Re: bettering open source involvement

[Suzanne Woolf <[e-mail]>]

Hi.

As an aside, DNS is one of the areas where there's actually a pretty active give-and-take between standards and open source development-- some of the leading implementations of the protocol are open source and have been for years, and there's been enormous benefit in an open "code to current spec; test multiple implementations; debug; revise code; revise spec" cycle for DNSSEC some years ago or some of the DPRIVE work more recently. 

One of the reasons, I think, is that interoperability among completely heterogeneous code bases is not optional. No one gets to say "my open source is all the spec you need, why would you want your own code base?" any more than they get to say "we have a spec, you should just follow it" or "you should just buy my proprietary gear to get this functionality".  There are a lot of implementations, new ones appear every day for various reasons, and while we're drifting towards the same lowest-common-denominator interoperability that everyone else gets in the widely-middleboxed internet, that minimum remains fairly high.

I admit I'm not sure why that is; I'm aware of a number of historical factors that seem relevant, but I suspect there are other important factors too. Perhaps it's simply the only way to maintain backwards compatibility.


Suzanne


On Jul 28, 2016, at 1:34 PM, Melinda Shore <[e-mail]> wrote:

> On 7/28/16 2:51 AM, Riccardo Bernardini wrote:
>> I have direct experience with FOSDEM (https://fosdem.org/2017/) that is
>> held every year end of January/beginning of February in Bruxelles.
>> Personally, I like it, I find it very active (~5000 participants/year)
>> and refreshing (after a year buzz-wording).  Joking I say that it looks
>> almost like IETF's incubator... :-) (the density of geeks per square
>> meter is similar :-)
> 
> I think that if we wait for people to come to us, we'll be
> waiting for people to come to us.  I spoke a few months ago
> at PyCon US, and while it was specifically about getdns (an
> advanced DNS library) it was also an opportunity to talk about
> DNS privacy and security issues, and that mitigations are
> specified and available, along with ongoing work in the IETF.
> It was an extremely positive experience, highly recommended.
> Also, non-IETF hackathons are a good thing.
> 
> Melinda
>