[alicebot-archcomm] Looking for <fetch> in all the wrong places?
Gary Robertson
alicebot-archcomm@list.alicebot.org
Mon, 04 Aug 2003 11:02:24 -0400
Ernest (earlier):
>Nice result of your last occupational therapy session - isn't it?
Gary:
I truly do not know what you mean by your statement above.
Ernest (earlier):
>*Each* of your examples need different protocols, interfaces and/or
>local/external/remote processing and can't be done by a simple fetch of
>a website.
Gary:
Agreed. I never suggested this could be done by a simple fetch utility, but
I am suggested that something similar or better is needed before many
bot masters would attempt any of it at all.
Ernest (earlier):
>Moreover some of them need context knowledge about the user - location
>or personal data like bank account i.e..
Gary:
My very thoughts. Now we are talking! It would not be so hard to link a
user to a database account that would manage this information. I do it all
of the time
in the web world.
Ernest: (earlier)
>The examples are just tasks a good private secretary should carry out.
Gary:
Yes... sort of like typing papers and letter mailing used to be, but now we
do it
all much more efficiently with our personal computer systems and cool software.
Ernest: (earlier)
>I think, a budget of 250.000 US$ will help creating such a bot.
>Notabene: AIML 1.01 will not be extended, it's just a lot of work
>creating a whole bunch of categories to cover the named tasks. The
>processing/calculating/interfacing can be done using <system>, <perl>,
><php> or <javascript> depending on the AIML interpreter used.
>At present I have no other urgent projects and could just start over
>with yours.
>
>Looking forward to your cash remittance.
Gary:
Ernest, I am glad you are having good fun with this. I am also. Isn't
this the
way it should be? Although perhaps your $ figure may be reasonable,
I think that if we do a good job documenting and promoting a bot master
friendly fetch utility or fetch methodology, we can split the work load
amongst dozens,
if not hundreds of interested parties!
I believe that once we get AIML to demonstrate a good <fetch> functionality,
we will begin to see AIML applications grow out of the "novelty" arena and
more into the
"want to have" or "got to have" arena.
Ernest, I favor whatever works. Perhaps you can demonstrate for
me a solution that suites you. If <system> and <perl> is the way to go, please
show me a good demo. Keep in mind that I REALLY want to use <coldfusion>
which is
very similar to <php>. So maybe you can show me a good example in
<php> and I will translate it over to <coldfusion>. I sense there is a
reasonable solution at hand that will make most folks happy. Thanks for taking
the time and effort to consider it.