Releasing Fest Assert 2.0 soon ?
Joel Costigliola
I think we should plan Fest Assert 2.0 release to show that this
project is active.
IMHO, Fest Assert 2.0 release should :
- be feature complete compared to Fest Assert 1.4 (the remaining being
postponed to 2.1)
- have a user guide with lot of examples and how to (e.g. how to write
custom assertions)
It would be a good thing to plan regular releases (every 2 months ?)
even if they are small, like jenkins project.
What do you think ?
Cheers,
Joel
Ansgar Konermann
Hi Joel,
first things first: thank you very much for the constant ongoing effort
you invest into FEST. I wasn't able to contribute nearly as much as you
did over the last months. This is great work.
Regarding the 2.0 release:
My main concern with the 2.x is that this is *the* chance for
introducing breaking changes into the API. Are all the planned/desired
breaking changes done? Are the core APIs (internal vs. API package,
extension mechanism) stable?
Skimming the source code, I see we have no "self type emulation" in the
2.0 line. This means we'd have to use a lot of co-variant overrides to
ensure a working fluent API. We had this self-type thingie in 1.4, so we
should use it in 2.0, too. Otherwise extensibility is seriously hampered.
So much for now.
Ansgar Konermann
> Regarding the 2.0 release: [...]
Completely bollocks.
Sorry, I accidentally pulled from the wrong repo - my fault. Will have
to look tomorrow. Obviously too late today ;-)
Best regards
Ansgar
Joel Costigliola
I think we should ask Alex and Yvonne if they have planned breaking
changes not yet in the code base.
Alex, Yvonne can you tell us ?
Another point I would like to mention is that the actual documentation
site is horridly slow (30s to display Fest assertions doc), that is
just not acceptable for our users (it also makes me mad when I want
to update the doc).
-> It is maybe time to move our documentation to Github.
Having said that, why not go further and move all the project to
Github (I think of bugtracking but I haven't played a lot with it so
we need to evaluate it).
Any comment welcome !
On Oct 21, 12:37 am, Ansgar Konermann