On Thu, Oct 27, 2016 at 10:48:24AM -0400, Josh Boyer wrote: > My concern isn't "how do we get Bikeshed tested". It's "how do we > make sure Rawhide continues to be tested if Bikeshed exists and > promises to be somehow a more stable rawhide". Simple: it probably wouldn't. > Rawhide testing today is very minimal. If Bikeshed is a thing that > appears to be the sweet spot that is essentially a stable rolling > release, what purpose does rawhide then serve to the average > user/tester? For that matter, I'd actually be concerned that it would > pull testers off of the actual releases as well. The average tester/user would prefer Bikeshed to Rawhide; Rawhide would just be a raw source of possibly-not-working-together packages. Hopefully, the increase in usefulness/stability would increase the number of early testers overall. I'm not -- yet -- worried about drawing testers from the actual releases simply because the number of people using rawhide right now is so low — about a fifth of what we see for betas and a fraction of a percent compared to people who run the releases. I'm also fine with the idea of instead making Rawhide itself the more-carefully-gated thing (maybe with the automatic side-tag idea), but last time we talked about this there seemed to be some preference for leaving Rawhide as it is and making a new thing. I don't really care as long as we call *something* Bikeshed. -- Matthew Miller <mattdm@xxxxxxxxxxxxxxxxx> Fedora Project Leader _______________________________________________ devel mailing list -- devel@xxxxxxxxxxxxxxxxxxxxxxx To unsubscribe send an email to devel-leave@xxxxxxxxxxxxxxxxxxxxxxx
