On 12/7/2010 5:38 PM, [email protected] wrote:
Sean,
Hope that additional to the fun, we can converge to an understanding how can we
increase Smalltalk popularity.
OTOH, I made the reference as a sample, I did not intend to have this as all
encompassing list of issues nor start a thread about comparisons with other
languages.
As I said earlier, what is 'trivial' or 'intuitive' for some is 'maverick' or
'bizarre' for others, it is in the eyes of the beholder... ;-)
My main message is more around this: although Smalltalk was a very advanced
technology when its inception, present status have almost all of its
characteristics embedded in present technology. Even, when newer concepts as
xUnit have started in Smalltalk, the _concept_ was absorbed in mainstream
technologies so again we're stuck with small delta between what Smalltalk does
'better' than the incumbent technologies we would like to replace.
I can't help but disagree with this most strenuously.
Almost nothing has touched Smalltalk's advanced technology. They don't
have the world view or paradigm for doing so. They are languages that
operate in dead cycles. They have dead processes and do dead post mortem
debugging. Yuck! (yes, I know that was redundant. :)
I am not a computer scientist or trained programmer. I am a creative guy
who is a businessman. I find Smalltalk empowering. I also use Python
when necessary, but Squeak/Pharo is my preferred language. I find myself
infinitely more productive in Smalltalk than in Python. Operating in a
live environment is incredibly empowering, enabling and productive.
Where else can I find this. C, C++, C#, Java, Python, Ruby, VB, nope,
none...
Demonstrating large scale Smalltalk projects have been canceled does not
prove Smalltalk's inferiority, nor that the competition (should one
choose to refer to them as such) has caught up or passed us. Especially
when said project was not canceled for technical reasons but political
ones and for software that was/is demonstrably inferior to the already
existing Smalltalk software. But what it does prove is that Smalltalk
has been and can be effectively deployed for large scale projects where
someone has the courage or political will to choose something different
than the status quo languages.
The decisions of PHBs (Pointy Haired Bosses, Dilbert) does not
invalidate anything about Smalltalk. After all their primary motivation
is job security, not project success. No one ever got fired for
choosing, MS, Java, IBM, Oracle, etc...
The biggest weakness for Smalltalk or at least our open source version
in Squeak/Pharo IMHO is its integration, cooperation or access story to
certain outside systems and technologies. And I speak as an end user who
cannot develop a plugin, or use Alien or FFI and program in C. In my
situation for example. I have to interface to either a Windows dll or to
a Java library to access financial servers. No choice. So I must choose
a technology which can interface their technology. In my case, I chose
to use Python to interface the Windows dll. I have written a minimal
Python app which accesses the dll and thus the servers to access the
data and my accounts. But the business logic I am writing in Pharo. The
Python app simply provides my Pharo app the data and executes the
actions directed by the Pharo app.
I would that I could have interfaced the dll (or Java library) directly
from Pharo. But despite that obstacle I am choosing to write my app in
Pharo. Most in this business are using VB/C#/Java. They do not provide
the flexibility or productivity that I have in Pharo/Smalltalk. They do
not enable the small guy like me as well as Pharo/Smalltalk. Yes, this
is my opinion. But guess what. This is a Pharo mailing list populated by
people who are for and proponents of Smalltalk. If we favored other
languages and environments we would be there. Many of those on the this
list are also pretty dog gone expert in the other technologies as well.
They did not make their decisions out of ignorance. What I don't
understand is that if you are so convinced that the advantages of
Smalltalk are so minimal and of no consequence, then why are you here?
I would like to see Pharo/Squeak bridge the gap to be able to interface
system components or other business libraries. I would like to be able
to use Pharo/Squeak anywhere I could use Python. I would like to see
that there are fewer and fewer technical reasons for not choosing
Pharo/Squeak. I can understand other choices. Other people are more
comfortable with other languages, systems, environments. They may have a
knowledge, systems, and process investment. But this can only validate
their decision for them and in no way invalidates the technology or
capabilities of Smalltalk.
There are many, many projects for which Smalltalk/Pharo/Squeak have no
discernible technological disadvantages. They do not require any systems
integration or use of the "native ui".
Enterprise. I agree with the sentiment that we don't need to worry about
it. Let Cincom address enterprise users. We need to empower the little
guy in the enterprise who has small side projects which do not have to
go through the same channels as the larger "enterprise" projects.
Projects for which they might choose Python or Ruby. Over time we can
improve our enterprise story and infiltration. Fortunately for us,
Smalltalk/Squeak/Pharo is in the long game. It has seen technologies
come and go and is still here. If we the open source Smalltalk community
continue to make the improvements we are making, then we will be an
increasingly viable choice for the entrepreneur, small businessman,
enterprise guy with the side project, the person with a project they do
on their personal time at home. It is the most enabling technology I know.
Well, I'll end my rant here and get back to being enabled in the
development of my business app.
Jimmie
