Hello Don,
Stewart Gordon wrote:
John Reimer wrote:
<snip>
2) You don't have enough information to go on to make that change,
unless you have dicussed this with Walter. Maybe he doesn't want to
recognize the old DWT as "standard" anymore, if it isn't actively
developed. Maybe neither library is "standard".
<snip>
Maybe you're right. But if Walter hasn't stripped Phobos DWT of its
'standard' title, I think technically it still applies. But it might
be more practical to consider neither to be.
I don't think that it was _ever_ accepted as "the standard library".
It's more that it was a proposed standard, and it died almost
immediately after being proposed for standardisation.
That's also possible. But then we have no idea since Walter didn't say anything
more about it. It's all conjecture.
Also, here is the reason the project died soon after being proposed. The
reason it /looked/ like his "blessing" nixed the project was because he proposed
it before having communicated with the pototential developers to find out
how involved/committed they were with the process. Instead, he just announced
it... and shocked a couple of us out of our skins :). As far as I knew
(since I was contributing to dwt at the time) none of us had enough time/skill/energy
to see the project through to completeion -- with the exception of Kris (and
maybe a couple others) who had loads of skill but perhaps little time, since
he was working hard on other important projects. We were mostly feeling things
out in the background to see how far we could go with the idea. Meanwhile
Kris had tried to help the process by contributing a Java to D converter
that he had thrown together to automate as much of the SWT conversion as
possible; he also convinced Walter to support internal classes and anonymous
classes in DMD (I think; I can't remember which or both: Kris made many important
proposals based on practical stress-testing of the language/compiler some
of which were ignored)... Also the compiler was still quite buggy making
it sometimes exasperating to work on large projects like this one.
Although there was some excitement behind the project, there was sizeable
lack of motivation: it was very easy to get people interested in the idea,
but practically impossible to find people who were willing to contribute
to it (this is still the case with new dwt, but to a much lesser extent since
the community has grown since then). Unfortunately Walter gave his blessing
without determining or knowing this. When the project crashed after this,
he apparently felt he had caused it to do so; he had not. Thus, I don't
blame Walter at all for its demise, although I think he would have avoided
some of the pain if he had chatted with a few of us first. Regardless, nobody
lost completely on that one because, awhile later, Shawn Liu took up the
guantlet from where we left off (which means he did a lot of work) and completed
the windows port. It enjoyed success on that platform for a time, but never
progressed to the cross-platform initiative.
Back to the present. Again, it would be easier if we just fix this situation
by changing the "dwt" newsgroup to "GUI" and forget about the reference to
"standard" for now. The time to "standardize" a GUI library is perhaps when
a project has proven its survivability and popularity enough to warrant the
title. Even so, GUI's are going to be particularly controversial, so it
may be wise for D to avoid standardizing any such thing for awhile. Just
as there are people that don't like the Tango "style" (a very /few/ people,
of course ;) ), even so there are going to be people that don't like dwt.
Finally, the original purpose of standardizing DWT was to help promote D.
I think both D and DWT have matured to the point that they can both achieve
that purpose without resorting to "standardization" of a GUI.
-JJR
