W3C home > Mailing lists > Public > public-hydra@w3.org > January 2015

Re: remove hydra:Resource and hydra:Class

From: Gregg Kellogg <gregg@greggkellogg.net>
Date: Wed, 7 Jan 2015 19:33:37 -0800
Cc: public-hydra@w3.org
Message-Id: <7B35200C-F914-40A2-A0E5-D252F30FECA8@greggkellogg.net>
To: Markus Lanthaler <markus.lanthaler@gmx.net>
> On Jan 7, 2015, at 2:05 PM, Markus Lanthaler <markus.lanthaler@gmx.net> wrote:
> 
> Wednesday, January 07, 2015 5:38 PM, Dietrich Schulten wrote:
>> ---- John Walker schrieb ----
>>> More formal change control should be applied once it is released. 
>> 
>> +1
> 
> Yeah, at that stage we can't really make any breaking changes anymore.

Yes, but I think we're a long way away from that. IIRC, there are a number of "decided" issues which have been pending to get into the spec for some time (collections, LINK operations, for example).

>>> Early adopters/implementers understand that things are still subject
>>> to change and accept the consequences.
>> 
>> Status "testing" says it clear enough. Still I thought we should be as
>> easy as possible on the community. But if its consense that we can and
>> will turn everything upside down, so be it :)
> 
> There's a saying that walking on water and developing software from a specification is easy -- if they are frozen :-)
> 
> On a more serious note, I think it's unavoidable to make breaking changes at this stage. We still have the luxury to have everything that depends on Hydra under our control. It certainly creates more work for implementers to keep their tools/libraries up to date but breaking things now is better than having to do so later.
> 
> I'm still on the fence about removing hydra:Resource and hydra:Class. For me they are hypermedia controls as well.. but it might be that we can design them better. I have to think more about it... and I guess we have more important stuff to figure out at the moment -- no, I didn't forget about paged collections :-)

I'm +1 on removing hydra:Resource & hydra:Class. For my use, the only things I trust are dereferencable are in my own domain, and are marked up by making the properties subclass hydra:Link; this is adequate for me.

Gregg

> --
> Markus Lanthaler
> @markuslanthaler
> 
> 
Received on Thursday, 8 January 2015 03:34:07 UTC

This archive was generated by hypermail 2.3.1 : Thursday, 8 January 2015 03:34:14 UTC