Upgrade StorageHow to prevent automatic attachment of "comments"
This is an old discussion that has not been active in a long time. Before commenting here, you should strongly consider starting a new discussion instead. If you think the content of this discussion is still relevant, you can link to it from your new discussion.
This isn't a "feature". Translators save all sorts of data, and can do so to any field. Not everyone finds all of it useful. I don't think it even makes sense to talk about it as a singular thing. We just might make a policy decision that web translators shouldn't save things like arXiv comments to notes, and should instead save them to Extra. But there might be some cases where the data is closer to what people think of as notes being for (e.g., a paragraph of text).
I suspect a technical issue here might be that some translators (like WorldCat) use RIS or BibTeX, which have generic mechanisms for creating notes. We'd have to decide if it makes sense to override those to Extra or delete them for web translators when importing the same RIS or BibTeX directly would produce notes.
But beyond those general decisions, there's really no point in talking about this across translators. We should make individual decisions, in separate threads, about specific sites. I'd start by moving arXiv comments to Extra if we think significant numbers of people find them useful.
Hypothetically yes this could be useful in some scenarios. But that's not what it's being used for now.
Additionally, although this does vary by translator, it could also be fixed by addressing the most common translators: in this case, I've only seen Worldcat and arXiv mentioned here, so those translators could just be fixed to not import this data and/or import it differently. Then if rarely a less common translator imported a note, we could manually delete it.
- arXiv we added it because it was explicitly requested
- Worldcat we're just adding it because it's in the note field in RIS and we're not explicitly removing it. This may happen for other translators too, in some cases simply because of bad metadata, especially with RIS: e.g. I just saw a DOI in a note.
- Other library catalogs we're using various fields in the 500 range of MARC in notes which contain anything from ToCs to additional publication info such as funding etc.
- For Espacenet, we're adding various Patent classifications.
Those are the ones I know about. I personally have always preferred having these out of my way and _not_ in the Extra pane. I'd also say that Extra isn't great for some of the longer ones, but overall I don't have strong feelings about this -- given that I review & write a lot of translators, the current behavior does probably reflect those preferences, though.
I'm happy to change this for new/updated translators if there's clear guidance, but I won't spend time cleaning up old translators, leaving that too people with stronger opinions.
If preferences do vary for this, then that goes back to the original request to add a preference option to disable this. But previously it was stated that it wasn't clear how to label/explain that feature, because it's too hard to explain to users.
If you're importing many items, adamsmith, you probably would want to check the data in each one for accuracy, or at least glance over notes you're deleting, so it wouldn't be entirely a batch operation anyway.
The issue of how a particular translator should deal with additional data aside, my personal preference would be for a global zotero setting that blocks import to Notes (i.e., if the translator produces a note, zotero should just discard that data), similar to the setting that blocks the import of tags
But Extra is really only an option for the ones that we're going out of our way in translators to extract, like arXiv comments. I think we should change those, but the ones that come from note fields in import formats have to go to notes no matter what, so the only possibility there would be a preference to discard them. But I think we would only want that preference to apply to translation other than file imports (so, from the web, Add Item by Identifier, feeds, metadata updating…), which makes it pretty hard to describe coherently.