Hi! > >> With the current mainline code, if we so want, we can just rm > >> panel-dsi-cm and implement it and the related DSI code from scratch > >> using the DRM's panels. That's a lot easier than going from panel-dsi-cm > >> to DRM panels gradually, step by step, changing all the components in > >> the pipeline piece by piece and keeping things working after each commit. > > > > Eek, that sounds like a "flag day" plan: > > > > https://en.wikipedia.org/wiki/Flag_day_(computing) > > > > Trust me, we're better off keeping things working for every > > commit. It's good to have people using and testing the code. > > I don't quite see the flag-day comparison. Manual update has never > worked. It is essentially a new feature. Do we want to add a new Manual update works for us, today, and you keep stalling it because some restructuring might take place in the future. If the restructure is ready, please commit it _now_. If it is not, can we get our displays working, please? > right when we're doing a major restructuring, knowing that the new > feature needs to be partly or even mostly rewritten during the > restructuring? Or do we delay that new feature and implement it after > the restructuring? The feature already works. We need it for our devices to be usable. > I believe it will save everyone's time quite a bit if manual update is > added after the restructuring. The downside is that we won't have this > feature until after the restructuring is done. How does it save anyone's time? Same ammount of work still needs to be done. Pavel -- (english) http://www.livejournal.com/~pavelmachek (cesky, pictures) http://atrey.karlin.mff.cuni.cz/~pavel/picture/horses/blog.html
Attachment:
signature.asc
Description: Digital signature
