I have noticed something and it might become problematic in the future. When I create a new item in Omeka, I enter all the Dublin Core fields, no problem. I then upload the PDF file that corresponds to that item. But when I go back to view it and click on the file thumbnail, the file has another set of blank Dublin Core fields. The item and the file itself seem to each have their set of Dublin Core fields.
I am using Omeka for a research project involving researchers abroad, and I am thinking that this issue might become a big problem if people download the files and the metadata entered under item does not remain associated to its correspondant file (since when you download a PDF from Omeka, the PDF name also changes to a series of numbers and letters.)
Does this mean that I have to enter the metadata both for items and then for the file associated to the item itself?
File metadata is optional. Some projects desire separate file metadata. File metadata is not available, by default, on the public website.
Since you are collaborating with a team, you may want to create a guide for everyone that discusses how to fill in certain fields, so that there is consistency across the site, and to advise all researchers which fields to ignore or fill.
Much agreed with Sheila that a guide is helpful -- and as it comes together it might call for some careful thinking on an abstract level about the relationship between an "Item" and its "Files". It's sometimes the case that they represent two different things, and so they might have different metadata, depending on what you need to represent.
Here's an example of one possible way the distinction could play out:
Say the Item in Omeka is the Mona Lisa. The creator there would be Leonardo daVinci.
Easy so far.
But now let's look at the files. Imagine I snuck a picture of the Mona Lisa on my cell phone when I visited the Louvre. daVinci didn't create the image _file_ -- I did, and so the creator info for the image file would be "Patrick Murray-John", while the creator of the _item_ remains "Leonardo daVinci"
In some cases and projects, maintaining that distinction is important. In others, it doesn't really matter. If the distinction doesn't matter, it's safe enough to disregard file metadata. If it does, then that's the sort of thing to detail in a guide.
Omeka makes those fine conceptual distinctions possible, but doesn't require them. If you don't want or need those distinctions, it's safe to provide guidance that just says not to worry about the file metadata.
Hope that helps,
Just to add something relevant to the discussion because I was wondering what to do with files connected to an item that I wanted to hide (scanned notes, copyright photos, etc).
Since it's not (currently?) possible to mark files as public or not, I used Patrick's suggestion regarding SimpleVocab a few days ago (http://omeka.org/forums/topic/simple-vocab-for-collections) to work with file metadata as well as collections. This, combined with what I learned in this post here (http://omeka.org/forums/topic/ability-to-set-files-to-public-or-private) allows me to easily filter specific files from display. I mark the files as 'Private' or 'Public' in the "Audience" element with SimpleVocab and hide Private files with the file_markup filter.
Perhaps this will be useful to other people, or something to consider down the road for the Omeka team. It would be useful to create distinct vocabularies for items, collections, and files, but so far problems with overlapping vocab have been minimal.
I would be very pleased if You could explain this more precisely. I'm working on an Omeka project and I've found that this would be very useful for my project. Because I am not familiar with coding I was wondering if you could tell me what is the right code you have to write and which files to change in order to get this working.
Thank You in advance.