Registry in RSS

[Douglas Burke]

On 1/13/10 3:09 PM, Tom McGlynn wrote:
> One thing that I've thought would potentially be very powerful would 
> be for observatories to tweet events of the form:
>   "It's now YYYY-MM-DD HH:MM:SS and I'm taking an image at RA,Dec with 
> CameraX"
> One could do this with regular VOEvents (and I've mentioned it in that 
> context), but my sense is that it's much culturally attuned to how 
> tweeting is done generally.
>
I've been meaning to set something like this up for Chandra (following 
on from my experiments with twittering the position of Chandra over 
certain landmarks), but haven't yet found the tuits (*). For the general 
public I'd like to say "Chandra is going to observe the galaxy cluster 
Abell 2057 with the HRC for the next ten hours.", which suggests 
possibly two streams (with and without detailed coordinates).  If anyone 
has any suggestions let me know.

One issue with archival style analysis is that (AFAIK) twitter doesn't 
guarantee that past events will be searchable/retrievable (due to the 
volume of highly-important tweets they are collecting :-) so I'm not 
sure how useful it would be in this case.

Doug

(* hopefully real-soon-now ;-)
> If there were a significant density of tweets of this form, 
> astronomers could easily choose to make simultaneous observations  and 
> develop spontaneous collaborations.  They could mine also mine 
> archived tweets to find if there was data that might be useful a given 
> project.
>
>     Tom McGlynn
>
> Rob Seaman wrote:
>> Not sure why this would be heresy - VOEvent is a pretty agnostic
>> standard.  Is the current specification sufficient to the task?  By
>> all means go for it.  Need new features?  Tell us what they are.  You
>> would want to use a Stream separate from the various celestial feeds.
>>
>> Rob
>> ---
>>
>> On Jan 13, 2010, at 12:43 PM, Mike Fitzpatrick wrote:
>>
>>> Not quite what I meant:  I know there are numerous feeds for actual
>>> "VOEvents", what I was proposing was a feed for "VO Events" like
>>> the availability of a new resource, overlaying the voevent
>>> infrastructure,
>>> e.g. not that the Catalina Sky Survey found an object, but that the
>>> CSS is now a broker for "Bob's-really-cool-GRB-followup-on-his-
>>> GalileoScope"
>>> program.
>>>
>>> The heretical part is that I see 'events' as a type of message, and
>>> little
>>> difference between a GRB "stream" subscription, a "KBO" one, and
>>> a "general msg" one.  Is the idea "there is something new on the sky"
>>> really that much different than "there is something new in the
>>> Registry"?
>>> Note I'm not quite yet suggesting IVOA address these cross-WG issues,
>>> neither do I see a need for an emerging discussion on a new protocol
>>> for Tweet/Rss that shares many similarities (at least wrt the 90/10
>>> rule)
>>> with existing standards.
>>>
>>> -Mike
>>>
>>>
>>> On Wed, Jan 13, 2010 at 12:05 PM, Roy Williams
>>> <roy at cacr.caltech.edu>  wrote:
>>>>
>>>>
>>>> Mike Fitzpatrick wrote:
>>>>>
>>>>> And for a bit of heresy:  Should/can a new
>>>>> Tweet/Rss be used in the context of VOEvent (where 'event' is
>>>>> generalized
>>>>> to something beyond the sky but is a type of 'VO' message)?
>>>>>
>>>>
>>>> This is already happening. The Skyalert twitter feed(*) carries
>>>> events from
>>>> several ongoing projects, and also carries the interpretation of
>>>> the event
>>>> ("it was a supernova").
>>>>
>>>> (*) http://twitter.com/skyalert
>>>>
>>>> -- 
>>>>
>>>> California Institute of Technology
>>>> 626 395 3670
>>>>
>>>>
>>