On Fri, Jan 30, 2015 at 10:11 AM, Jason Ji <jason.y.ji(a)gmail.com> wrote:
Thanks for your feedback. To clarify a bit, we're
not thinking of using
LiquidThreads as it is - we have a different extension we will be building,
with some different needs than LQT has. For example, we may not need any
integration with watchlists. So our thought is that we might fork LQT and
modify it to suit our needs. We're still very early in the design phase.
The bad part of LQT is not about interaction with watchlist. It will be
essentially untouched by any trimming short of complete rewrite.
Max - when you say just use Flow, do you mean we
should fork the Flow code
base and work from there, or that we should just install Flow? Flow looks
interesting, but we're not sure it will have the features we need, and our
timeframe is likely to be shorter than the timeframe of Flow development.
If you fork something, you will have to maintain it forever - why not put
the same effort in contributing to mainline instead? And Flow is quite
complete for most use cases, and its team is mostly working on adding
support for various crazy workflows user communities have created in more
than 10 years without a good discussion system. I don't think you need to
wait for these.
Is there somewhere I can go read in detail about the
bugs and unfixable
problems with LQT? We might not fork LQT at all, but we were also thinking
of using wiki pages to store comment text. So if that idea is fundamentally
broken, it would be great to know why.
I already explained why, bugs are here:
https://phabricator.wikimedia.org/search/query/ojED3mdcIKDQ/
--
Best regards,
Max Semenik ([[User:MaxSem]])