W3C home > Mailing lists > Public > public-sysapps@w3.org > March 2015

Re: Informal CfC to close the SysApps WG

From: Anders Rundgren <anders.rundgren.net@gmail.com>
Date: Thu, 26 Mar 2015 22:51:39 +0100
Message-ID: <55147F6B.8070705@gmail.com>
To: Wayne Carr <wayne.carr@linux.intel.com>, "public-sysapps@w3.org" <public-sysapps@w3.org>
On 2015-03-26 19:42, Wayne Carr wrote:

As a non-member I have nothing to add regarding organizational issues.

However, regarding the "replacement", I'm less convinced that permissions is a panacea, particularly not for subsystems having multiple functions like secure hardware.  For such systems I strongly believe that packaged high-level services would be simpler, more secure and not impose difficult permission prompts on users.  What's missing is a standardized way of connecting such services to the Open Web:
https://lists.w3.org/Archives/Public/public-webappsec/2015Mar/0155.html

Such a scheme would also speed-up innovation; starting a multi-year standardization effort for every kind of hardware you can find is unrealistic.  There are also a bunch of one-of-a-kind solutions needing a better way connecting the web to the platform:
http://tomforb.es/dell-system-detect-rce-vulnerability

Anders R
webpki.org/papers/web2native-bridge.pdf


> I think it's likely this WG will close. We're going to quit the WG, and we're handling what we need with the CfC that ends tomorrow, but we'll do a couple of others that may be useful to the WG participants so their opinions are known.  (So, just to be clear, we support the two CfC's I'll start -- but it also doesn't matter to us what the result is -- this is just to help the WG participants have their opinions known by W3C staff).
>
> At the bottom, there is a list of some reasons people may want the SysApps WG to close.  Those are not part of the proposal.  They're there just as background. Responses to the informal CfC could list reasons to keep the WG open.  I actually can't think of any other than hoping something will change and the work will get done, but that can happen by creating a new WG which may be easier than making the case now for approval for a new charter.
>
> Call for Consensus Proposal:
> The participants in the SysApps Working Group believe the SysApps Working Group should be closed.
>
> Please respond to this list by end of day (anywhere) Thursday, 2 April 2014.  Silence will be considered agreement.  As always, responding to the poll is preferred.  Since we don't have a Chair, there won't be an attempt to assess consensus.  The responses will stand on their own for W3C management to consider. Also, if you disagree with taking this poll, that would be a useful response as well.  I'll do a simple count, yes or no, summary after its over.
>
>
> ---- Background - not part of the proposal and can be skipped -
>
> Some reasons for closing the SysApps WG
>
> 1. WG Charter expired 1 October 2014.  If a WG Charter is not extended (simply an email from Director to AC extending it) or renewed through rechartering (Advisory Committee Review and Director consent), the WG should close when the Charter expires.  It is not good for the W3C as an organization to have work continue with no review from the Advisory Committee and Director on whether W3C resources should be expended on it.
> 2. No request was made to extend the Charter or to re-charter in 6 months by the end of this CfC.
> 3. WG has no Chairs.  Both no longer in WG.  One no longer with employer while in WG.
> 4. Work on specifications is not underway.
> 5. The key execution model and security model envisioned by the Charter were not able to gain consensus.  A better approach seems to be to explore Service Workers and new Trust and Permission extensions to the usual Web security model. This WG was explicitly for standalone apps outside the usual Web Security model (otherwise the DAP WG would have done it), so the likely direction this will take is outside the intended scope of this WG.
> 5. There aren't multiple implementations for the specs so no prospect of reaching REC.  Two implementers have either already quit (Google) or are about to (Intel).
> 6. Specs that can't complete the requirements for REC or where there still isn't consensus on how to proceed are better done in Community Groups (CG).  CGs have a patent licensing model based on contributions that does not depend on completing specs with multiple implementations. WGs have no patent licensing if the spec never completes.  CGs also has a permissive copyright license that makes it easier for contributors to continue work outside the group if the effort fails (and risky, experimental specs often fail)
> 7. If it becomes clearer at some point in the future how to do what SysApps was created to accomplish, a new WG can be created, but there is no reason to have a WG stay in a "zombie" state waiting for that.   Closing the WG doesn't mean having specs like the Charter planned for wouldn't have been good.  Closing the WG means this approach can't achieve its goals at this time (e.g. because two will not implement).
> 8. Key parts of the initial scope are underway in Community Groups aimed at the Web security model (Bluetooth and NFC). It seems better now to head in that direction.
>
>
>
Received on Thursday, 26 March 2015 21:52:10 UTC

This archive was generated by hypermail 2.3.1 : Thursday, 26 March 2015 21:52:10 UTC