[Commotion-dev] OLSRd plugin for service discovery

Paul Gardner-Stephen paul at servalproject.org
Tue Feb 12 10:52:23 UTC 2013


Hello,

On Tue, Feb 12, 2013 at 11:39 AM, Dan Staples
<danstaples at opentechinstitute.org> wrote:
> Geobounding is a great idea. I assume that requires some sort of
> location tagging, perhaps via GPS?

Yes, so at present we have the capacity to advertise the geobounding
boxes of bundles, which would naturally require a GPS fix to include
in a bundle's manifest.  Similarly nodes would need a recent GPS fix
to decide whether to accept a geofenced bundle.  We do not currently
add the geotags, nor do we check them, but the facility is there,
albiet untested.

> And can you explain a little more what you mean by "allowing the natural
> localised diffusion of bundles to allow localisation of information"?

Data will tend to distribute from where it is released on the mesh.
If there are more bundles than can be synchronised across the entire
mesh, then it is expected (but not proven) that data will be more
likely to be available nearer where it was produced.

Paul.

> On 02/11/2013 07:00 PM, Paul Gardner-Stephen wrote:
>> Hello,
>>
>> This whole idea of locality of information, and value being related to
>> proximity to some area is something that we very much agree with.
>>
>> I think that there are a variety of complementary ways that it can be
>> supported by the network.  This TTL concept is a great one. In Serval
>> Rhizome, we also have implemented a geobounding box, as well as just
>> allowing the natural localised diffusion of bundles to allow
>> localisation of information.
>>
>> Paul.
>>
>> On Tue, Feb 12, 2013 at 8:45 AM, Christian Huldt <christian at solvare.se> wrote:
>>> Well, the main point is that some local information is more interesting
>>> in the local neighbourhood.
>>>
>>> My first (in some kind of ordering, probably chronological) idea was an
>>> old museum with thick stone walls (namely the Swedish national museum
>>> with paintings and sculptures), and one access point per room with a
>>> small webserver - advertised by avahi with att TTL of zero hops - with
>>> information of the pieces in that particular room.
>>>
>>> And then - out of annoyance with the seemingly weak community spirit
>>> around here - I was thinking of a mesh along some street with lots of
>>> tourists payed for by local shop owners hosting their own access point
>>> with some presentation/hours/offers advertised by some rather low TTL
>>>
>>> I think there are quite a few areas where local information has a
>>> certain quality, that's why I tried to propose a "radius of interest"
>>> parameter to rss version <something> back in the day... (because I would
>>> love to zoom into Paris before going there and see local shows,
>>> happenings, whatever the week before I go there (went to Amsterdam in
>>> the 80's just to find that I missed a Burning Spear concert by one day
>>> just because I didn't see the poster in time)
>>>
>>>
>>> Seamus Tuohy skrev 2013-02-11 20:24:
>>>> I can't speak for Dan, but i think that exhibitions and museums are
>>>> exactly the kind of use cases that would interesting to see for mesh
>>>> service advertisement with TTL. Could you expand on your use-cases. I
>>>> would be interested in what you have in mind.
>>>>
>>>>
>>>> s2e
>>>>
>>>>
>>>> On 02/10/2013 02:20 PM, Christian Huldt wrote:
>>>>> Dan Staples skrev 2013-02-10 06:37:
>>>>>> I just finished an initial version of a plugin for OLSRd that propagates
>>>>>> mDNS/DNS-SD advertisements based on a service's TTL value.
>>>>> I can see a use-case for this at exhibitions and museums which probably
>>>>> is a bit out of line for the original purpose of this project, but could
>>>>> perhaps give some more users and hence testing?
>>>>>
>>>>>
>>>>> _______________________________________________
>>>>> Commotion-dev mailing list
>>>>> Commotion-dev at lists.chambana.net
>>>>> https://lists.chambana.net/mailman/listinfo/commotion-dev
>>>>>
>>>>>
>>>
>>> _______________________________________________
>>> Commotion-dev mailing list
>>> Commotion-dev at lists.chambana.net
>>> https://lists.chambana.net/mailman/listinfo/commotion-dev
>>>
>> _______________________________________________
>> Commotion-dev mailing list
>> Commotion-dev at lists.chambana.net
>> https://lists.chambana.net/mailman/listinfo/commotion-dev
>>
>
> --
> Dan Staples
>
> Open Technology Institute
> https://commotionwireless.net
>
> _______________________________________________
> Commotion-dev mailing list
> Commotion-dev at lists.chambana.net
> https://lists.chambana.net/mailman/listinfo/commotion-dev
>



More information about the Commotion-dev mailing list