|
PureBytes Links
Trading Reference Links
|
Can't speak to the OMZ files, however the data structure used in the .ORD and
MASTER.* files uses an integer (not word) data type with values ranging
from -32,768 to 32,767 = 65535 days = 179 years with base of 1/1/1900. My
testing indicated that the data field itself could not handle my pre-1900
historical data and would have no problems handling dates through a good part of
the next century. It's the other parts of the software which can't handle Y2K.
Earl
-----Original Message-----
From: Gerrit Jacobsen <Gerrit.Jacobsen@xxxxxxxxxxxxx>
To: omega-list@xxxxxxxxxx <omega-list@xxxxxxxxxx>; Bob Fulks
<bfulks@xxxxxxxxxxxx>
Date: Friday, July 17, 1998 3:36 AM
Subject: Re: Is TS 5 next generation or not?
The date field in the internal Omega 4.0 data structures is WORD
which covers 65535 days = 179 years.
Unless Omega expands this field length someone will be out of luck.
Either the 18th century grain price followers or the people who are
trading forward months on nuclear waste certificates.
A radical step would be to increase the length of the field which
should not be tooooooo difficult. But this would propably
involve some changes everywhere in the code.
The cheap solution would be to change the meaning of the date field
and make TS compatible in between 1900 and 2079 until the last TS 4.0
user has died. (Although I am pretty sure that some Mark Brown
reincarnations will survive)
Gerrit Jacobsen
> I suspect the real reason that the Y2K 4.0 patch will follow the
> release of 5.0 is that there will not be a Y2K "patch". Such
> "patches" are too expensive to build and test and have no revenue
> impact (no direct return-on-investment). Even if the code changes to
> 4.0 are minor (unlikely), the testing is still a major effort. And
> how many of you will volunteer to beta test a Y2K patch on your
> critical data. Not me.
>
> What usually happens in such cases is that all 4.0 users will get a
> "crippled" version of 5.0 with certain 5.0 features there but
> disabled. It is very easy to build software such that certain
> features are disabled. So I would bet that 4.0 customers will either
> upgrade to 5.0 for a fee, or get a crippled version for free. In any
> case, we will be using 5.0 code whether we want to or not. 3.5
> customers are probably out of luck.
>
> Just a guess but highly likely based upon my experience.
>
> Wouldn't it be nice if Omega would tell it's customers what the
> plans are? I guess that is asking to much.
>
> Bob Fulks
>
>
7 PM -0400 7/16/98, Timothy Morge wrote:
>
> >Your possible explanation about early releases of TS ver 5.0 not
> >being year 2000 compliant makes some sense, I guess...in a
> >haphazard design sense [which I could now imagine Omega using].
> >However, if it isn't that difficult for them to 'compartmentalize'
> >their design and research products, it should be pretty darn easy
> >to take about 75 percent of their code staff and put them on
> >finishing a year 2000 patch right now. I mean, they no longer have
> >to worry about releasing version 5 on time...it was initially
> >planned to be released last quarter at the latest.
>
Gerrit Jacobsen
http://www.tickscape.com
|