[ruby-talk:9503] Re: 101 Misconceptions About Dynamic Languages

On Thu, 18 Jan 2001, David Bolen wrote:
> "W. Kent Starr" <[e-mail]> writes:
> 
> > My personal take on deployment issues is consumer end-users.  These
> > people want an easy install, which any application not distributed
> > in compiled form does not deliver.
> 
> Well, that's not necessarily true.  The installation has to appear to
> be a self-contained "compiled" program, but that doesn't mean the
> application itself really has to be.  Although perhaps it's just
> playing games with what "compiled" means.
> 

I guess "transparent" might be a better word, or even "idiot-proof"? :-)
As in click-install-run.  Which is about as much as a typical end-user in a b2c
environment (as well as many b2b) wants to really know (or imo _shoul_ really
need to know). Increasingly, software at the b2b level is becoming a "service"
rather than a "product" and I believe this trend will continue into the
consumer level. 

Actually, there are systems with tcl/Tk whereby everything is put into a single
zip file in executable form, which is ery much what you are talking and which I
think would answer the need, proided the exe didn't end up being a huge
multimegabyte monster. (Associated data is a separate issue, of course).
Although gloppies are passe`, I think the "exe fits on one disk' rule of thumb
is still a good target to shoot for.

And, with the tcl/Tk approach, noted above, if you keep the wish versions for
multiple systems handy, you can do cross platform builds on a single box
(whatever flavor you prefer).  Nice touch :-)

I guess what I lean toward (but don't express very well) is that I believe Ruby
is an exceptional opportunity and I would like it to address this issue and
that of GUI, which then covers most of the required bases. I like the fact that
if it doesn't have (or more often I just can't locate) a feature I want, I can
just 'modify' Ruby to make it happen. This is in contrast to my last post
(which Dave Thomas politely corrected me) re changes to 'read'.  I was thinking
in terms of lifo for data and tail for text (reading from the end of the file).
Of course, the means to that functionality already exits in Ruby -- one just
needs to look for it (duh)

So, my original, pre-that-read-post thing is that there is a lot of unexplored
functionality inherent in Ruby as it now is to be explored.  This does not mean
re3al 'bugs' don't need to be fixed but it does mean that many of the
'features' we discuss here may already be there and we can with a little
thought and effort we can 'find' them.  Does this mean Ruby is the best
solution for evey problem?  No...such a thing does not exist. Do I believe Ruby
has a shot at being a major contender? Without doubt!  We just need to address
the deployment and presentation (GUI) issues.

My 2 cents.

Kent Starr
[e-mail]