[squeak-dev] Re: Thinking about Environments

David T. Lewis lewis at mail.msen.com
Fri Dec 19 23:50:10 UTC 2014

On Fri, Dec 19, 2014 at 11:06:38PM +0100, Herbert K??nig wrote:
> Hi,
> Am 19.12.2014 um 22:18 schrieb Chris Muller:
> snip..
> >The audience I want to include into the Squeak community are not just
> >developers but "users" too.
> ...snip
> considering myself a Squeak user I theoretically see the benefits of 
> environments, but assume I'll rarely need them. So yes, leave them in. 


I don't know how much I will make use of Environments, but from everything
that I have read and seen, this is a simple and intelligently designed
approach to the problem. I also recall that it was enthusiastically supported
by Andreas Raab, which to me is a very strong endorsement of the approach.

My sense is that Environments is 98% done. We should make every effort to
ensure that the last 2% gets finished, and that the Environments system is
put to some good use. By that I a mean a working example of using it in
some useful way, no matter how big or how small.

> Besides being helpful when using foreign libraries I too use some names 
> only to avoid conflict with existing classes.
> Practically only this week on the newbies list two people have been 
> bitten by the fact that environments broke project save and this hasn't 
> been fixed.
> I remember other people stumbling over this before and I keep using 4.4 
> for development and only use 4.5 for deployment because I use projects 
> for documentation.

I think you are exactly right. And IIUC, the issues with project save are
related to a short list of issues that Levente has been identifying. Sorry
I don't have a link right now, but Levente has been asking for help on this
for quite a while. So we need to address the "Levente list" of issues and
get Squeak back to the point where project saving works properly.

>From my point of view, if we could do two things I would be happy:

1) Fix the "Levente list" and get project save working. I think that's
the 2% that still needs to be done for Environments.

2) Use environments for something. Maybe just a hello world example to
illustrate, or maybe something related to a real project. I don't care
what it is, I'd just like to have a working example that I can look at
for reference.

Probably other people will point out that more tool support is needed,
but personally I would be happy with just these two things.

> Craigs approach seems plausible too but it's not yet main Squeak.

I think it is more than just plausible, it is potentially quite revolutionary.
But is a different set of ideas than Environments, and I don't think that
either one needs to exclude the other.


> Cheers,
> Herbert

More information about the Squeak-dev mailing list