That's not going to work with AUTOREV even if you don't have nobranch=1 there, but it will make SHA1 builds work since the sha1 is still accessible.
On 21 June 2018 at 12:03, Martin Jansa <[email protected]> wrote: > And how is it going to test latest revison in e.g. 5.10.1 branch which > you've removed in upstream? > > On Thu, Jun 21, 2018 at 7:14 AM Samuli Piippo <[email protected]> > wrote: >> >> On 20 June 2018 at 15:19, Martin Jansa <[email protected]> wrote: >> > On Wed, Jun 20, 2018 at 02:57:39PM +0300, Samuli Piippo wrote: >> >> On 20 June 2018 at 14:29, Martin Jansa <[email protected]> wrote: >> >> > 3) we cannot use nobranch=1, because that breaks AUTOREV >> >> >> >> How does it break AUTOREV? >> >> We are using nobranch=1 when testing non-released Qt versions and >> >> haven't seen issues with it. >> > >> > And are you using AUTOREV? How is it supposed to know which branch to >> > track? >> >> Had to test this myself and I still don't see problem here. >> As long as there still is "branch=${QT_MODULE_BRANCH}" in the URI, >> nobranch=1 doesn't seem to make difference for AUTOREV. >> It only stops the SHA1 validation, when you do have SHA1 in SRCREV. -- _______________________________________________ Openembedded-devel mailing list [email protected] http://lists.openembedded.org/mailman/listinfo/openembedded-devel
