Rob Whitfield <rob at customfun.com> wrote:
> I'd like to use squeak to serve as a platform independent means to collect 
> user input and drive a series of command line utilities.  I've downloaded 

Sounds fine.

> squeak and worked through a number of tutorials.  Squeak is cool!  However, 
> at this point I have no idea how I'd create a squeak program to open a 
> file, read some data, display the data, allow the user to enter some 
> additional data, start a separate executable, etc.
> Can anyone point me to documentation that would be of use in creating a 
> standalone squeak application that does the things mentioned above?  Is 

Well, my advice is that first you read a beginners book on Smalltalk
with Squeak running on the side of course to try stuff. There are also
quite a lot of tutorials etc for Squeak (check the Squeak swiki).

Then, when you have learned the basics of the language you should start
learning how to find your way in the image. The image is like a big
Library and if you don't know how all the books are categorized etc in a
Library you won't find your answers!

Some things off the top of my head:

1. Method finder.
2. Alt-m, Alt-n, Alt-b, Alt-q, Alt-p, Alt-d, Alt-i, Alt-I (these are the
most important I think) and friends.
3. "find class..."
4. Reading class comments (press the '?' in the system browser) - those
that are there! :-)
5. Use Hierarchy browsers
6. Good to repeat: Use Alt-n (senders of). If you have found a method in
a class that seems interesting this will bring up other methods using it
so that you can "learn by example".

Note: "Alt" for Win32 - on Linux it seems to be in a flux and on Mac I
think it's called "Command". 

> there a good book that I should buy?  Also, when I start squeak it comes up 

There are a few Squeak books - especially two by Mark Guzdial. Those
books are "compilations" of chapters written by a number of people and
some chapters are for newbies, especially in the first white book. So
unless someone else here knows a better book that might be a good
starting point.

> with its own environment containing elements of the development 
> environment.  How do I start squeak in a way that would run my program so 
> that my users don't need to know anything about squeak?

Well, what platform are you on? We just deployed an application on Linux
and Win32 and we do it in two steps:

1. Do a few manual things like removing all windows, destroying global
flaps, check Preferences etc. (that can off course also be done
programmatically).

2. Run a class method in the development image that:
	-cleans up a few things of our own (checks that there are no stray
instances/Processes of ours around etc)
	-strips the image from stuff our app doesn't need. (based on stripping
code available on the Squeak swiki)
	-saves this new image on another name

This gives us an image that doesn't need the sources or the changes file
and is about 6Mb with Morphic still in - it only needs a VM to run. Note
that we do NOT prepare the image in any way to "autostart" stuff when
coming up - like a first window on the application. You can do that of
course in a number of ways, but we instead opted to use the facility of
the VM to pass a little script-file as first input:

On Win32 we have a .bat file called "proxy.bat" (the name of the
application is 'Proxy') with one line in it:

.\Squeak.exe squeak.image proxy.st

This starts the image and the script proxy.st will be evaluated
immediately.
The script "proxy.st" is also just a oneliner that looks like:

VHProxy startProxy

...and that is the class method that starts the application. One of the
advantages of putting "startup" code in a script file like this is that
we can have multiple files for different apps all using the same image.
;-)

regards, Göran