Hi Magnolians!
A little discussion I would like to start before Christmas: the way file-upload
works in Magnolia (Images, Documents).
We're looking at replacing/modifying some of the logic handling file-uploads.
Shortcomings of magnolia's current implementation include:
· No separation between images and documents
· No way to force "the correct choice" - user can pick a PDF when they
should be picking an image, they can pick a folder instead of a document, etc...
· No way to make an internal Link (or document Link) field required
("required" setting for controls does not work on uuidLink)
· No way to force file-types on upload - user can upload anything, from
.exe to .psd, no filtering of allowed file-types is possible
· No preview when picking images
· No handling of embedded image metadata (EXIF, XMP)
· No difference between "New File" and "New Image" --> all document
metadata is the same
· Useability: if you make a mistake in the "New File" dialog, the field
with the file to upload is cleared, forcing you to choose the file again
We're addressing these issues with some custom modules, which:
· Provide specific Trees (based on DMSAdminTree) for Images and
Documents
o Trees are based on DMS content (use the dms repository) but show only
images or documents, respectively
· Provide different Upload dialogs for our custom trees, allowing
different metadata on images and documents
· Provide specific uuidLinkDocument and uuidLinkImage controls forcing
the user to pick documents of a specific type
· Provide an improved repository browser, with image preview, image
upload, and using our trees
This is mostly working well.
However, to handle issues related to file-upload, I am now running into some
problems:
· How to force upload of specific file types?
· How to extract metadata from images at the right moment?
The Problem is that the current implementation for the "New Document" dialog
does everything in one step: the user enters the metadata and picks the image,
and then uploads everything at the same time.
I was thinking of a 2-step solution:
1. first the user chooses and uploads the document
· the system can check the document type against configured allowed
types
· the system can extract the metadata from the document to provide
default values for the next step
2. in a second step, the user is presented with the metadata form and can
enter the values for metadata
· any metadata found in the image is prefilled in the dialog fields
What do people think? Is this the right way to go about it?
Can you suggest any easier way to implement this?
Thanks for your attention,
Richard Unger
Vienna, Austria
PS: For those interested, a little background to motivate why I am asking these
questions:
We're implementing a Magnolia-EE based System for a large government sector
customer - the final system will run about 50 Websites, maintained by 150 or so
editors in different departments. Very few of our editors have strong computer
skills.
For this reason ease of use for the editors is very important to us - things
that might seem trivial to us programmers or to an experienced web-content
person can be serious problems for less advanced users.
Magnolia's GUI is very intuitive in general, we have already received a lot of
positive feedback from our testing editors.
However, one area that is consistently causing problems is the image and
document handling.
----------------------------------------------------------------
For list details see
http://www.magnolia-cms.com/home/community/mailing-lists.html
To unsubscribe, E-mail to: <[email protected]>
----------------------------------------------------------------
