Si, the syntax is <start using stack "someStack">, which puts the
script of stack "someStack" into the message path ie. available to
all other open stacks.
The script of a mainstack of an app is already in the message path,
so you wouldn't need to put a mainstack 'in use'.
<Start using> is probably the most common way that library stacks
(stacks containing lots of re-usable handlers) are implemented. Many
of us have built up collections of generally useful handlers which we
put in a library stack, and put 'in use' (both in compiled apps, and
in the IDE) so we don't have to re-write them for every project.
For instance, quite a few people coming to Rev from other languages
ask here about a 'trim' function (to trim white space from the
beginning and end of a string), which is not a built in function, but
is very commonly wanted, so many of us will have written one like:
function trim pString
return word 1 to -1 of pString
end trim
and rather than having to write this function for every project, you
can put it in the script of a library stack containing many such
useful functions, and copy that stack into each project you start,
and put it 'in use' at startup.
It also makes it possible for people to write libraries of handlers
that they can then share with or sell to others. RevOnline contains
many libraries that people have decided to share, and so there is a
great deal of extra functionality available to us that is not part of
the regular Rev install.
Best,
Mark
On 3 Apr 2007, at 09:49, Simon HARPER wrote:
So as a newbie, would I be right in thinking that you can call any
function from the library stack in any other stack, or must the
library stack be the mainstack. Also you say "put 'in use' at
startup" do you have to do this via a specific command?
Cheers
Si.
====
Simon Harper
2.44 Kilburn Building
University of Manchester (UK)
Pri: [EMAIL PROTECTED]
Alt: [EMAIL PROTECTED]
On 2 Apr 2007, at 23:20, Mark Smith wrote:
Well, I'm certainly not going to presume to criticise your method,
but in the spirit of "you show me yours, I'll show you mine", what
I tend to do is to have a stack in my apps which is put 'in use'
at startup, and which has various getters and setters. The actual
data it returns could be in cps, script locals, text files,
databases or somewhere on t'internet, the calling handlers don't
have to know. So:
answer gcOK() in a handler calls the gcOK() function in my library
stack which might be
function gcOK
return "OK?" -- literal
end gcOK
or
function cgOK
return the okMessage of me -- cp
end gcOK
or
function gcOK
return sOkMessage -- script local
end gcOK
or
function gcOK
get URL (http://www.google.com/search?q=OK";
--parse html and find what you need
return whatYouFound
end gcOK
etc.
This approach proved very handy indeed recently, when the data
(retrieved from the web) for one of my apps changed format, but
all I had to do was re-write a couple of getter functions, the job
was done, and I felt quite smug.
Script locals are quite a neat way of achieving data-hiding when
used this way.
Best,
Mark
On 2 Apr 2007, at 23:53, Graham Samuel wrote:
Forgive me if this conversation has ended, but my internet
connection has been in meltdown... just got back on line.
I most frequently use globals because there aren't global
constants. I use them very largely for strings containing stuff
like error messages or even very simple strings like "OK", so
that I can refer to these indirectly in scripts, thus allowing me
to switch (human) languages by redefining the globals in just one
script of the program. I guess I could have used custom property
sets with exactly the same effect, and with the advantage that I
wouldn't have to initialise them during the startup of my app,
but like many others I didn't understand these when I started,
and I tend to re-use stuff I wrote before. I guess there isn't
much difference between writing
answer gcOK -- 'gcOK' is a global with a string in it.
and
answer (the gcOK of stack "allTheConstantStrings") -- 'the
gcOK' is a property of some object.
but the second statement seems to have more characters in it,
since it involves referring to the object in which the property
is stored. If there are a lot of such references, my scripts are
going to get longer.
I also use globals when I have a quantity which needs to be used
in different scripts in different stacks, i.e globally: a very
obvious point, but I really don't see what is wrong with that. I
do accept that I have to be disciplined about changing their
values. I do use properties (I tend to use these for global
status stuff like 'the soundOn of this stack'), parameter-passing
and message-passing extensively, but to me globals feel right for
quite a lot of things.
I shall now wait for someone to tell me why this is a really
wrong-headed approach. I'm always willing to learn - really.
Graham
----------------------------------------
Graham Samuel / The Living Fossil Co. / UK and France
_______________________________________________
use-revolution mailing list
[email protected]
Please visit this url to subscribe, unsubscribe and manage your
subscription preferences:
http://lists.runrev.com/mailman/listinfo/use-revolution
_______________________________________________
use-revolution mailing list
[email protected]
Please visit this url to subscribe, unsubscribe and manage your
subscription preferences:
http://lists.runrev.com/mailman/listinfo/use-revolution
_______________________________________________
use-revolution mailing list
[email protected]
Please visit this url to subscribe, unsubscribe and manage your
subscription preferences:
http://lists.runrev.com/mailman/listinfo/use-revolution
_______________________________________________
use-revolution mailing list
[email protected]
Please visit this url to subscribe, unsubscribe and manage your subscription
preferences:
http://lists.runrev.com/mailman/listinfo/use-revolution
