Re: [libreplanet-discuss] Building a free/libre platform for reviews

[Dave Rolek]

On 08/05/2016 06:23 PM, Erik Moeller wrote:
> I wanted to give you an update on this project, in case you feel like getting
> involved now there's a bit more to look at.
> <snip>
> * You can create new reviews. For now, you identify things to review by
>   providing a URL. As you can see on the front page, the advantage of this is
>   that we can already review pretty much anything -- such as software 
> packages,
>   movies, or books. The disadvantage is that we don't have easy
>   discoverability / search yet. But it's a way to get started.
>
>   Over time we want to interface with open repositories like Wikidata,
>   OpenStreetMap, and so on.
Hi Erik, it's good to hear from you and of the progress being made on
this project.  Keep it up!

As for the quoted part above, using a URL as an identifier might cause
difficulty in grouping reviews on the same item.  Here are just a few
examples (but the problem space is potentially "infinite"):

Wikipedia article:
https://en.wikipedia.org/wiki/Free_software
https://en.wikipedia.org/wiki/Free_Software

Wikipedia articles, significant difference:
https://en.wikipedia.org/wiki/FLOSS
https://en.wikipedia.org/wiki/Floss
(sorry, I know it's on the "discouraged terms" list, but it was the best
example I could think of in short time)

Book w/ ISBN:
https://en.wikipedia.org/wiki/Special:BookSources/978-0-9831592-0-9
https://www.amazon.com/Free-Software-Society-2nd/dp/0983159203/ref=sr_1_1?s=books&ie=UTF8&qid=1470453842&sr=1-1
https://www.amazon.com/Free-Software-Society-2nd/dp/0983159203/ref=sr_1_1
https://www.amazon.com/Free-Software-Society-2nd/dp/0983159203/
https://www.amazon.com/Free-Software-Society-2nd/dp/0983159203
https://www.amazon.com/dp/0983159203/
https://www.amazon.com/dp/0983159203

Software `mitmproxy`:
https://mitmproxy.org/
http://docs.mitmproxy.org/en/latest/introduction.html
http://docs.mitmproxy.org/en/latest/mitmproxy.html
https://github.com/mitmproxy/mitmproxy/
https://github.com/mitmproxy/mitmproxy

etc.
There are a few patterns above that certainly fall under convention but
are not strictly standard, e.g. the trailing slash could be treated as
significant.

I'm honestly not sure how one could tackle this*, but I would be curious
to hear your thoughts on this and whether the design may (at some point)
move away / limit the usage of URLs to identify the "target" of the review.

*I hope there is some project out there that attempts to document
"equivalent URL" patterns, but I'm not aware of any.  That would at
least help some of the above, but it wouldn't help when content is
hosted in multiple places or e.g. in the book w/ ISBN case.

Once again, it's great to hear of progress in this project!

dmr