On 27 August 2015 at 22:52, Lieven Govaerts <[email protected]> wrote: > On Wed, Aug 26, 2015 at 8:52 PM, Ivan Zhakov <[email protected]> wrote: >> Status update: >> ASF INFRA team performed test import for us: >> https://issues.apache.org/jira/browse/SERF >> >> The following data was converted: >> 1. Summary >> 2. Reporter and current assignee >> 3. Comments with original date and author >> 4. Labels >> 5. Issue status and resolution > Looks good. > Cool!
> I liked the difference between "New" and "Accepted" in the old repo, > is that still possible? Seems like "ACCEPTED" is available in Jira, > atleast in the filter. > We need custom workflow for this. I prefer to stick we simple standard workflow at least for migration. >> 6. Issue creation and last update date >> >> The following data was *not* converted: >> 1. Attachments -- I didn't find easy way to download them. I'm also >> not sure that JIRA supports importing attachments. >> 2. History of changes. >> Things I'm going to do for final migration: >> 1. Add footer with original comment author for unmapped users. Done. >> 2. Investigate ways to migrate issue history (source data is not easy to >> parse). Done. >> Bert suggested to remap revision numbers to match ASF repository >> revisions. But there is a risk that conversion replace something else >> that match rNNN pattern. So I'm inclined to leave them as-is. > > You could add a "(asf repo: [rNNN+offset])" right behind the old > number. But I don't think it's that important, these revs are > typically used in issues that are resolved anyway. > It could be possible, but there are still risk to mangle original comment. I still think that effort doesn't worth the result and inclined to leave it as is. I've asked INFRA team to perform another migration for us with improved import data. I hope it will be final import. -- Ivan Zhakov
