Rockbox mail archiveSubject: Re: Haxx will stop hosting most Rockbox services after 2020
Re: Haxx will stop hosting most Rockbox services after 2020
From: Solomon Peachy via rockbox-dev <rockbox-dev_at_cool.haxx.se>
Date: Sun, 15 Mar 2020 09:16:23 -0400
On Sun, Mar 15, 2020 at 11:13:36AM +0100, Baptiste BARON via rockbox-dev wrote:
> *Forums : Same as above, dump the database and mirror the forums on a third
> GitHub branch (github.com/Rockbox/rockbox-forums). We might ensure a
> transition with an update later on.
IMO, migrating the existing forums is a largely pointless affair; nuke
'em from orbit and create something new, if at all. :)
> * Gerrit : Is it easy to setup once the main repo is GitHub ? What is used
> for the CI ? Can the CIs be migrated as well ? May we see the code somewhere
AFAIK using gerrit means gerrit has to host the code, period. Gerrit's
flow is largely incompatible with a GH-style pull-request model.
It's not agreed upon that keeping Gerrit is worth it overall, but
migrating the existing patch backlog is likely to be a challenge. Some
(most?) of it is irrelevant but a lot of it isn't.
The CI/build system is home-grown, its code resides on git.rockbox.org,
and isn't really compatible with any existing CI flow. I really don't
think it's worth rewriting, if only because it's a decent amount of
work, what exists today works very, well, and the bulk of the builders
are maintained by currently-active developers.
> * Flyspray : Development for Flyspray has resumed some time ago, most
> up-to-date version is 1-0rc9, per the website. They also have a GitHub repo
> as well. Can we see the code for the current FlySpray used for Rockbox, so
> we can test migrations ?
Speaking from experience, migrating to the current version of flyspray
is likely to be straightforward. The question is whether or not to just
nuke it from orbit or not, as most of the issues in there are so old to
be of little relevance today.
* * *
All that said, at the end of the day, talk is cheap, and it's going to
take someone stepping up to do the actual work.
IMO, from a "bang-for-the-buck" perspective, it makes more sense to
cut-n-paste the existing tooling off of the haxx infrastructure onto
sometihng else. After everything is consolidated, we can determine what
tooling makes sense to rework without a dangling sword overhead.
I'm still willing (and able) to do this, including hosting everything
but the forums, but I won't have the personal bandwidth until April.
-- Solomon Peachy pizza at shaftnet dot org High Springs, FL ^^ (email/xmpp) ^^ Quidquid latine dictum sit, altum videtur.