Did you see the following presentation (about a VW to Pharo port):

 

Apart Framework: Porting from VisualWorks

 

 
<https://www.slideshare.net/pharoproject/apart-framework-porting-from-visual
works>
https://www.slideshare.net/pharoproject/apart-framework-porting-from-visualw
orks

 <http://files.pharo.org/conferences/2019PharoDays/07%20aPart.pdf>
http://files.pharo.org/conferences/2019PharoDays/07%20aPart.pdf

 

?

 

No, I missed it.  Thanks for that.

 

I have a lot of algorithmic meat in domain classes.  Those are easy to port,
even if I must manually trip over the minor VW/Pharo differences, and fix
them as I come to them, or just run my own VW script to do mass edits on
file-outs.  I don't mind this part of the conversion much, and it would
cover most of what I have port.  

 

I have some special GUIs (2D and 3D) that I need to recode anyway, probably
in Spec and Woden.  The 3D stuff is now done in Jun OpenGL, which has seen
its day, and I suppose won't be updated to use the OGL shader model.  The
last update for VW, I think, is from 2015.  I've had enough of Jun anyway,
and would prefer to use Woden in Pharo, or port it to VW, and work on in
there.  But the GUIs in recent years are not the big issue.  The domain
objects are.

 

Mostly, I value reading efficiency.  We spend most of our time reading
programs so that we can determine how to write some change or new feature
without breakage or incongruity.  I value reading at high rates and with
high comprehension, small methods, and rigidly controlled naming
conventions.  The ultimate aim is to have the machine reason about my code,
with the same or greater grammatical and semantic awareness.  This happens
now only at a superficial level with classes, methods, and referential,
slot-based integrity checks.  I'm happy they are there, but they are not
nearly enough to make humans extremely productive and powerful at shaping
machine behaviors.  I think we are overworked and that our machines are
hardly being used.  There is a science to increasing reading speed and
comprehension in a programming context (and getting the machine to help you
in the domain layer), and you can't use that science very well with speed
bumps at the beginnings of all the class names.   Therefore, I very much
value naming classes exactly as I need.  Even this is an interim approach.
If things go well, classes and methods will go away in the future, but I
digress from the near-term task.

 

I saw the bit near the end of the slides about Spec2.  Does this really use
GTK to make native GUIs, and does this mean we can have real OS windows on
Windows?  I miss Alt-Tab-ing for navigating through windows, and would like
to see it in the Pharo, but that can't happen without OS windows.  Any
Windows Pharo users miss Alt-tab-ing?  

 

Can Spec2 be tested in Pharo8 now?

 

Please point me to the best resources on Spec2.  It looks very interesting.
I found these:

 

https://pharoweekly.wordpress.com/2019/04/18/pre-ann-spec-2-0-gtk3-bindings/

 

"

And that's it. Pharo 8.0 will come with Spec 2.0 and users will be able to
benefit of it immediately 

Pharo 80 will not need Gtk3. We will provide packages for Gtk3 for the
people that need it. We will also release Spec20 soon so that people can try
Spec 20 - independently of Gtk3

"

 

https://github.com/dionisiydk/Spec2

 

I don't see Spec2 in Pharo 8.  I suppose I need to load it.

 

Is the Spec book still relevant?  Is Spec2 mostly about the back-end changes
for use of GTK?

 

 

Shaping

 

> On 2 Jun 2019, at 12:12, Shaping < <mailto:[email protected]>
[email protected]> wrote:

> 

>                 

>  

> I'm fairly sure Environments hasn't been ported to Pharo.  Namespacing is
still on open question on Pharo.

> The little I know about the topic is that its complicated since it can
introduce as many problems as it solves.

> its something that is wanted, but priorities have taken precedence.

> A bit work being done in the background should support namespaces when
there are resources to "get it right" ...

>
<http://forum.world.st/Namespaces-was-Re-Behold-Pharo-The-Modern-Smalltalk-t
p4983800p4990734.html>
http://forum.world.st/Namespaces-was-Re-Behold-Pharo-The-Modern-Smalltalk-tp
4983800p4990734.html

>  

> For the moment, best best is a few tips on converting from namespaces.

>
<http://forum.world.st/Migration-from-VW-to-Pharo-td5098328.html#a5098334>
http://forum.world.st/Migration-from-VW-to-Pharo-td5098328.html#a5098334

>  

> Thanks for the links.  

>  

> Not having a quick and easy way to create namespaces in Pharo to ease
porting of VW code is my main impediment.  If Pharo namespaces existed, I
would just deal with all the other problems, and push on, instead of using
them to justify returning to VW.   

>  

> I can learn to deal with Iceberg.

>  

>  

> Shaping

 

 

Reply via email to