Valentina and OS-X (Darwin) revisited...

Greg Saylor gsaylor at net-virtual.com
Wed Nov 20 14:10:01 EST 2002


>> Hello,
>> 
>> A long time ago (8 months or so) I had posted about
>> using Valentina under
>> OS-X Darwin...   Unfortunately the lack of this
>> ability has held up my
>> project for over a year now - and this was only
>> discovered after having
>> purchased Revolution...  So Revolution has basically
>> been a complete bust
>> for me due to this issue... ;-(   I even convinced
>> one of my clients to
>> purchase a license as well, but thankfully the
>> mistake of purchasing it
>> for use as a database application was one that I
>> made on my own dime and
>> not my client's...  
>> 
>> I just tried the revdb under Darwin and it doesn't
>> work either:
>> 
>> #!Darwin
>> 
>> on startUp
>>         get revdb_connect("Valentina","","faa.vdb") 
>>               
>>         put "Hello WOrld!" into tresponse
>>         put tresponse
>> end startUp
>> 
>> 
>> NOTE: If I comment out the "get revdb_connect" line
>> this script does what
>> is expected (it prints "Hello World!" to the
>> console)...  If I don't I
>> get the following result:
>> 
>> [host100:Revolution 1.1.1/components/engines] root#
>> ./test
>> Darwin exiting on signal 10
>> [host100:Revolution 1.1.1/components/engines] root#
>> 
>> 
>> What this all comes down to is that I have to get
>> this project start --
>> which should have been finished by now.. And my
>> Revolution license which
>> has been doing nothing but collecting dust is coming
>> up for renewal...
>> 
>> Is it possible to either use Valentina with the
>> Darwin engine or at least
>> to get the OS-X engine to function correctly from
>> the command line?  
>> 
>> It remains baffling to me that the developers of
>> these respective
>> products have not yet integrated their products to
>> work together at the
>> command-line...  Practically every database product
>> worth any amount of
>> salt has the ability to manipulate the data using
>> command-line tools...
>> 
>> It is enough of a burden to have to learn this new
>> transcript language
>> and development environment which is extremely
>> inefficient (at least to
>> me)...  I like to have vi open in one window and
>> execute my scripts from
>> another window.. The fact that I have to keep
>> draggin' a mouse all over
>> the screen just to get something to work is simply a
>> distraction -
>> especially when trying to conceptualize some process
>> which should operate
>> entirely automated without any human interaction....
>> 
>> Sorry if I sound bitter, but I am... I have never
>> bought a product before
>> that I was so excited about and after almost a year
>> still been unable to
>> realize any value from it....
>> 
>> Thanks for listening to me gripe!
>> 
>> - Greg
>> 
>> P.S. THe "windowless OS-X" as well as writing
>> Applescript junk to speak
>> to the OS-X engine are not acceptable options here
>> either... I have found
>> both to be unreliable and clumsy to script automated
>> processes with - if
>> they would even work for this anyways....
>> 
>> P.S.S. I have also tried:
>> 
>> #!/usr/bin/open
>> ~/Revolution/components/engines/MacOSPPC
>> 
>> on startUp
>>         put "Hello WOrld!" into tresponse
>>         put tresponse
>> end startUp
>> 
>> 
>> However, the MacOSPPC engine does not seem to accept
>> "scripts" because it
>> just starts up the development environment instead
>> of executing the
>> code.....  (THis would probably be the best solution
>> - to make the
>> MacOSPPC engine work like a UNIX engine when called
>> from a command-line
>> script, if possible)...
>> 
>> 
>> P.S.S.S. I have tried lots of other stuff too - this
>> just doesn't seem to
>> work but I can't remember everything I tried....
>> 
>
>
>Hi Greg,
>
>Just to clear a few things up, as they are in my head:
>1) There is no 'Darwin' version of Valentina.
>2) If you're running the 'cgi' version of RunRev, you
>have no access to the libraries such as the database
>and printing libraries.
>3) Those libraries should be present in the version
>2.0 of the 'cgi' app ; but you'd still be limited in
>what you can do under Darwin as it has no ODBC or
>Valentina.
>
>Thus I'd recommand the following path:
>1) Run full MacOSX on the machine
>2) Build your own cgi-server with RunRev ; Gary
>Rathbone from this list has posted an example stack on
>the RunRev Newbie Board.
>
>This way you can make sure all the libraries are
>loaded and the connection to the Valentina database is
>opened at the start.
>Moreover, you could more easily maintain your cgi's as
>they're in a single stack instead of a collection of
>text files.
>
>Hope this helped, even if it wasn't the answer you
>were looking for.
>
>Best regards,
>
>Jan Schenkel.
>
>=====
>"As we grow older, we grow both wiser and more foolish at the same time."
> (La Rochefoucauld)

John,

Thanks for the reply!...

Regarding your clarification:

1)I don't quite get what you are saying....  There is a "Macho" version
of Valentina - isn't that one which will work on Darwin or am I confused?...
2) By the 'cgi' version do you mean this "Darwin.tgz" I downloaded from
http://www.runreve.com/engines11/Darwin.tgz ?...
3) [ blank look ]

Regarding your recommendation:
1) Already got that.
2) Where is the RunRev Newbie board?... I must be so new that I can't find it.


Is he talking about running some sort of client/server application with a
Revolution stack running as the server piece?.....  If so will it be able
to handle multiple simultaneous requests?...

The idea of maintaining text files inside of stack is precisely what I
want to get away from.. I want to be able to edit my files with vi or
bang out scripts for doing database manipulations from some other tool
(like PERL)...  The big thing for me is that what we are workign on is
moving a portion of a client/server application which currently resides
in Oracle down to PC's and Mac systems..  The current process of
collecting data and keeping the database updated is almost 100%
automated....  I want to ensure that this process continues
transparently....  So what I ened is a way to build, index, basically I
need to be able execute all of the Revolution transcript functions from
the command line - plain-and-simple...  If you are suggesting that I can
setup a server which will allow me to do this, it is interesting to me...
If you are saying I still have to bury my code inside of a GUI
development environment where developers can't get at it to fix it
without popping up some GUI application, that's a lot less interesting to
me (and in fact, I would probably get laughed out of the office if I
tried to suggest it)...  It's goign to be tricky enough to get them to
learn this new transcript language (I am sure there will be a lot of
resistance to it - as there has been with me)....  So in other words, I
cna play a little bit of Voodoo magic here and hide an OS-X server in the
back room where nobody has to deal with it, but when they write their
scripts they are going to want to be able to develop those scripts using
the development tools they already have for writing command-line apps and
to be able to do it in basically the same way....  They have over a
million lines of code devoted to this application and to have some other
piece hanging off a ledge somewhere else is just going to be very
distracting when trying to track down bugs or add new build procedures....

As a side note, I have noticed on several OS-X development sites that
does appear to be possible to call Carbon and/or [whatever] functions
from a command-line environment...  I'm not sure what all the terminology
is, for some reason Mach-O, CDF, and Carbon seem significant....  But,
that there are ways to cross those boundaries and link in libraries from
one of the other environments...  Though this part of my research ont he
subject goes back a long ways, that seems to be the result of what I came
up with....

Thanks, again, for your reply!..

- Greg




More information about the use-livecode mailing list