This one is call ‘hypercat’ and though it sounds like the result of feeding coffee to a feline it is instead a discovery and presentation format based on JSON
Pro’s
Uses modern tech
Uses the imp notion of using the internet as the abstraction layer rather than a physical or web-based “box”
A lot of companies are involved including BT, ARM, Intel, and IBM
Con’s
Google isn’t involved, so it’s a discovery protocol without the biggest discoverer
Apple isn’t involved, and with the IOT community hoping they’ll kickstart IOT as they did with tablets, thats a problem
Interesting
That IOT is not just in geek news now is in itself hopeful
The expert in the report is a “doc’ brown” so the protocol may only work at 88 miles an hour
I’ve looked at HyperCat quite a bit in the past couple of weeks, and even began working on a HyperCat class for agents (that would allow you to programmatically construct catalogs).
There are things I really like about HyperCat, and things that I think may doom it fail.
It’s based on HTTP, abstracts to the hardware as a service, and is (or at least, appears to be) extendable.
But it also reminds me a lot of REST - and I’m talking about true hypermedia REST, not “exposing resources through URLs and generally following GET/PUT/POST/DELETE conventions for CRUD operations” REST.
REST (in theory) allows any API to be crawled, and understood through nothing other than web requests. You make a request to the root of the API, and it tells you all of the resources it controls (and URLs to those resources). Make a request to one of the resources, and it will tell you sub-resources, and actions that can be taken on that resource.
That’s all really great in theory, but in practice there are less than a handful of true REST APIs that I know about because they’re unnecessary and somewhat difficult to build (even with good resources/tools for building them).
I have a lot of really similar feelings towards HyperCat… I think it’s really great in practice, and I hope it pushes people towards building service oriented hardware, instead of hardware with services tacked on… but I’m not hedging my bets on true HyperCat (or any protocol at this point in time) being the One Language To Rule Them All.
Also - I would love to hear other people’s thoughts on HyperCat (and other communication type things). My background is software - and the part of software I really enjoyed was crafting APIs… so this is the kind of thing I get really nerdy about
I am also excited by things like this as well, this is for at least three of reasons
I think as a programming model it makes sense to retrieve and expose data using standards, as well as not “reinventing the wheel” it means reusing the code becomes easier and there are web resources for people who have “been there” and “done that”
From a support perspective using standards makes debugging easier
it encourages connectivity and opens your market up to things you would never expect of example impr’s have plugged a disco light into their dev environment!
As a principIe I would always encourage people to write their code in a multi-tier way and to interlink those tiers with standards
(take it from someone who’s day job is to debug code)
However I can’t find a HyperCat service ready for us to try out, it looks like it is still a specification, and all implementations have been inside the group. We need to be able to try it if they want the world to adopt the specification. If someone has found a service with developer accounts, please let us know!
A lot of the links on the hypercat page are broken
The cool part of hypercat isn’t implementing hypercat on a single device - it’s a world where there are thousands/millions of devices exposing hypercat catalogs, and we can easily query / link between them
@DolfTraanberg - HyperCat is HTTP + JSON - which means the imp is absolutely ready for an implementation. You could fairly easily hard code some of the response formating, but it would be great to have a class where you could add items, item-metadata, submit it to external catalogs, etc!
@beardedinventor there is evrythng.com another Brit company and thier Active Digital Identity concept which is supposed to be discovered-able.
The guys behind it HyperCat are all X http://www.alertme.com/ and they say their main focus is helping customers build scaleable IoT solutions. HyperCat was a nice bit of UK Gov IoT funding so wonder it get the take up it needs to flourish.
evrythng looks really interesting - I remember reading about it (quite) a while ago, but thinking it was much more of an art project than anything else… giving a voice and a history to your objects.
It seems like the UK Gov’t funding is helping, there is a growing HyperCat movement in the UK (and more broadly, Europe).
If there was a central HyperCat repository you could add your imp devices to (and a class that made it simple to build out a HyperCat catalog in your agent), would people be interested in exposing some data, etc? I think the Electric Imp community is in a pretty interesting place to help drive this
I rather like it as it’s decentralized and not awfully rigid at this point (though rigidity would help gain some acceptance, Pilgrim appears to be purposefully letting it settle a bit and seeing how people use it).
I know it’s been used in anger to link some UK government-sponsored IoT projects together, so it’s proved useful in some areas already, but as beardedinventor says, a central repo - at least to point to other repos - would be nice to help get it adopted.
at least to point to other repos - would be nice to help get it adopted.
There’s the rub you want some independent body to manage the standard.
And then hope this doesn’t go the way of EDIFACT where commercial VAN providers just charge too much to access the service in this case accessing repos local, regional central etc.
