Am 10.01.20 um 10:16 schrieb Marten Feldtmann:
That happened once in the history of Smalltalk and the big player was
IBM ...
Well, twice actually ;-)
Many people might not know that HP once was a Smalltalk vendor with
their distributed Smalltalk (which was actually a white-label copy of
VisualWorks iirc/iiuc)...
Big corporations do not warrant the success of a technology. We'd still
be using OS/2, BS2000, whatever... today if that was the case.
The much bigger impact on Smalltalk's demise was that all vendors back
in the nineties told their customers to switch somewehere else. One of
them even never delivered what they had suggested as a replacement...
I tried not to jump in here, but I am still a bit surprised how many
people state that Smalltalk was so superior to (whatever hyped thing of
the day).
How many of us have built a nice iOS or Android App in Smalltalk? Do we
even havy anything that might allow to do that in a power efficient way?
How many of us have built a modern-looking SPA web application in
Smalltalk? Smalltalk-only?
These are areas where Smalltalk is quite weak and may not excel in any
time soon. Let's be honest about it. None of the commercial vendors have
announced anything in that direction, and to my knowledge, none of the
open-source ones have anything more that the usual JS bridging.
Dart/Flutter promises to finally be a platform on which you can write
apps for both mobile platforms, the web and maybe even native with one
single code base. Does that sound like a dream to developers who need to
pick a tool? I guess so.
Is there any point in telling people how great Smalltalk is when writing
a mobile app or an SPA is hard to almost impossible?
I know there are prototypes and even shipped apps on mobile platforms.
Some are available and can be used, but are complex and mostly
undocumented. One is private and doesn't even have a (public) price tag
(yet?).
Not really comparable to downloading Flutter and starting to write your
first mobile app on a Saturday afternoon. Not sure the productivity
advantage we keep praising (be it existent or not) will be noted much by
somebody who needs to pick a tool now.
So let's stop trying to convince people with things that mattered some
20 years ago. Even the function point thingie we keep carrying in front
of our bellies (Capers-Jones was it?) is a lie when you want to build an
application for today's markets.
Smalltalk is great. If you don't need a mobile app to accompany your
product. It's great if your GUI doesn't have to be sexy as hell or you
are happy reinventing wheels. It is great if you only ship to PCs or on
the web and don't need a lot of interaction in the browser. Anything
else is hard in Smalltalk. If it's not, it is undocumented.
I am not sure if energy spent on these "syntax fits on a postcard and,
btw, we have the balloon" articles could be better spent doing
something about the problems I mention here. I was in the same boat in
the late 90ies and early 2000's with my blog and articles and stuff. All
I found out was that nobody actually cares about these old hat stories.
Heck, a lot of people these days don't even care about maintainability.
You don't like it any more? No prob, we can redo it in this great new
(JS) framework anyways.
I am not saying anything is wrong about the maintainability of Smalltalk
code or greatness of our IDE support. I like it and use it every day. I
love programming in Smalltalk and hunting bugs in Smalltalk and fixing
customers' problems in Smalltalk. I can fix a bug while the user tells
me about theit problem and I am convinced not many other technologies
can support me in this as good as Smalltalk does. But, unfortunately, I
spend way too much time reinventing GUI glue code fo the web. And,
unfortunately, I need to implement a lot of stuff in Javascript. Oh, and
unfortunately, there is nothing as good as this if I wanted to ship a
nice mobile app to augment our service. When it comes to these, I am not
productive. I am not even close to any bleeding edge, and if I want to
be, I need to do it on my own. Is it more productive to reinvent a
Smalltalk version of VueJS just do be able to do the same as VueJS in JS
than it would be to learn JS and VueJS and do some part of my work in
those?
I wanted to stay out of this thread, because it leads nowhere. But now
that I've typed all this, I will push the send button and regret it in a
few minutes...
Joachim
>and actually that really showed impact to the Smalltalk market.
Lots of consultings were running around, get pretty much money to teach
COBOL programmers how to use Smalltalk (or to be more precise: learn how
to click programs together).
That hype perhaps lasted a few years ... and then IBM switched to Java
... so they never can go back.
Marten
Am 09.01.20 um 23:16 schrieb horrido:
It would be really nice to have some big tech company adopt Smalltalk, like
Amazon, Apple, Facebook, Google, Microsoft, Netflix, Uber, etc. That would
hit the ball right out of the park. Alas, I don't see that happening. I'm
afraid JP Morgan, Siemens, and Thales aren't good enough.
--
-----------------------------------------------------------------------
Objektfabrik Joachim Tuchel mailto:jtuc...@objektfabrik.de
Fliederweg 1 http://www.objektfabrik.de
D-71640 Ludwigsburg http://joachimtuchel.wordpress.com
Telefon: +49 7141 56 10 86 0 Fax: +49 7141 56 10 86 1
