:42 -0600 (MDT)
X-Authentication-Warning: freeze.metastock.com: majordom set sender to owner-metastock@xxxxxxxxxxxxx using -f
Received: from mars.superlink.net (root@xxxxxxxxxxxxxxxxxx [204.97.220.9])
by freeze.metastock.com (8.8.5/8.8.5) with ESMTP id KAA19033
for <metastock@xxxxxxxxxxxxx>; Tue, 4 Aug 1998 10:46:33 -0600 (MDT)
Received: from superlink.com (609-35.superlink.net [207.76.66.35])
by mars.superlink.net (8.8.8/8.8.8) with ESMTP id MAA19379;
Tue, 4 Aug 1998 12:39:03 -0400 (EDT)
Message-ID: <35C76098.B56F3555@xxxxxxxxxxxxx>
Date: Tue, 04 Aug 1998 12:27:21 -0700
From: Vitaly Larichev <vitaly@xxxxxxxxxxxxx>
X-Mailer: Mozilla 4.04 [en] (Win16; I)
MIME-Version: 1.0
To: metastock@xxxxxxxxxxxxx, support@xxxxxxxxx
Subject: Re: FW: future of MSWIN
References: <001501bdbf5e$1db3de00$872084d0@xxxxxxxx>
Content-Type: text/plain; charset=us-ascii
Content-Transfer-Encoding: 7bit
Sender: owner-metastock@xxxxxxxxxxxxx
Precedence: bulk
Reply-To: metastock@xxxxxxxxxxxxx
X-Loop-Detect: 1
> From: Equis Support recently:
> ....System Testing for multiple securities is not a large request either....
Are you serious? The presumption of many on the list is that you collect and
compile requests from users. Then you would remember how many times, perhaps
not as recently, we've asked, even made long, thoroughly thought out lists of
our suggestions on how to improve MS. They did include System Testing for
multiple securities, sequential explorations, much better file management, etc.
etc. So often we don't voice our support over and over again assuming that our
opinions have been registered in some Equis files (sounds almost like Eqs-files
<g>). Seems a mistake.
As for System Testing, I honestly thinks it's not usable at all if you are to
build a system. To assess a quality of a system, you have to gather STATISTICS
on it! That is to run it on multiple securities. Having had it, you may then
check a standing of a particular security vs. "all others" (within the trading
system). But the way System Testing is now, you don't have anything to check
against unless you have a many years of experience to substitute that gut
feeling for the statistics! So it's not System Testing, it is Stock Testing.
You can compare systems on a secuirity, of course, but again it's not about an
overall quality assessment of the systems. Why to waste time finding out which
one of two lousy systems is worse on that particular secuirity?
Another minor programmably, but nagging in everyday use thing - a logic in "IF"
condition: it calculates ALL expressions before checking the branch condition,
i.e. in IF(A>0,B,C) it calculates A,B,C before checking if A>0, although both
values of B and C are not needed due to A>0 condition (it's the case for MS
5.11, cannot say about later versions, but if the expression
If (2>1,1, 1/(1-1)) gives you now an error calculating, that's it). The
problem which may seem too puristic, is however very much practical. If your
every night test takes an hour or so, that could be due to overhead in
redundant calculations. I would love to save time by nesting "IF" conditions
in such a way that the first one, fast and easy, would eliminate a majority of
candidates whereas others, more time consuming would deal with much smaller
number of stocks, sort of sequential approach.
Without cheers on the issue, Vitaly
|