I’m doing a lot of slide show work of late, so I was wondering if any of my readers knew of an in-image equivalent of the gfwa script I use on the blog? Basically, I’d love to know of (or have written?) a script that, given an image, a license, attribution strings, etc., would spit out the image plus an unobtrusive overlay containing the proper attribution information, so that I could put the image into the slide and preserve the requisite licensing information.
Something like adding the following in the lower righthand corner of this:
but less ugly (and more license-compliant; any CC-expert readers might want to weigh in in comments on the state of the art for citation in non-web-y formats like pictures/slide-decks.)
To be really kick-ass the script would probably need to take a font color and corner (upper left, upper-right, etc.) but I’ll settle for having that hard-coded for now.
Lessig handles this by including a list of attributions, links, etc., on the final slide, which may be appropriate for particularly image-dense presentations, but I’d like to experiment with this for now- it feels more appropriate anyway, especially if it can be done in a low impact, aesthetically pleasing way.
Thanks…
13 Comments
I don’t have one but I would absolutely love to have something like this. I save all the images I use from Flickr in a folder with a filename that matches the name I give credit to on the slide. Then every time I use it, I copy in the picture and then add a text field that says “Photo by “. Having more detailed information on the picture itself would be great - better for the author and easier for me to use.
This isn’t a solution, but a suggestion to find a solution. I’ve done similar things using the imagemagick libraries. It would take a bit of trial and error to make it robust enough, but you could definitely accomplish what you need using it… all in a command line script environment.
Stormy: thought you might like this. ;)
Mark: I was assuming either imagemagick or gimp scripts would be part of the solution.
The basic command could probably be something like:
convert ORIGINAL_IMAGE.png -font Helvetica -pointsize 14 -fill white -box ‘#00000080′ -gravity southeast -annotate +0+5 ‘Attribution String’ png:- | composite -gravity southeast -geometry +0+25 CC-LICENSE-IMAGE.png - ANNOTATED_IMAGE.png
That’ll take the original image, add your attribution in the bottom right and output a png image to stdout. Then we use composite to combine the license image with the newly created image to produce the annotated version.
Wrap that with something to add the correct license image, etc., and you should be set.
hrm, poking at this a bit more. how about http://fenris.org/cc-licenses.tgz
definitely something that could be customized more than it is. note that the attribution string should be quoted (obviously), but that it can include \n for a multi-line attribution
I don’t have it automated at all, but see how I do image attribution and notice in slides for the past year or so in recent decks at http://www.slideshare.net/mlinksva/slideshows (text at bottom of slide where image is used, including on most covers).
I’m very interested in having active attribution and license links in the reuse context if at all possible, so simply modifying a bitmap doesn’t really help.
In order for automation to work across a zillion different ways an image could be used and programs that could facilitate same, attribution and license metadata has to somehow travel with the image, then programs can use that as appropriate.
The obvious option is to opportunistically embed metadata in the image file (eg CC recommends using XMP because it is flexible and works across many formats). Complementary to this, and I think more interesting, would be for the OS (speaking conceptually, not sure at what level or what mechanism) to remember the source (ie URL) and source context (ie another URL) of all its files, from which metadata can be discovered (and cached, etc.).
You asked for code, I gave you bs. Apologies!
cec: awesome, i’ll try that out tomorrow.
Mike: I realize that metadata is ideal. (Do Flickr or Picasa set it correctly in images it delivers?) The reality is, of course, that these will go into ppts and thence to pdf, so metadata will be lost, and I’d like to do second-best in the meantime.
cec: That is very sweet. Now it just needs to integrate wget to take a flickr URL instead… might well look at that tonight.
thinking about it, the script should probably also grab the license image dynamically. maybe add some options for different license versions, optional output file name, etc. I may take a poke at that tonight
cec: might want to look at how gfwa does that: http://muellerware.org/projects/getfwa/getfwa.user.js
(Really, I assume that a real javascript whiz could do all this in there, but I am not that person, as my current attempts to hack tinyurl into that are showing.)
We’re tracking this as a wishlist idea now - http://code.creativecommons.org/issues/issue7.
Oh - so I wrote up a chunk of this in javascript a few weeks ago. It works pretty well. Tonight I wrapped that in a GreaseMonkey script: http://www.fenris.org/projects/annotate-flickr/
thoughts or suggestions?
You may have catch it :
http://www.flickrleech.net
Source : http://creativecommons.org/weblog/entry/9970
4 Trackbacks/Pingbacks
[...] but which might be appropriate for someone in the community who wants to contribute. Luis’ idea from earlier this week is the first. I hope we have a giant pile of ideas (and a corresponding [...]
Post a Comment