Hi Daniel, I don’t think there’s any enforcement in the OSGi spec regarding leading 0 (actually, I think it’s "blurry" as there’s no details in one way or another).
That’s why I said more a "best practice". Generally speaking (not OSGi related), I don’t see any good reason to use leading 0 in version ;) So, as the spec is not very strict about that, we should support it in the resolver. Regards JB > Le 24 mai 2021 à 17:16, Daniel Krügler <daniel.krueg...@gmx.de> a écrit : > > Am 24.05.2021 um 17:10 schrieb Jean-Baptiste Onofre: >> Hi Steven, >> >> Yeah, the state should already contain the "cleaned" versions. >> >> I will take a look. >> >> Anyway, as best practice, I would avoid leading 0 ;) > > I completely agree and I don't understand why there should be additional > work invested to "fix" that. The OSGi specification is clear that any > leading zeros in version numbers are invalid. Anyone attempting to > provide such a bundle should correct the version number and remove the > leading zeros. > > - Daniel >