- From: Joshue O Connor <joconnor@w3.org>
- Date: Wed, 10 Jul 2019 11:09:23 +0100
- To: "White, Jason J" <jjwhite@ets.org>, "public-apa@w3.org" <public-apa@w3.org>
- Cc: RQTF <public-rqtf@w3.org>
Thanks Jason. Just some brief comments inline.. On 09/07/2019 16:51, White, Jason J wrote: > Thank you, Josh, for these informative notes. I look forward to further discussion within the Research Questions Task force and the APA Working Group more broadly. > > One question that comes to mind immediately is, in each case, how comprehensively the Web of Things protocols enable the use and operation of the device. That is a very good lens for our discovery phase questions but is even higher up in the stack that current thinking within the WoT group. As I understand it, much of WoT work is currently focussed on the network protocols etc, so a lower level. But a use case that comes to mind based on your question is, if user device needs to respond to node 'states' or device status in a network - such as in the navigation use case - then these nodes could drive or lead the user device in some way. The antithesis is that if there is an application that will handle this data, and just respond to what is coming from the network then thats fine, and arguable an application only problem - but in the future, as you mentioned, there may be cases where sensors do actually control a user device somehow. I hope this makes sense, and isn't too meta *grin. > Suppose a device has its own user interface that is not adequately accessible to the user. However, there is an accessible UI available for remote operation that depends on Web of Things protocols. Do the protocols sufficiently support the functionality of the device to enable the user to operate it remotely, or are there important (perhaps even essential) functions that can only be used via the "on device" UI, and which thus create barriers to access for the user due to missing features in the Web of Things implementation? Good question. What we need to do is to try to catch where we see potential gaps, even in the lower network layers that may bubble and become more complex abstractions that need to be mediated in a way that ultimately the user can understand (if it gets that far up the stack). My hunch is that at the very least if may be a case the AT needs to adapt to consume JSON-LD, which can extend a Thing Description to point to other suitable semantics, or languages that are supported by WoT architecture. > I would welcome comments on whether this statement of the problem makes good sense. > Appreciated, thanks. Josh -- Emerging Web Technology Specialist/A11y (WAI/W3C)
Received on Wednesday, 10 July 2019 10:08:11 UTC