|Re: [eclipse.org-committers] Installing SVN|
Hi Ed,What tooling are you using? I have no problems tracking history of renamed or moved files. Nor do I have such problems when I refactor using Eclipse. Perhaps you need an upgrade? Try the latest version of Subversive. When you upgrade, don't forget to include the "JDT ignore extensions" feature that takes care of the problems with the .SVN folders that you complain about.
Regarding atomic commits and batching. I have never done a batched check-in using multiple comments so the lack of such a feature has never been an issue for me. Using SVNKit I can check in several projects in one single transaction and that's always been good enough. My guess is that most people use one comment per commit and in my opinion its not correct to argue that batching is a requirement for atomicity. SVN certainly has atomic commits. If I check-in with SVN I can be 100% certain that if it failed, nothing happened. That's rather important when the network is sometimes unreliable.
The two main benefits of tracking directories are in my experience:1. You can compare directories. Subversive will show a pane with files that has been added, removed, or modified. Clicking on a modified file will show you what changed in the file. 2. The revision of the directory reflects the last change made beneath that directory. I.e. you can use the revision like you'd use a tag in CVS.
Perhaps the Subversive folder comparison has an issue with svn:ignore. I've never encountered that problem. I'd suggest you bring it up with the subversive people.
Regards, Thomas Hallgren Ed Willink wrote:
You win some and you loose some. SVN has atomic commits. It might seem like a trivial feature for those who rarely experience network problems but for us who do on a somewhat regular basis, it's very valuable. SVN can also revision directories as well as files. Very hard to do if you're using a file system. SVN doesn't loose track of history just because you move or rename a file. All of those features are contributed to the fact that SVN uses a database and hence, isn't subject to all the limitations you have in a file system. Revisioned data is after all not files. It is fragmented pieces of information. The CVS model is IMHO severely limited.A good summary of why SVN should be better. Unfortunately SVN or at least its tooling just doesn't work. An atomic commit requires that I can batch up numerous different commit candidates with distinct commit comments. I can't; the tooling requires a separate commit per comment. I know of no benefit from revisioning directories, but suffer numerous pains when directories have outgoing changes (e.g. svn:ignore) while their contents have incoming changes. The default Subversive connector doesn't support folder comparison to try to understand the problem. I have yet to see SVN track a rename. A Windows Explorer move or copy is treated as delete+create. An Eclipse refactor is similarly delete+create. And if you want a bit of fun, try cleaning up the bin trees after JDT has copied the .SVN folders! OK, maybe one day I'll learn to switch off auto-build in a new workspace before I do check outs. Until the tooling works, SVN causes much more pain than joy I'm afraid.The CVS tooling is just too smooth to move on.Regards Ed Willink _______________________________________________ eclipse.org-committers mailing list eclipse.org-committers@xxxxxxxxxxx https://dev.eclipse.org/mailman/listinfo/eclipse.org-committers IMPORTANT: Membership in this list is generated by processes internal to the Eclipse Foundation. To be permanently removed from this list, you must contact emo@xxxxxxxxxxx to request removal.
Back to the top