Hello
> It is indeed clearer.
> Previously I thought that "migrating tools to Spec2" means migrating Spec
> tools to Spec2; now I see it's rewriting all tools to use Spec2.
We cannot continue to stack layers and layers of “dust".
For P9 we will have to have a massive cleanup again.
I started to browse the code. Now we do not have enough people looking at
modularity.
> However, I still wonder about all shiny new tools being developed by feenk
> as part GToolkit 2 - documenter, coder, etc. as well as new iterations on
> inspector, debugger.
> What is their place in the grand scheme of things?
Good question. Why not but not at all prices :)
- tests tests tests
- real models because we will not rewrite all the time everything
- code review and code review
[ I do not like to have subclasses of announcer and other beast to
maintain
I was against Glamour even if I encouraged doru to do it and
retrospectively I was right :) ]
I’m not sure that we all want to code in Coder :) but we will see.
Our move is to make sure that we can move slowly in the right direction:
Bloc / Brick
Better VM
Better infrastructure (this is why I cleaned/shrunk beacon and added in
P8)
Better Tools
What we see is that Bloc deserves 6 months of work and a nice documentation.
After we will have to design widgets (or revise the ones of feenk).
There is a path between innovation and making sure that companies can
deploy robust applications (this is why GTK will help us for that)
but we should find the right path.
You see we cannot just say ok let us throw away all the PharoThings effort
and others :) because we could have follow the feenk path last year
but was the implication.
Stef
