Signing events

[rdenny at dc3.com]

> On the other hand, if we – you and I, for the sake of argument -- put aside those objections and press ahead with using your system now, but the rest of the community is scared off by the objections, we're also not actually achieving much.

I know.  On the other hand, when the need arises, and there is a field-proven solution that's been operational for years...

> I think I already pulled the scalability card, actually… :-)

You ain't seen nuthin'  The real scalability card is "but what about LSST which will be sending a million events per microsecond? Whatever. My claim is that the dreamed-of rate of events from LSST will exceed the capacity of actually using them by orders of magnitude, and LSST will need to filter the hell out of the for quality.

> For what it's worth, my current intention is to experiment with my own implementation of your PGP-based system and see how well it works for me in practice (I realise you already report success with Dakota, but I am a great believer in learning-by-doing). I continue to be worried about the compromises it entails, though, and eagerly hope that an alternative will emerge which addresses them.

Excellent. Feel free to call me at +1 480 396 9700 or Skype 'dc3dreams'. I'll call you back if regular telephone. The only compromises I can see are hypothetical what if's that, to me, are opposed to the notion of message ("packet") integrity.

> I'm not well versed in the history here; sorry. I suspect I've used the term vTCP in the past, and apologise if it's not to your taste. If you can suggest an alternative – preferably one that (a) doesn't occupy many more characters, (b) is unambiguous (there are many IVOA notes!) and (c) doesn't imply canonicity (there are multiple VOEvent transport systems in use) – I for one will be happy to adopt that instead.

Sorry for the tangential issue. Alasdair Allan and I published the VOEvent Transport Protocol in an IVOA Note. It has things in it, and a formalism that is recorded, that go beyond the "vanilla TCP" (vTCP) that resulted from a paper napkin design by Allan and Robert White years before. It would be nice to refer to what is in that document as Transport. I see no need for an acronym, but if needed, then VTP or something. 

  -- Bob
-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://www.ivoa.net/pipermail/voevent/attachments/20120306/31bf17d1/attachment-0001.html>