Hi all, looks like this subject is getting a bit controversial on IRC but
nobody wants to start the discussion about this, so I'll volunteer.

The goal of this thread is first of all for Thonsew to explain in detail his
introduction of Tokens, the benefits he expect from them, and the benefits
he was actually able to *measure*. Also, when he expects his work to be
stable/production-ready, especially given the bug-fixing experience of the
last days.

Second purpose is for us to expose how those changes are affecting us, the
improvements or side-effects we've noticed. We can discuss the technical
merit of Thonsew's token framework, and hopefully have a better idea of
whether it's a good addition to Wesnoth.

--

My personal point of view:  I haven't looked in depth into Thonsew's
changes, and I suspect my C++ isn't advanced enough to undertand most of
them. All I can say is that I see the many bug reports and commits
addressing those, and I'm under the impression that the code will stay
unstable for a long time. Quite normal, you can't do large-scale changes
without bringing many issues to the surface, and introducing a flock of new
bugs. *My* main problem though is that I'm doing very sensitive UI work with
the whiteboard, trying to perfect the user experience and finally finish the
WB. There's no point trying to polish something when the underlying code is
unstable, so the question to ask is when we want to reach 1.10... it's been
a long while already.

So, did Wesnoth speed up or save memory at all with the Thonsew Tokens? Do
we keep changes, wait it out, branch, revert, mass suicide? Comment away.

Cheers,
gabba
_______________________________________________
Wesnoth-dev mailing list
[email protected]
https://mail.gna.org/listinfo/wesnoth-dev

Reply via email to