|
PureBytes Links
Trading Reference Links
|
Hi Tomasz,
Thursday, May 29, 2003, 4:50:48 PM, you wrote:
TJ> It is not random becuase there is NOTHING random in computers.
Not supposed to be, I understand. But then not everything as it's
supposed to be. ^^_^^
TJ> The algorithm for cache works generaly as Nigel described (the
TJ> least recently used symbol is flushed) but you should remember
TJ> that all "Foreign" calls in any of your indicators also cause
TJ> 'retrieval' of data.
Fair enough. But if I'm looking at the same sheet and scrolling
through the stocks, the calls would all be exactly the same, because
the same indicators are being applied to each issue, right? So why
is the behavior (not the same) random? <g>
TJ> It works that way that if you start the computer the first
TJ> displayed stock is retrieved from the plugin (the data are as in
TJ> your metastock database) and cached. and so on and so on until
TJ> you exceed cache size. if you updated your MS database sometime
TJ> in the MIDDLE of 'scrolling process' then cached data are old and
TJ> symbols that are not in the cache will be 'new'. That's all
TJ> philosophy. Nothing really random. And I am not going to change
TJ> this as it works well for (almost) everyone.
So let me see if I understand this, because we are both using a
foreign language here. ^^_^^ If I view a symbol before the update,
it's "old", and will retain it's data until I refresh, regardless of
what I do with the database. (Let's assume for this argument that I
am not running out of cache size.) But, if I view a symbol AFTER the
database is altered that I did NOT view before, that symbol is "new",
and therefor refreshed. Is that right? This of course would be a
possibility to account for the behavior, as I may not view every
stock in the database prior to entering new data, and there might be
hits and misses with the views (not necessarily in order).
TJ> And as I wrote you there is no option NOT to read the last bar.
Don't want that. Forget about it. <g>
TJ> I suggest that your pre-market update should NOT contain zeros
TJ> but a COPY of previous quote but with zero volume. This seems
TJ> far more resonable than adding quote with price set to zero,
TJ> especially considering the fact that price = 0 is unsually
TJ> invalid and causes trouble when someone uses semi-log charts.
Ah . . . Tomasz. Do you really thing I'm inputting zeros here??
They get put there apparently as placeholders by The Downloader,
since I enter dates up to 30 minutes or so before data. How on earth
I would enter the previous day's data automatically (as I load new
dates) I don't know. In any case, that would skew the chart I'm
looking at anyway.
Do I understand the behavior now?
Best,
Yuki
------------------------ Yahoo! Groups Sponsor ---------------------~-->
Get A Free Psychic Reading! Your Online Answer To Life's Important Questions.
http://us.click.yahoo.com/Lj3uPC/Me7FAA/CNxFAA/GHeqlB/TM
---------------------------------------------------------------------~->
Send BUG REPORTS to bugs@xxxxxxxxxxxxx
Send SUGGESTIONS to suggest@xxxxxxxxxxxxx
-----------------------------------------
Post AmiQuote-related messages ONLY to: amiquote@xxxxxxxxxxxxxxx
(Web page: http://groups.yahoo.com/group/amiquote/messages/)
--------------------------------------------
Check group FAQ at: http://groups.yahoo.com/group/amibroker/files/groupfaq.html
Your use of Yahoo! Groups is subject to http://docs.yahoo.com/info/terms/
|