Currently we need to manually create every collection in panels. I want collections; but find it cumbersome to create a new folder every time I add a new series of comics.
I used Chunky before switching over to panels and it had the feature to automatically create collections based on the filenames of the imported files. Currently Panels already somewhat has this function by suggesting the collection name upon import, but only there and it requires a manual action of creating the folder.
I think - if you decide to implement this request - there can be a few ways to implement it;
Next to the sort option, have a view option that allows collections to be created based on the filenames. Sort of like an overlay that does not touch your manually created collection (as other users will prefer those to manually manage their library) but provides and alternative view for those that don’t want to bother with organizing.
have a toggle in the settings menu to let panels auto create collection based on file names and disable manual collection creation). User can manually change filenames if they are sorted incorrectly to get them in the right collection (this is what Chunky does).
We’ve always avoided creating anything automatically because it’s improbable that it would work for everyone. Some people like to have complete control over their library, and others prefer convenience over customization. It’s pretty hard to balance, and that’s why we’ve always opted not to do anything on your behalf.
“have a view option that allows collections to be created based on the filenames”
We want to implement something similar to this. We’d love to have a feature that would suggest a name based on your file names. But as easy as it might sound, it’s not a trivial problem. For now, the cost/benefit ratio has kept us from doing it.
“have a toggle in the settings menu to let panels auto create collection based on file names and disable manual collection creation”
We would have to maintain two different implementations with very different approaches. For now, we prefer to keep it more uncomplicated and have only one way of organizing the library. (that is quite complex already)
Also, we are in the progress of building a file-system-based library. Where collections and comics are files and folders on your disk, also, you’ll be able to nest collections. These two factors will make the content organization different if we keep two different approaches.
Nevertheless, we’ll keep making Panels library more accessible and easier to organize.
Thanks for the explanation. I understand the complexity that this can bring and the need to keep the codebase manageable. I’ll just wait and see if the new file-system-based library will open up some possibilities for less manual library management.
It would be great it Panels would automatically create collections based on the series metadata in an archive. If the metadata is there you don’t need to worry about making assumptions, because the file is telling exactly how it should be grouped.
It would be nice if the “new collection from selection” feature suggested a collection name made up of the longest common substring of all the selected items. It could just have the collection name pre-filled in but changeable.
(Yes, I’m aware that Longest Common Substring isn’t really an efficient thing to calculate, but it shouldn’t be too crazy for this use case where the selection is probably fewer than 50 items).
It would be nice if the “new collection from selection” feature suggested a collection name made up of the longest common substring of all the selected items.
It’s been added in the latest beta. Maybe some of you already noticed it.
Thanks for the feedback @awh
About this. We don’t have support for metadata. Mainly because there is no guarantee the comics files have them, it will only solve the problem for a small group of users who have put effort into ensuring their files are correctly tagged.
We want to add support for metadata in the future, joining forces with the Anansi Project https://anansi-project.github.io/. (which is based on ComicRack’s ComicInfo.xml. Once we add support for it, we will explore options to use that information to create easier ways to access the content.
I don’t think we will ever get rid of the file-system-based library in favor of an auto-organized library based on file metadata. But I see a future where Panels can provide an alternate “view” of the library using the metadata. That way, users with completely tagged libraries will have an auto-organized view, but the comics will be “physically” organized the way they are on the file system.
If you drop all your comics at the library’s root, but they are all correctly organized, one view of the library will show all of them at the same level, but the other one will show them auto-organized. If that makes sense…
We are still far from there. I hope that’s clear. But we will get there at some point.
It works really well! Also, thanks for making it smarter than my suggestion, because thinking about it, the “longest common substring” here also contains the space, the # and the 0, which wouldn’t be right, but your algorithm correctly didn’t try to suggest it:
thx. Yeah, I wrote a “longest common subwords” algorithm, which discards incomplete words.
I did a lot of testing, and I’m quite happy with how it turned out.
Though I’m happily surprised it works fine with kanjis That I didn’t consider. I’m glad it works fine.
Sure. File system level storage is, I think, irrelevant to most users, unless the app makes it matter because that is how they have to interact with it or something. The suggestion for grouping is at the visual layer and how the files are stored shouldn’t matter.
With regards to metadata, I realise it may not matter to many, but I feel like those who spend time to tag their comic archives are the kind of people for whom a nice reading app appeals, and also the kind of folks who find an app that doesn’t support it to be annoying/lacking. It feels wasteful to my time if when I import files I have to organise them yet again. Computers are supports to make things easier for us. The prime candidate for this, in my opinion, is that Panels doesn’t auto-detect reading direction if it is in the metadata. That is job one for me. I spent the time to make sure the info is there for right-to-left archives. Having to manually adjust it in the app is just the app not bothering to look at the information available to it so it can make my reading experience just a little smoother.
I appreciate the quality of the app and the time you put into it, and have even been a premium subscriber for a good while, but I don’t actually use the app because it’s not a convenient reading experience, for me.
Metadata is part of that, but the fact I have to do so much manually, or that has already been done, means that just reading via a Komga webpage is a much smoother experience. Even with the recent improvements to sync reading progress etc (thank you for that), reading is still a hassle because:
the up next/in progress view is not attractive (nor front and center)
the reading order isn’t auto-detected
and I can’t continue with the next chapter when I finish the current one.
It’s like the app puts up a road block I have to actively work around at every point I need to make a decision.
thx. Yeah, I wrote a “longest common subwords” algorithm, which discards incomplete words.
That I didn’t consider. I’m glad it works fine.