Community technical support mailing list was retired 2010 and replaced with a professional technical support team. For assistance please contact: Pre-sales Technical support via email to sales@march-hare.com.
Matthew McClellan wrote: >>> is the bug id a string or a number? >> >> As Tony indicated it is stored in the RCS file as an ID, so it could >> be anything. However at various points in the code it is treated as >> an integer - so best to stick to numbers only. The 'message to >> bugid' stuff in the current commercial release that should be in the >> next CVSNT release also treats it as an integer (I think some of the >> patches I submitted to TortoiseCVS may also treat it as an integer). >> >> It would have been 'better' to treat it as a string all along, and >> the RCS format gives us that option in future - it's particularly >> tricky to integrate to bug systems like JIRA that use a 'prefix' on >> the bug number - however it does have it's advantages too (even in >> Jira it really is a bug number with a project prefix and treating >> the bug number as a bug number means it can be used cross project). Yes, but I'm not sure what in the case of Jira using a number cross project would do good -- issue "PJA-123" is likely to be about something completely unrelated to issue "PJB-123" :) So if you have code related to several tracker projects in the same repository, that can get confusing. > I just recently started using the "Mark Bug" feature, and we happen to > be using JIRA for most of our development issue tracking. In our > current environment we have a few separate projects in JIRA whose > code is all managed in the same CVSNT repository - running > 2.5.04.3226; this particular senario needs that prefix to be included > in the "Bug Number" textbox on commit - otherwise there would be no > way of matching that Bug to the correct JIRA project. Anyway, so far > things have worked great; the CVSNT audit database correctly stores > the entire string in the BugId column, the commit messages are > consistently correct, TortoiseCVS (from March-Hare) must be sending > the Bug Number correctly too, and even CVSMailer can find the Bug > comment and add the correct link to the JIRA issue in the email > messages. Thanks for the success report from the trenches :) Gerhard