+1 on the folder name humanly-readable

ok, hmm. Let's take an experience I add recently, I needed to export a sub-library of my Zotero library so that I could print it on another computer. To do this I looked for the .pdf in the directory "MyUserName\Documents\Zotero bibliographie\storage\..." but then it was really messy.

Check this image to understand what I'm talking about : http://funkybudha.net/wp-content/uploads/zotero_storage.jpg

In blue: I'm not inside a AppData\... directory, no, I'm inside a User\Documents\... . I' m Expecting to have human-readable document system here. You could see that as a usability request.

In yellow : You have the actual random integer directories. On the right, you see the name of the pdf [renamed with zotero] showing meta-data (author, year) as expected for this type of document.

Now, I'd need to see the same meta-data reflected on the directories to help me pinpoint rapidly the documents I need.

This is one.

(The number two is that I lacked the possibility to export only the pdfs out of my library. Yes, I did it, then searched for the pdfs inside the export, etc. But it was not an "elegant" method to do so.)

Believe I did half the answer.

The problem with smart search is that one can't separate pdfs of a sub-library from another. One get all of them at once and even in multiple copies... So no, adamsmith, a smart search is actually not workable for retrieving sub-library group of pdf as the storage directory structure does not reflect the zotero library structure.

no one denies that this would be nice in general - it's a question of priority and I remain unconvinced that this is an urgent/high priority issue.

If you want to export a part of your library, just export as RDF with files - which is what, I think, you actually ended up doing (if you want just the files from that, search for pdf in the folder will do) - while the downside of this is that you have one additional step, the upside is that it allows you to export using Zotero's flexible collection-tag structure rather than a fixed directory structure:
Note that, crucially, Zotero collections are (purposefully) not equivalent to folders on your harddisk - one item can be in several collections and thus a directory structure cannot reflect the collection structure in Zotero - simply because they're two different things. And forcing one into the other is not elegant, either.

And the whole point about saved searches is that you _can_ have human readable data in a folder in your documents directory - you just need to use a built in function of your operating system - thus saving Zotero devs the time to re-program something that's already possible.

Dan/Trevor - this would be a good topic for the new Zotero Tips/Tricks series on the blog.

Once again - the point is not that this wouldn't be _nice_ in Zotero, but let me give you a small list of things that need yet to be implemented in Zotero and see if you really believe this is more important:

- Duplicate detection
- implementation of new CSL (allows for original date of publication and much more)
- Better note management (allow shuffling of notes, hierarchical notes and more)
- Hierarchical data structure - allow for chapters to be grouped as sub-items to e.g. an edited volume
- Improve tag functionality - esp. introduce colored tags for better visibility
- Improve functionality for other word processors through improved RTF-scan or similar feature
- Improve the style repository for searchability
- improve online display - allow for selective publication of libraries, allow sorting of online libraries etc.
- improve data input and output from/to other software especially Endnote - e.g. it's not possible to export files/file links along with an Endnote/RIS export.

and that's just a small selection of things that aren't possible to do without a workaround.

I think you may need a pat on the back.. Let me give you one: Zotero is really a great tool !! Hope you the best ;)

just for clarification: I have no job with Zotero.
I come here to do something useful while I'm procrastinating ;-).

Also, I don't see what I'm doing here (or elsewhere on the forum) as "pushing back" in any sense.

I view this - and open source software more generally - as a community effort that benefits from people getting a sense of what's going on, how the software works, some of the underlying concepts, and what type of considerations drive decision making etc. In my mind that's creates a sense of "ownership" of software that encourages people to contribute in one of the multiple possible ways and help make Zotero better. For that reason I also agree very much with miggug that it'd be helpful - and worth some dev time - if we could get some insight into the plans of the dev team at least for the medium run.

I guess there are probably other possibilities for solutions to this, but I don't really understand why a name like "[author][year]" or "[author][year][hash]" wouldn't be just as simple as the current system.

It is not hash, but a key that is permanent for an item. One of the problems in this is that author and year might change during the life cycle of an item and might not be available when the item is created.

But if you want to play with the command line, you can probably solve your problem by modifying these scripts

http://forums.zotero.org/discussion/16148/create-a-humanreadable-folderstructure-from-zoterostorage/

https://github.com/mronkko/ZoteroCleanOrphans

(They are separate scripts, I did not put the first one on GitHub for some reason.)

Drag and drop only works out of zotero only works if I expand each refererence and highlight the actual pdf file to drag and drop.

Press the "+" key to expand all items and select the attachments you want or create a saved search (say, by Attachment File Type) and use Ctrl/Cmd-A to highlight just the search matches. Then use drag and drop.