[vlc-devel] Rust work (discussion)

Pierre Ynard linkfanel at yahoo.fr
Tue Sep 22 22:08:10 CEST 2020


Hello Lyndon,

Thanks for your earlier patches.

> Taking an incremental approach inevitably gets some Rust into end-user
> hands sooner (which is great), but yet on top of the time reviewing
> plugin conversions, additional team time needs to be spent reviewing
> all of the compatibility layer code (and possibly participating in
> its development), chasing any bugs with it, and having to spend time
> adjusting it to reflect applicable changes in the C interfaces over
> time and testing those change, for as long as it exists (until a full
> conversion is completed). Which is additional time stalling other
> enhancements, even if it is broken up over time. This doesn't apply to
> the all-in-one approach.

It's a tried-and-true approach in software project management to focus
on incrementality and compatibility, as it yields much more realistic
prospects of success and adoption.

It's not just well worth the cost. In fact, projects that depart
from this put themselves at risk of turning into a fiasco, even with
consensus and huge commitment and resources behind them: for example,
IPv6.

> Whether or not we can agree on a compatibility layer being worthwhile
> for enabling Rust in the short term, we both want Rust;

Well personally I'm not on board with that.

-- 
Pierre Ynard
"Une âme dans un corps, c'est comme un dessin sur une feuille de papier."


More information about the vlc-devel mailing list