[Date Prev][Date Next] [Thread Prev][Thread Next] [Date Index] [Thread Index]

Bug#786895: lintian: incompatible-java-bytecode-format warning needs update for Java 1.7



On 05/26/2015 11:49 PM, Niels Thykier wrote:
> On 2015-05-27 08:09, tony mancill wrote:

[...]

>> However, couldn't we use versioned build-deps on default-jdk + the
>> virtual runtime dependency for the binary package to accomplish the same
>> effect?  For any software that requires Java7, we would build-dep on
>> default-jdk (>=2:1.7~) and require java7-runtime for the resulting
>> binary package.  (The versioned build-dep has already been used in some
>> cases to excluce gcj-jdks.)
>>
> 
> STOP! Versioned dependencies on default-java for pulling a particular
> Java version IS A BAD IDEA!
> 
> We have twice now added an epoch to default-jdk et al, because we had to
> revert the default java on several architectures.  So far it has always
> been openjdk -> gcj, but it could just as well be openjdk-8 to openjdk-7
> next time.
>   As soon as the epoch is bumped, you get the old java version while the
> dependency is still satisfied!

Ah yes, thanks for pointing this out.  This seems like a generally
serious unfortunate consequence of epochs.

> To be honest, I mostly of the belief that using the java major version
> as a part of the version for default-jdk et al. is a misfeature.  It
> appears to trick people into believing it is useful method to ensure a
> minimum java version, only to be surprised by an "oops, we have to
> revert to gcj and have to bump the epoch".
> 
>> For Java8, the pattern would be the same, and it won't be long before
>> we'll have to accommodate sources that *require* Java8.  The versioned
>> dependency would allow us to differentiate between source packages that
>> require 8 (or 9) and those that will still build with Java 7, etc.
>>
> 
> No, I am not convinced it will (as per above).  Perhaps we should have a
> look at versioned provides as an alternative.

I didn't propose this because I thought we couldn't use a versioned
dependency on a virtual package, but it appears that dpkg 1.17.11
provides some support.

I'll move future discussion to debian-java, since as Markus pointed out,
this bug report is regarding the lintian warning.

Thank you,
tony

Attachment: signature.asc
Description: OpenPGP digital signature


Reply to: