New issue
Have a question about this project? Sign up for a free GitHub account to open an issue and contact its maintainers and the community.
By clicking “Sign up for GitHub”, you agree to our terms of service and privacy statement. We’ll occasionally send you account related emails.
Already on GitHub? Sign in to your account
Shredder: Can't seem to make "Render script from Selected" work with partial selections #375
Comments
It should also work when you select the root node and generate from that. Or when you generate from a directory that is self-contained, i.e., all originals and duplicates are in it. Anyways, there was some buggy behavior mixed with surprises. Here's what's happening: When you click Generate from Selected |
Oh, it did, but that's just "Render script from All", so I already had a way to do that. 😄
Sure will, re-running that compare now. Thanks! |
Things are greatly improved with the new version! I can now make partial selections of individual files or directories, and generate a script processing only those files. Thanks for the quick attention on that! There are still two areas where things are a little different than might be expected, though they're not critical at all compared to the basic selected-items functionality which is now working well.
Other than those two aspects, everything seems to be working exactly as expected. I'll play with it a bit more just to see if I encounter any other unexpected results. Thanks again! |
Good, good. To the other points:
|
Fair enough.
Not just a folder, but individual files as well. Deselect them, when they were previously selected, and if any of their parents are selected they'll still be "stealth" included in the selection. (I mean, it's not "stealth" in the sense that the parents are still visibly selected, but... it's at the very least counterintuitive. Like I said, you explicitly deselect something, you expect it's removed from the selection.
From an implementation standpoint, I can see that. But I've thought about it a bunch more and from a UX standpoint I just can't agree. It's problematic both practically and theoretically/philosophically:
Not to imply that the selection model, or the issues with it, are unique to rmlint, BTW. One place you'll find the same UX is in Nautilus' List View mode. If the "Allow folders to be expanded" preference is enabled, it turns Nautilus' list view into an expandable tree much like Shredder's. And it's the source of some pretty buggy behavior in Nautilus, too.
I apologize for belaboring what feels like it should be a minor point of GUI implementation. But I know that every aspect of rmlint's filesystem operation was carefully considered, to the point that it was the topic of an entire treatise on the subject. (A document to which I still refer anyone who casually suggests adding "simple" file-syncing functionality to some network service or whatever. Because assuming you don't want to end up accidentally losing your users' data, there is no such thing as "simple" file syncing.) |
Long story short: |
I have a two-directory compare that I ran with the following settings:
Then I tagged one of the two trees (my master copy), selected both trees (master copy and working copy), and let the comparison run. It returned a large set of duplicates, as expected, and automatically marked everything in the working copy to delete, that's also present and marked to preserve in the master copy. Again, as expected, and all good so far.
If I "Render script from All", no problems, I get a script that'll delete every dup in the working copy.
But I don't necessarily want to do that. I want to pick and choose some parts of the working copy to delete. And that's the part I can't seem to make work.
No matter how I make a partial selection, the resulting script comes up with the
AUTOGENERATED OUTPUT
section completely empty. I've tried:And so on. Nothing works, unless I select the ENTIRE results tree, and then I get the exact same script that "Render script from All" would produce. Even if I then deselect some files/directories, following a select-all, those files are still included in the generated script, same as if I'd just done "Render script from All". (This is presumably because the very root of the tree is still selected — but whenever that root node is not selected, I get empty scripts.)
At least with the comparison settings and directory selections I'm working with, it's not clear to me how "Render script from Selected" can be convinced to do anything at all.
The text was updated successfully, but these errors were encountered: