Developer discussion for Creative Commons technology and tools

[Dan Mills <dan AT creativecommons.org>]

Not necessarily, see the two examples I gave Greg. And there are others which we aren't interested in today, but we should know are possible in the future. For example:

http://www.gnu.org/licenses/gpl.html

That page has a ton of stuff that is not part of the GPL itself.

Again - that's not what we want to do at this time, but it's conceivable that at some point we might.

Dan

On Thursday, April 18, 2013 at 11:04 AM, Maarten Zeinstra wrote:

Well actually ideally it would exactly the same document, but with different css and no js, right?

Cheers,

Maarten

-- 

Kennisland 

| www.kennisland.nl | t +31205756720 | m +31643053919 | @mzeinstra

On Apr 18, 2013, at 20:01 , Dan Mills <dan AT creativecommons.org> wrote:

On Thursday, April 18, 2013 at 10:52 AM, Nathan Yergler wrote:

On Thu, Apr 18, 2013 at 9:32 AM, Mike Linksvayer <ml AT gondwanaland.com> wrote:

On Wed, Apr 17, 2013 at 3:57 PM, Kat Walsh <kat AT creativecommons.org> wrote:

If we were to do this, the legal code would be maintained in a separate file

from the HTML, in a format that maintained all of the essential information.

For example, formatting such as bold or italic text that has legal

significance, section headings, etc., would all be considered essential and

part of the legal code itself. This legal code file would likely be

maintained using Markdown[1], or something similar to it.

The web page with the licenses would be generated from this legal code file,

by converting it to HTML and adding non-legal code formatting, text, and

navigational elements. However, since the legal code file would not have to

be touched, it would be impossible to accidentally make a change to the

legal code itself by changing other elements of the page.

I may have suggested something like this long ago, but I'd probably

stick to HTML as the canonical version now. That canonical HTML should

be as minimal as possible, just including enough structure and

annotation to make it possible for external CSS and _javascript_ to make

look pretty and dynamically add further annotation in a variety of

contexts, and for plain text to be generated without manual post

processing.

While you could continue to use _javascript_, etc for injecting that

sort of customization, I think the burden for creating and maintaining

that sort of code is greater than that for a script that takes a

template document and runs in the actual content.

I very much agree. Client-side JS absolutely has its place, and I have no problems with using it (heavily, if needed), but it's not some sort of escape hatch for modifying pages without modifying the page that is served up. That's just obfuscation, and it's harder to maintain.

Regardless of the markup format for the "immutable" document, I think

my primary concern is making it easy for a software agent to "follow

its nose" from the license URI to the immutable legalcode. (I

*thought* there was follow-your-nose markup from the deed to the

legalcode, but I don't see it now, so maybe I'm mis-remembering.)

Figuring out what the right predicate is shouldn't be super difficult,

and would fit in the existing ecosystem.

Were you thinking of a link? "The license on this page was generated from [link]" ?

 

For what it's worth, we added support for "stripped down" legalcode in

2010 (I think). For example,

http://creativecommons.org/licenses/by-sa/3.0/legalcode-plain. That

file is generated from the static HTML, and having

markdown/restructured text/something less expressive would have made

life a little easier.

Hah.

Yeah, so that plain format could be close to being acceptable as a *source* if we really want to use HTML (modulo the stylesheet and JS tags).

Dan

_______________________________________________
cc-devel mailing list
cc-devel AT lists.ibiblio.org
http://lists.ibiblio.org/mailman/listinfo/cc-devel