Save Changes - erroneous meta-data

Question

I'm impressed with the Salesforce Development cookbook, but one thing I can't get my head around is how you manage some of the meta-data that the Dev Save Changes plan will pull down (maybe I'm missing something). Most package.xml files will contain profiles - for my org - when I currently fetch metadata using Ant (manually) it pulls down so much info including some inclusions that we can't deploy (but salesforce seems to add) like Social-Persona info - that I need to manually remove or exclude from GIT. How might you handle this in an automated workflow ?

craig.bradshaw · Answer

Thanks Mark. Interesting reading. I'll have a play about with that some of what Alex suggests.

I've pulled together the meta-data retrieve which works well. Always seem to get unexpected/different results when using tools like MavensMate - the Ant retrieval works well tho.

Thanks pointing me a that blogg.

Mark Klinski · Answer

@craig.bradshaw, Profiles are indeed a bit tricky. As such, we've decided to not to track Profiles in our metadata repos currently (ie. our production package.xml doesn't include Profiles). There are some creative approaches to dealing with Profiles, though. For example, Alexander Brett has some code for Salesforce deployments that deals with Profiles a bit more gracefully.

Nic Brough -Adaptavist- · Answer

Ah, ok, that makes good sense.  I'm afraid I'm not qualified to tell you much beyond that I know several places that use Salesforce and have adopted some or all of the stuff you've seen in the video.

What are the issues you are having with Bitbucket and Bamboo though?