The creation of a new order of things

It seems to me that what you've been doing is for those with a fair amount of Smalltalk skill already. Those of us who only mess around with it in our off hours (it isn't on the approved software list at work) are not going to be very helpful in such an effort.

Still, I have pointed to your work as an example of why I think Smalltalk is still relevant. I had showed someone VW and Squeak. Squeak, with all the hideous colors, is a turn-off to most adults. VW's GUI, especially on Linux, looks like it hasn't been updated since the Win 3.1 days. Pointing to your Webkit and former XUL work enabled me to say that those issues are being resolved.

In the end, you have to follow your own path, but don't feel that no one appreciates your efforts. I do, and I'm pretty sure Jim does, since he toots your horn regularly.

Hi, Antony,

I like your work very much, and I was astonished by the fact Cincom never looked at it.

For example, your improvements on the L&F for OSX make usable the Mac, giving a decent look. The original L&F of VW is really ugly and, on a platform where elegance is everything, bad Look is really a non sense.

The idea of WebKit is being investigated not only in VW, and it can be an example of a possible path to follow, etc.

I suppose Cincom is looking only at the Seaside stuff lately (not to say this is bad, but I think what you wrote about your Ruby experience should make someone thinking... I looked at Ruby years ago, and now it seems about to be mature...), and, as I noted during the years, they not allways take into account good ideas, and loose some on the way. This can be a resource problem, or just an error, because errors happen, or a different point of view.

I agree on the fact that your kind of stuff needs support from the producer, and this is not here.

If you think also at the work done by customers on the MAC VM and not included yet, you see you are not alone.

I would like to add something more, and quite rude: Cincom, it seems to me that sometime you never look outside of your room. Not invented here seems sometime a problem on your developement group. I hope to be wrong, because VW is still my preferred tool as a programmer.

Regarding the fact that VW is not a comunity for developing OS project, this seems true, expecially compared to Squeak (where sometime there are to many, also overlapping). this could be perhaps a problem of numbers.

Hi Giorgio,

Thanks for your kind words. Without being rude: I know you commented on some of my posts previously - could I ask if you considered contributing to any of my projects? I'm really interested in why these projects in particular failed to get any traction, and after all we are the community!

I must also say that I'm not 'blaming' Cincom in any way, and I'm particularly not concerned that they didn't 'pick-up' any of my work. They have a very difficult tactical problem (as I point out in a later post), and I know that the people I have spoken to at Cincom are well aware of it. Personally, I blame rampant monkey-patching :)

Well, I've been following your posts with interest. LLVM is very interesting technology, and I'm very interested in any attempt to use it with Smalltalk. It seems more suited to statically-typed languages, though. I haven't been able to figure out how one might use it to implement PICs, for example, which seems crucial to making Smalltalk fast. I was kinda hoping you'd come up with something clever. ;-)

The look and feel stuff is nice, and I've been using ASBIcons in VW images since it's easier on the eyes than the usual icons. 

I doubt you'll get many people interested in a file-based Smalltalk. Non-Smalltalkers will find the syntax weird, and Smalltalkers will miss their browsers. I agree that the VW tools can be annoying, but I think a better tack would be to fix them rather than rejigger things so you can use TextMate. 

Two packages that I find indispensabe are NoCutDelete and WindowsFaithfulPaste, which are available in the public Store repository. They make the VW text editor tolerable if not good. 

I'm not a big fan of Store either. I'm getting pretty close to releasing Monticello 2.0 for Squeak, which I'm hoping will get ported to VW at some point. That would be a big improvement over Store, and much easier than figuring out how to flatten things into files so they can be versioned in Git or whatever. 

I do like your instinct to try to bridge between Smalltalk and big pieces of useful code from the rest of the world. The Smalltalk community often gets caught up in trying to reimplement everything "our way," instead of just taking advantage of what others have already done. 

I think integrating WebKit and LLVM are wins - they both represent lots of effort by smart people to solve ugly hairy, problems and lots of feedback from the real world. Doing the equivalent in Smalltalk just won't happen because it's way more effort than the community can generate.

Making VW file-based is bound to fail, though, for the same reason. The file-oriented tool-makers don't support Smalltalk (with, eg. syntax highlighting), and the Smalltalk community doesn't deal with files. With a hybrid approach, you can't take (full) advantage of mainstream efforts, *or* the more modest efforts of the Smalltalk community. The community you can draw on for support is almost non-existent, and you'll end up doing it all yourself.

Anyway, I've been watching your progress with interest even though I think some of your ideas are a bit quixotic. :-)

Well, I've been following your posts with interest. LLVM is very interesting technology, and I'm very interested in any attempt to use it with Smalltalk. It seems more suited to statically-typed languages, though. I haven't been able to figure out how one might use it to implement PICs, for example, which seems crucial to making Smalltalk fast. I was kinda hoping you'd come up with something clever. ;-)

The look and feel stuff is nice, and I've been using ASBIcons in VW images since it's easier on the eyes than the usual icons.

I doubt you'll get many people interested in a file-based Smalltalk. Non-Smalltalkers will find the syntax weird, and Smalltalkers will miss their browsers. I agree that the VW tools can be annoying, but I think a better tack would be to fix them rather than rejigger things so you can use TextMate.

Two packages that I find indispensabe are NoCutDelete and WindowsFaithfulPaste, which are available in the public Store repository. They make the VW text editor tolerable if not good.

I'm not a big fan of Store either. I'm getting pretty close to releasing Monticello 2.0 for Squeak, which I'm hoping will get ported to VW at some point. That would be a big improvement over Store, and much easier than figuring out how to flatten things into files so they can be versioned in Git or whatever.

I do like your instinct to try to bridge between Smalltalk and big pieces of useful code from the rest of the world. The Smalltalk community often gets caught up in trying to reimplement everything "our way," instead of just taking advantage of what others have already done.

I think integrating WebKit and LLVM are wins - they both represent lots of effort by smart people to solve ugly hairy, problems and lots of feedback from the real world. Doing the equivalent in Smalltalk just won't happen because it's way more effort than the community can generate.

Making VW file-based is bound to fail, though, for the same reason. The file-oriented tool-makers don't support Smalltalk (with, eg. syntax highlighting), and the Smalltalk community doesn't deal with files. With a hybrid approach, you can't take (full) advantage of mainstream efforts, *or* the more modest efforts of the Smalltalk community. The community you can draw on for support is almost non-existent, and you'll end up doing it all yourself.

Anyway, I've been watching your progress with interest even though I think some of your ideas are a bit quixotic. :-)

I was very interested in your WebKit work and the look and feel stuff too, but for some reason I was under the impression that these things were being done for a private project and that they wouldn't be available to the community.  After a while I stopped paying attention.  Sorry if I got that wrong.

Carl: it's all Open Source. I personally have no commercial need for those features - it was just my attempt to improve VW as a product. I made that point in a couple of posts, but I guess I should have said it more often.