[Labs-l] A (21) day in the Labs

Wed May 1 23:46:47 UTC 2013

Ryan Lane <rlane at wikimedia.org> wrote:

> [...]

> Things take time. Having the WMF behind Labs doesn't mean things will
> happen faster, but that it'll be properly resourced from a hardware
> perspective, will have adequate levels of funded staff and will fit into
> our infrastructure in a way that's manageable. It also means things will be
> puppetized, documented, and open to be run by the community.

> Our current goal is to have the replicated databases completely ready for
> the tools labs project by the Amsterdam hackathon. Ideally it'll be
> accessible by all projects by that time.

If time was standing still, we would have bigger problems.
Of course "things take time", but the amount of time varies
significantly between different operators.  Outside of mo-
nopolies like WMF, efficiency is a major factor in competi-
tiveness.

On Tools, there were months of planning to set up replica-
tion.  Then we get one state of the union
(cf. http://permalink.gmane.org/gmane.org.wikimedia.labs/926),
only to be denied the next day
(cf. http://permalink.gmane.org/gmane.org.wikimedia.labs/935).
Paid personnel, working for the same (rather small) organi-
zation, with no language barriers.

The "Database plan" has an item "Write schema review tool".
Even though Marc has marked this as "Done"
(cf. https://wikitech.wikimedia.org/w/index.php?title=Tool_Labs/Database_plan&diff=68306&oldid=65869),
the actual code
(cf. https://gerrit.wikimedia.org/r/#/admin/projects/operations/software/redactatron)
still lacks anything that generates triggers.  But not only
because http://xkcd.com/1205/ was published Monday, the big-
ger question is: A "schema review tool" is a *blocker* for
replicated databases?  Really?  We need some fancy UI for
schema changes that happen a few times per year at most?

This suggests that there is a lot of unnecessary friction at
WMF.

> If folks work with us rather than fight with us, we'll get things done
> faster, together.

I haven't seen evidence that any such correlation exists in
WMF software development.  On the contrary, contested fea-
tures were finished on time against considerable opposition,
while for example the community invested much energy in
string manipulation functions and had to wait years for Lua
to come along.

Tim