How good is the binary diff in Jazz SCM?
We are delivering binaries as part of our build and wanted to know how good the binary diff is in Jazz SCM. For instance, if I am checking in a zip file and all that changes are modified dates in the zip file (ie the build was compiling the exact same source as the previous build), how much of the new binary needs to be saved in the checkin compared to the original?
The reason I ask is that if jar/zip files are not handled efficiently I have the (second best) option of checking in expanded JAR files which would only need to commit individual changed classes rather than the whole jar/zip file. Then at deployment time, the class folder can be re-JAR'd. Best Regards, Chris. |
5 answers
On Wed, 19 Nov 2008 12:08:01 +0000, chris.melikian wrote:
We are delivering binaries as part of our build and wanted to know how If you have delta compression enabled then you should be getting very high compression ratios if the only things that differ are the mod dates in the zip file. - Dmitry |
I have the default values set in the advanced properties Possibly, it depends on how different the files really are. You can try using the repo size reports to see the effects of delta compression. - Dmitry |
Thanks Dmitry. I have the default values set in the advanced properties section. Are these sufficient for say a 1% difference in a 10MB zip file?
Best Regards, Chris. |
Thanks Dmitry, but it seems I'd need to get my manager to pay for the Standard edition as reports aren't included in the Express edition.... :-)
I'll see if I can achieve the same test using a DB2 client. |
chris.melikian wrote:
Thanks Dmitry, but it seems I'd need to get my manager to pay for the FYI, some limited reporting (in the form of some trend reports in the dashboard only) is now available in Express in 1.0.1. james Jazz Reports Team Lead |
Your answer
Dashboards and work items are no longer publicly available, so some links may be invalid. We now provide similar information through other means. Learn more here.