[libre-riscv-dev] Tests in soc/fu

[Cole Poirier]

On Jun 21 2020, at 9:32 am, Luke Kenneth Casson Leighton <lkcl at lkcl.net> wrote:
> i mean that the load on the server is incredibly high.
> 
> for those 8 seconds the loadavg is up at *1.5*, sustained.

Hmm that's odd, I would have thought since it's in js it would have been
very heavy client side not server side... Regardless, I'm confident I
can write something that sends a minimum number of requests fetching the
minimum necessary state.

> if two or more users try the same page simultaneously, particularly during
> meetings, the server is in deep doodoo.
> 
> it is only a vCPU with 1GB RAM.
> 
> i would very much like to see us using a distributed bugtracker.  the only
> one i know of is bugseverywhere and it does not have a writeable web UI
> (only a readonly one).
> 
> plus we would have to transition from bugzilla.
> 
> which sigh has to happen at some point because we cannot rely on
> centralised internet-connected services.

Indeed, can this be done through git, or
conflict-free-replicated-datastructures? If you can send me the prior
art you think we should use or build upon that will be very helpful.

Cole