- From: James M Snell <jasnell@gmail.com>
- Date: Fri, 23 Oct 2015 12:24:51 -0700
- To: ☮ elf Pavlik ☮ <perpetual-tripper@wwelves.org>, Michael Vogel <heluecht@pirati.ca>
- Cc: public-socialweb@w3.org
- Message-ID: <CABP7RbcRKtaJirxuTd2qHe57dtwjQh+GsONn36FZrrxwf4_LPw@mail.gmail.com>
Michael, thank you for the specific comments. Keep in mind that much of the complexity comes from JSON-LD itself and not AS2.0. In fact, the AS2 spec takes strides to *reduce* the complexity of JSON-LD. On the specific topic of displayName vs. displayNameMap, I feel your pain. AS actually recommends *against* using most of the options available in JSON-LD for expressing language tagged values. It can, and likely should, be simplified further. There's only so much we can do tho. W3C guidelines require a certain level of i18n capability that currently does not exist in JSON. So to achieve that capability in AS2, we either need to borrow from JSON-LD or invent it ourselves. If we tried the invention route, we'd likely end up with something almost identical to JSON-LD. So let's just adopt a *subset* of what JSON-LD provides. On the difference between displayName and displayNameMap... These were specified this way because of feedback from this WG. Essentially, it works best for JSON-LD processing but that's about it. There are a number of simplifications that can be made tho and I will write up a proposal to do so. As far as the implementation goes, having already gone through the process several times, it's not as bad as one might fear. Annoying, yes, but the details are easy to code around. But you're right, this particular part of the spec can be improved. I appreciate the specific implementation focused feedback, it is quite helpful and welcome. - James On Oct 23, 2015 11:14 AM, "elf Pavlik" <perpetual-tripper@wwelves.org> wrote: > forwarding in case some of use don't subscribe to > public-socialweb-comments@w3.org > > i recommend everyone subscribing to public comments! > > BTW Michael contributes (possibly maintains) Friendica :) > > > -------- Forwarded Message -------- > Subject: Unnecessarily complicate definition? > Resent-Date: Fri, 23 Oct 2015 17:09:16 +0000 > Resent-From: public-socialweb-comments@w3.org > Date: Fri, 23 Oct 2015 19:08:49 +0200 > From: Michael Vogel <heluecht@pirati.ca> > To: public-socialweb-comments@w3.org > > Hi! > > I just had a look at http://www.w3.org/TR/activitystreams-core/ > > Normally I really like JSON. But when I see the examples, then I think > that someone tried to make the worst out of the both worlds "JSON" and > "XML". > > With a structure like this you may have to implement some kind of > "JQuery for JSON" to fetch the relevant data. > > Data with only a single language could be written this way: > > { > "@context": "http://www.w3.org/ns/activitystreams", > "@type": "Object", > "displayNameMap": { > "en": "This is the title", > } > } > > Or that way: > > { > "@context": "http://www.w3.org/ns/activitystreams", > "@type": "Object", > "displayName": "This is the title" > } > > Or even that way: > > { > "@context": "http://www.w3.org/ns/activitystreams", > "@type": "Object", > "displayName": { > "@value": "This is the title", > "@language": "en" > } > } > > And maybe there is some other way possible as well. > > That really means that you need a parser that queries all these > variants. That really makes the things complicate. > > I really would suggest a single definition. Flexibility maybe nice when > defining this stuff. But it is a nightmare in the implementation. > > Michael > > > > >
Received on Friday, 23 October 2015 19:25:26 UTC