My main worries were about silently breaking the code of existing users, who may already jave that kind of ugly conditionals.
If there is confidence about that won't happen (nobody has expressed an opinion against in the ML, so we can assume that), then I'm ok with having this change in 1.6.x too. Thanks for clarifying Aled! El 16/09/2013 01:47, "Aled Sage" <[email protected]> escribió: > We certainly need it urgently in the 1.6.x branch. If it's not in until > 1.7 release, then we'll work around it with an ugly `if > (provider.equals("softlayer")` block to change MBs passed in to the GBs it > currently expects. Anyone else who's using generic jclouds code to run > against softlayer + other clouds will have to do the same (or have > inconsistent units in their config files or some such). > > My gut feel is that pain for new users will be greater than the pain to > existing users of fixing it - new users will all get it wrong if they > follow the standard jclouds examples (i.e. they use MBs); existing users > have been alerted via the mailing list (and it can be in the release notes). > > Also note that https://github.com/jclouds/jclouds/pull/77 (which fixes > https://issues.apache.org/jira/browse/JCLOUDS-163) means softlayer is > probably unusable in jclouds 1.6.x currently, so there might not be > "existing users" for 1.6.x. > > --- > Reply to this email directly or view it on GitHub: > https://github.com/jclouds/jclouds/pull/147#issuecomment-24483258
