These are chat archives for linkeddata/dokieli

18th
Jan 2016
Dmitri Zagidulin
@dmitrizagidulin
Jan 18 2016 14:26
@csarven - I just discovered that a friend of mine is doing a research thesis on something that very much sounds like Dokieli! :)
so I just pointed him to it.
Amy Guy
@rhiaro
Jan 18 2016 14:36
@dmitrizagidulin openannotation -> w3c Annotations WG; dokieli is using some of the AnnoWG specs
Dmitri Zagidulin
@dmitrizagidulin
Jan 18 2016 14:36
thanks! :)
Amy Guy
@rhiaro
Jan 18 2016 14:38
@dmitrizagidulin Who's your friend?
Any link to work?
Dmitri Zagidulin
@dmitrizagidulin
Jan 18 2016 14:39
guy named Michael Broughton, working at university of Koln in germany. I'll see if I can get a link, but I think he's just starting his thesis
something about trying to convince a part of the scholarly editor community to move away from Text Encoding Initiative flavored XML markup and more towards something more rdf/linked data like
"Scholarly editors are transcribing and marking up old texts using Text Encoding Initiative flavored XML. The two biggest problems they run into is that A) they keep having tag overlap because they want to say things about these texts that are from different value hierarchies (most notably, talking about the text encoded in the document and the realities of the text-bearing artifact (aka document) itself. And B) when one editor marks up a text, it is practically IMPOSSIBLE for another editor to use it."
^ etc.
Amy Guy
@rhiaro
Jan 18 2016 14:45
ohh, interesting. Transcribing is like an orthogonal problem I hadn't even thought about... but yeah, related I guess. Awesome if dokieli could cover that use case!
Dmitri Zagidulin
@dmitrizagidulin
Jan 18 2016 14:45
yeah, definitely gonna steer him towards it
Amy Guy
@rhiaro
Jan 18 2016 14:46
'editor' in the context of that paragraph is a human being, right?
Not a tool/system?
Dmitri Zagidulin
@dmitrizagidulin
Jan 18 2016 14:46
yeah, I think so
definitely the first usage of it, and probably that last one too
Amy Guy
@rhiaro
Jan 18 2016 14:46
This sounds like a humanities thing
Dmitri Zagidulin
@dmitrizagidulin
Jan 18 2016 14:46
it is, yeah
but he was looking at the Open Annotation spec, which is why I asked here
Amy Guy
@rhiaro
Jan 18 2016 14:47
great
Sarven Capadisli
@csarven
Jan 18 2016 15:29
@dmitrizagidulin @rhiaro dokieli sort of covers some TEI-like UCs in which you can embed arbitrary descriptions (in Turtle, JSON-LD, TriG) into the document (see the menu -> Embed Data). So for instance, if you want to say something more about a heading (which wouldn't normally go into the body text of the article), you can do it from there. The UI needs to improve on that front.
Dmitri Zagidulin
@dmitrizagidulin
Jan 18 2016 15:29
nice, ok.
how is Pundit (http://labs.europeana.eu/apps/pundit) related?
Sarven Capadisli
@csarven
Jan 18 2016 15:32
I haven't reviewed Pundit yet.. but it looks along the lines of hypothes.is
As for annotations, we're quite current with the Web Annotation data model. Open Annotation was precursor to the current W3C Web Annotation WG.
Dmitri Zagidulin
@dmitrizagidulin
Jan 18 2016 15:37
makes sense.
Sarven Capadisli
@csarven
Jan 18 2016 15:41
In case you haven't tried it yet, if you sign-in and leave an annotation (the last button in the toolbar after highlighting), it will write the annotation to your storage using the text quote selector approach i.e., it captures the exact selected text, 32 chars prefix, and 32 chars suffix. It stores that information in the annotation (at your storage), and then two things happen: 1) the current document fetches that annotation and inserts it into the document (visible on the side of the document), 2) leaves a notification for it on the document's inbox so that when that document is loaded again, it will go through step 1.
If you experience issues, please file them :)
Dmitri Zagidulin
@dmitrizagidulin
Jan 18 2016 16:08
thanks!
Sarven Capadisli
@csarven
Jan 18 2016 16:10
Oh.. 2 is not completely finished. It is leaving the notification at the inbox, however it is not loading the notifications from the inbox. Should be done by the end of day :P
Tim Berners-Lee
@timbl
Jan 18 2016 16:27
@csarven? do you know when when I saved a dokeili document to timbl.com I couldn’t edit it?>
Sarven Capadisli
@csarven
Jan 18 2016 16:45
@timbl I replied to your email.. And that's fixed now. Default document mode is 'read', so changed it to 'edit' when the user does a 'save as' (the URL for it includes ?edit=true).. which goes away on load (rewrites the URL history).
We are already doing ?edit=true for 'new' documents, and doing it for save as makes sense.
I don't know at the moment if there will be document modes other than 'edit' (besides 'read') but for now I prefer not to persist ?edit=true in the URL. It will minimize the number of different URLs being bookmarked/copy-pasted...
Sarven Capadisli
@csarven
Jan 18 2016 22:05
@sandhawke re: issue #100 , Thanks! That's a great set of features to have documented.