Can you elaborate why? Are you concerned about the implied complexity or anything else?
On Thu, Oct 8, 2015 at 2:09 PM, Bill Farner <[email protected]> wrote: > I do indeed think the scheduler should not be doing this type of diff > operation, and it should be up to an API consumer. > > On Thursday, October 8, 2015, Maxim Khutornenko <[email protected]> wrote: > >> This isn't about JSON in the context of ExecutorConfig. The ticket >> assumes converting the entire TaskConfigs into JSON and then comparing >> them with the help of JSON diff library. I should have made it clear. >> >> Does this answer your concern or do you think we should not provide >> TaskConfig diff details at all? >> >> On Thu, Oct 8, 2015 at 1:59 PM, Bill Farner <[email protected] >> <javascript:;>> wrote: >> > I thought JSON diffing was scrapped? I really don't think we should do >> > this. >> > >> > On Thursday, October 8, 2015, Maxim Khutornenko (JIRA) <[email protected] >> <javascript:;>> >> > wrote: >> > >> >> Maxim Khutornenko created AURORA-1517: >> >> ----------------------------------------- >> >> >> >> Summary: Decide on TaskConfig diff library for the >> >> getJobUpdateDiff API >> >> Key: AURORA-1517 >> >> URL: https://issues.apache.org/jira/browse/AURORA-1517 >> >> Project: Aurora >> >> Issue Type: Task >> >> Components: Scheduler >> >> Reporter: Maxim Khutornenko >> >> >> >> >> >> Need a JSON diff library to compare TaskConfig instances. See design doc >> >> for more details: >> >> >> https://docs.google.com/document/d/1Fc_YhhV7fc4D9Xv6gJzpfooxbK4YWZcvzw6Bd3qVTL8 >> >> >> >> >> >> >> >> -- >> >> This message was sent by Atlassian JIRA >> >> (v6.3.4#6332) >> >> >>
