| Summary: | [backport][Import/Export] Timestamps updated on archive export and on project import | ||
|---|---|---|---|
| Product: | [Eclipse Project] Platform | Reporter: | Prakash Rangaraj <prakash> |
| Component: | UI | Assignee: | Prakash Rangaraj <prakash> |
| Status: | VERIFIED FIXED | QA Contact: | |
| Severity: | normal | ||
| Priority: | P3 | CC: | bokowski, christophe.bismuth, deboer, mober.at+eclipse, wb-rel |
| Version: | 3.5.1 | ||
| Target Milestone: | 3.5.2 | ||
| Hardware: | PC | ||
| OS: | Windows XP | ||
| Whiteboard: | |||
| Bug Depends on: | 263305 | ||
| Bug Blocks: | |||
|
Description
Prakash Rangaraj
This bug is for tracking the backport of the fix in bug 263305, which supports maintaining original file attributes and time stamps on import/export to archive files, to Eclipse 3.5.2. Boris,
The patch is available in the original bug (attachment# 124648 [details]). Need your approval for checking it to 3.5 branch.
Prakash, do all the APIs involved agree with respect to their notion of "time"? Is there a possibility to get it wrong with multiple time zones, or summer time/winter time changes? I did a quick check and according to their API docs, all the used API calls are in sync referring to "Java time". As in the patch on bug 263305: /** milliseconds since the epoch (00:00:00 GMT, January 1, 1970) */ long java.io.File#lastModified(); /** seconds since January 1st 1970 */ long org.eclipse.ui.internal.wizards.datatransfer.TarEntry#getTime() /** modification time of the entry. * Converts DOS time to Java time (number of milliseconds since epoch). */ long java.util.zip.ZipEntry.getTime() /** number of milliseconds since the epoch (00:00:00 GMT, January 1, 1970) */ org.eclipse.core.resources.IResource#setLocalTimeStamp(long); One last question from the department of the overcautious: Does this have a noticeable performance impact? If not, +1 for putting the fix in 3.5.2. (In reply to comment #5) > One last question from the department of the overcautious: Does this have a > noticeable performance impact? :-) Not that I know of. A glimpse at the patch doesn't show any. I tried few samples and didn't notice any difference. (In reply to comment #5) > One last question from the department of the overcautious: Does this have a > noticeable performance impact? Not with a sane file system, I would say, since applying file attributes should always be a lot faster than reading/writing file contents. And now that bug 296647 is fixed, I guess that the Eclipse File System is sane (again). :) +1 for 3.5.2 Prakash, can you please commit this change in time for the next M build? Patch released to Maint stream Verified in M20100120-0800 |