On 9 November 2013 02:33:52, fabian fabian@unpopular.org.uk wrote:

Yes I tried this . . . once. I create the DjVU file and got it onto commons but then when I tried to get it on Wikisource, I discovered there were forbidden characters in the file name which I had not been warned about before. After a bit more mucking about I gave up. All a bit frustrating really.

I've never heard of forbidden characters before, beyond the usual for all computers. You never did reply to my question, nor take this to the help forum.

Assuming you mean "File:Tudorschoolboyl01vivegoog.djvu" (as another user suggested), I could not reproduce your error. However, while I was examining it, I took the time to clean it up, replace it with a better copy (I try to avoid Google scans, they are frequently extremely poor quality) and had it renamed, per Commons policy, to "File:Tudor School-boy Life (1908) by Juan Luis Vives.djvu". (NB: The translator counts for copyright purposes, so I changed the licence too.)

There is still a small problem, Proofread Page isn't automatically recognising it as a DjVu file, but that can be fixed manually.

If you still want to try, click the wikisource logo on the Commons page, create and save. (Manual fixes that would be useful: add the <pagelist /> tag to the Pages section; change Scans to "djvu"; and change Progress to "To be proofread". I have no idea why it isn't doing this automatically at the moment.)

On 9 November 2013 04:57:47, Fæ faewik@gmail.com wrote:

This is one of the areas that having a chapter employee/named wikisource expert help could sort out, doing this in parallel to the competition might be a smart approach. Any volunteer that finds this confusing and has a key document that would be of high value to the projects, could just email a link or a photocopied document (via freepost) to a chapter contact. Only good for a limited number of documents and not whole books, but a good area to offer help, or indeed a training event for those that need a push to learn how to DIY and have some projects in mind (nods to cy.ws).

Creating a good djvu file (or even a pdf) is a bit of an art as with standard free tools it can be hard or impossible to set embedded image resolution etc.

My preferred three-step method is:

1) Upload it to the Internet Archive. 2) Let them deal with it. 3) Download the finished file.

That only works for page scan files (ie. JPEGs); I don't know how you would deal with photocopies.

Someone could do that, perhaps starting with something like Dropbox for the volunteer's scans. I turned my method into a help page on Wikisource (Help:Internet Archive) if anyone is interested. I can't do it: I frequently reach my bandwidth limit as it is, and may have to actually downgrade soon.

Speaking of photocopies, many standard office Multifunction Devices (MFDs), the modern version of an office photocopier, can easily make PDF files. Commons/MediaWiki has a problem with more recent version of PDF, however, but that's a separate issue. (I think Archive will convert PDFs as well as scans, I've just never tried it.)

The actual book scanning is the hard bit. I built a temporary V-cradle out of old cardboard boxes to do it when I tried. Lego might be a better medium for the future. I also only have one digital camera, so I had to turn it around and go the other way for a second pass (a professional book scanner has two cameras, pointing at either page, so it only needs one pass). That's too much of a spit and bailing-wire approach to really teach anyone. (Destructive book scanning would be easier but libraries might object.)

- Adam