Re: [gaia] Comments on: draft-irtf-gaia-alternative-network-deployments

[Mitar <[e-mail]>]

Hi!

On Mon, Apr 11, 2016 at 7:04 PM, Arjuna Sathiaseelan
<[e-mail]> wrote:
> Isnt a CN like Guifi doing exactly this?

Sorry if I was unclear. I was not saying that guifi.net does not have
those properties, they do (and this is one more argument why they
should be included) and they are simply amazing and highly
inspirational, but that from small hints here and there, like examples
used, it is noticeable that things are coming from their particular
case. Which is somewhat understandable, because they are doing amazing
stuff, but it might be one of the reasons for some bias. Because if
one thing is true about community networks is that they are as diverse
as possible. Even one network itself is often without any homogeneous
form or direction. But once you have this size of the network, it is
maybe easy to start thinking only about numbers: how many users they
have, what is the reach, how much they then pay for their Internet,
and so on. guifi.net is definitely an interesting example because of
their size they created their own economy inside the network, people
working full time on growing the network, companies moving in to use
them as backbone, and so on. All this is beautiful and it is a story
to tell, but it is only one side of a coin. If you do not look for the
other side of the coin, you might miss it. And when you look at a
network like guifi.net is it easy to get amazed by the first side of
the coin already, to not even ask "but is there something more to
those community networks".

> Are you saying that a specific section is not written correctly or are you
> saying the whole document does not illustrate the points you have raised. My
> confusion comes from this.

I commented point to point. So some parts of the document seem to me
to be missing important additional information are one sided, some
parts of documents seem to me unnecessary for this particular
document, and some are really off, like the routing section one, that
one needs quite a bit of improvement (especially because it is the
easiest to validate, as it is highly technical). In most cases I was
trying to propose concrete corrections/improvements to address the
issues raised.

I think the document is a great draft. I think it is moving in the
right direction, but sadly, I do not see it finished. Too many aspects
are still missing.


Mitar

> On 10 April 2016 at 19:53, Mitar <[e-mail]> wrote:
>>
>> Hi!
>>
>> I am participant in the open wireless network in Slovenia, wlan
>> slovenija (https://wlan-si.net/) and I am writing to provide some
>> feedback on the draft. I am glad people are working on it. My list of
>> comments is in no particular order, mostly in order of reading. I am
>> sorry if this will be long, but there are many issues with the draft.
>>
>> Section 4.2. I think this is a very limited set of motivations. From
>> our network and from my experience talking to other community
>> networks, I would claim that many share also much more altruistic
>> motivations. For example, our whole network is build around principle
>> that some of us have abundance of Internet connectivity at home
>> (FTTH), so we can share part of that openly with everyone. More of us
>> will share it, more people will have access to Internet. Of course
>> there are other motivations why people participate, and everyone has a
>> different mix of those motivations, but I definitely think that the
>> list should be extended to include:
>>
>> - free sharing of Internet connectivity, altruism
>> - various forms of activism (network neutrality guarantees,
>> anti-censorship, decentralization to minimize control, showing that
>> alternative is possible)
>> - building a new type of commons
>> - not being just a consumer, but active participant, wanting to have a
>> say in operations
>> - provide local services to local people, tighten/reconnect the
>> community (eg. http://tidepools.co/)
>> - provide alternative service in case of natural disasters and other
>> extreme situations
>> - wanting to have a space for experimentation and teaching of others,
>> empowering others to take their Internet connectivity into their own
>> hands
>>
>> An example of another network describing itself with much more of what
>> I am writing than what is currently on the list:
>> https://sudoroom.org/wiki/Mesh
>>
>> All those are goals and motivations (and those are just few I
>> remembered without much thinking) for many community networks. I do
>> not think the list in section 4.2 can ever be comprehensive, but I was
>> really taken aback from its economy-centric bias at the moment. Most
>> community networks do not operate on that level. It is of course
>> present, but there are some other primary motivations why people are
>> doing it.
>>
>> Section 4.4, technologies employed:
>>
>> "Low-cost optical fiber systems are used to connect households in some
>> villages."
>>
>> Isn't this section about list of technologies? Why are those villages
>> mentioned? Optical systems can also be used elsewhere.
>>
>> Moreover, in wlan slovenija we developed free optics free-space (no
>> fiber) system Koruza (http://koruza.net/), which are useful especially
>> in high-density urban environments because of no interference.
>>
>> Section 4.5, typical scenarios:
>>
>> I do not see usefulness of this categorization, because almost any
>> network I know of outgrow and changed through time inside all these
>> categories. Community networks maybe start somewhere (like urban or
>> rural area), but then they grow and spread over the whole country,
>> then start connecting with other countries.
>>
>> Section 5, classification of alternative networks:
>>
>> Introduction again focuses on incentives. That is a very limited
>> perspective. I would claim that most alternative networks go beyond
>> just incentives, but exist because of various beliefs: about how
>> networks should operate, who should have control over them, the
>> importance of commons and community stewardship of commons, etc.
>>
>> Section 5.1, community networks:
>>
>> As I explained, goals and motivation here seems a pretty limited list
>> and I think you should include at least ones I listed above. But
>> probably you should do a survey and ask community networks what are
>> their goals and motivations, instead of trying to guess them. (If you
>> have done such a survey, I would love to see data.)
>>
>> Technologies used: TDMA should be there, Ubiquiti gear does it by default.
>>
>> Typical scenarios: all
>>
>> Community Networks are large-scale: not necessary, they can also be
>> small, city only. Some are large scale, some are small, some are
>> focused on one region, some are distributed around larger region,
>> connected with VPN together.
>>
>> "There is a shared platform (e.g.  a web site) where a minimum
>> coordination is performed.  This way, community members with the right
>> permissions have an obvious and direct form of organizational control
>> over the overall operation of the network (e.g.  IP addresses,
>> routing, etc.) in their community (not just their own participation in
>> the network)."
>>
>> This can be true, but it is not necessary so. There are community
>> networks with much larger focus on decentralization. Especially with
>> IPv6 a lot more autoconfiguraiton is possible.
>>
>> Also, phrase "this way" is strange in this context. I do not see how
>> control over the routing of the network has anything with the
>> existence of the shared platform? In wlan slovenija network we also
>> use a web shared platform, but that platform does not control the
>> network. It just helps in coordination. Network would operate even
>> without it, just people would have a bit harder time coordinate about
>> use of IP space, and learning how to configure their nodes.
>>
>> So having a shared platform can have very little to do with how core
>> network's resources (IP addresses, routings, peerings, DNS entries,
>> etc.) are managed.
>>
>> "A Community Network is a network in which any participant in the
>> system may add link segments to the network in such a way that the new
>> segments can support multiple nodes and adopt the same overall
>> characteristics as those of the joined network, including the capacity
>> to further extend the network.  Once these link segments are joined to
>> the network, there is no longer a meaningful distinction between the
>> previous and the new extent of the network."
>>
>> Isn't this definition of the Internet? Just replace "segment" with
>> "autonomous system" and "community network" with "Internet". :-)
>>
>> So what exactly is the property which differentiates community
>> networks? Maybe there is none. Maybe community networks are simply
>> trying to (re)build the Internet, but this time having infrastructure
>> owned by participants, which are not just consumers, but are
>> participating.
>>
>> I think this is the most important characteristic. Organic growth, and
>> that people own the equipment, and thus the emergent network. This
>> should be more clearly explained. So it is not so much about
>> technology in community networks, but about who owns and controls
>> equipment (people/users/participants), and who coordinates the network
>> growth (people/users/participants). The line between users, providers,
>> participants, people gets blurred.
>>
>> "In Community Networks, everybody keeps the ownership of what he/she
>> has contributed.
>>
>> Not necessary. They can also give it away (for example equipment for
>> backbone nodes). In general, mostly people do not really track
>> ownership of equipment, because they do not care about ownership if it
>> is operating according to common principles. And if it is not, it
>> should not be in the network no matter who owns the equipment.
>>
>> Section 5.4, crowdshared approaches, led by the users and third party
>> stakeholders
>>
>> "VNOs pay the sharers and the network operators, thus creating an
>> incentive structure for all the actors: the end users get money for
>> sharing their network, the network operators are paid by the VNOs, who
>> in turn accomplish their socio-environmental role."
>>
>> I am not sure if all networks which can be grouped under this section
>> really do that. In the draft itself the https://openwireless.org/
>> movement is listed, but no shares or money or any incentive structure
>> is in place there for people to share their extra Internet
>> connectivity.
>>
>> So I would change this paragraph to:
>>
>> "VNOs can be organized to pay the sharers and the network operators,
>> thus creating an incentive structure for all the actors: the end users
>> get money for sharing their network, the network operators are paid by
>> the VNOs, who in turn accomplish their socio-environmental role. But
>> VNOs can also operate on gift-economy principles, where participants
>> contribute to the commons by sharing their resources knowing that this
>> benefits all."
>>
>> Section 6.2.1,  Media Access Control (MAC) protocols for wireless links
>>
>> "Wireless standards ensure interoperability and usability to those who
>> design, deploy and manage wireless networks."
>>
>> Wireless standard ensure low-cost of equipment due to economies of
>> scale and mass production. This is the main reason I think why WiFi is
>> so popular in alternative networks. You can get a device for $10. And
>> you can get this because everyone is producing this hardware, and the
>> hardware is not just made for the alternative networks. In contrast,
>> traditional ISPs use hardware which is developed only for them. Even
>> if they are big ISP, number of units is still much smaller than number
>> of WiFi routers produced for the global market.
>>
>> Question about list of WiFi standards in this draft. Why is that
>> needed? Why properties and descriptions of those standards should be
>> in this draft? They belong to Wikipedia, or their own standard. In
>> this draft we can just reference those standards. Like what of all
>> this text is relevant to alternative networks? It is general text
>> which is true for any use:
>>
>> "802.22 [IEEE.802-22.2011] is a standard developed specifically for
>> long range rural communications in TV white space frequencies and
>> first approved in July 2011. The standard is similar to the 802.16
>> (WiMax) [IEEE.802-16.2008] standard with an added cognitive radio
>> ability. The maximum throughput of 802.22 is 22.6 Mbps for a single 8
>> MHz channel using 64-QAM modulation. The achievable range using the
>> default MAC scheme is 30 km, however 100 km is possible with special
>> scheduling techniques. The MAC of 802.22 is specifically customized
>> for long distances - for example, slots in a frame destined for more
>> distant Consumer Premises Equipment (CPEs) are sent before slots
>> destined for nearby CPEs.
>>
>> Base stations are required to have a Global Positioning System (GPS)
>> and a connection to the Internet in order to query a geolocation
>> spectrum database. Once the base station receives the allowed TV
>> channels, it communicates a preferred operating white space TV channel
>> with the CPE devices. The standard also includes a coexistence
>> mechanism that uses beacons to make other 802.22 base stations aware
>> of the presence of a base station that is not part of the same
>> network."
>>
>> Section 7.1.2.2, mesh routing protocols
>>
>> "A large number of Alternative Networks use the Optimized Link State
>> Routing Protocol (OLSR) as defined in [RFC3626]."
>>
>> Not really. Networks use OLSR as implemented by http://olsr.org/,
>> which is far from the standardized OLSR in RFC3626. For example, in
>> practice, ETX metric is used, which is not even mentioned in RFC3626.
>>
>> Also Babel should definitely be mentioned, it is used in many
>> networks, and it is even (properly) standardized as RFC6126. So if you
>> want to include a routing protocol for alternative networks with IETF
>> standard, you should include this one for sure.
>>
>> Section 7.2.1, traffic management when sharing network resources
>>
>> It is interesting that so many people believe that there have to be
>> some special prioritization done for sharers to be able to use APs.
>> But it is not necessary true. Often people connecting to open AP nodes
>> will be much further away from the AP than the sharer. So often just
>> this distance already influences that the users of open network are
>> prioritized less (they have higher packet loss, lower bitrate, which
>> is also good to limit the lowest allowed bitrate).
>>
>> In community networks is also pretty common to run the network itself
>> on different frequencies than the APs. Some first generation mesh
>> networks ran everything (backbone over ad-hoc) and client-serving APs
>> on the same channel, but with 5 GHz spectrum and cheap dual-band
>> devices this is often separated now.
>>
>> Section 7.3., services provided
>>
>> What is this section? A non-comprehensive list of services on the
>> Internet and networks in general? This looks pretty useless section
>> which would not inform anyone reading this draft of anything about
>> alternative networks.
>>
>> If something, then it would be interesting to talk about specialized
>> services developed just for community/alternative networks:
>>
>> - Inter-network peering/VPNs: https://wiki.freifunk.net/IC-VPN
>> - Local wikis like: https://localwiki.org
>> - Community oriented portals: http://tidepools.co/
>> - Network monitoring/deployment/maintenance platforms
>> - VoIP sharing between networks, allowing cheap calls between countries
>> - Sensor networks and citizen science build by adding sensors to devices
>> - Community radio/TV stations
>>
>> What is interesting that some networks do not even provide Internet
>> access. For example, in Croatia, historically, there were wireless
>> communities which made networks in villages just to be able to play
>> games.
>>
>> Section 7.3.2.1, web browsing proxies
>>
>> "Other services (file sharing, VoIP, etc.) are not usually allowed in
>> many Alternative Networks due to bandwidth limitations."
>>
>> That would go against net neutrality and anti-censorship principles
>> which are important in many other alternative networks. So the
>> question how informative this sentence is, because for some maybe this
>> is true, for some it is not, and some probably are build just around
>> this. Others probably address this with innovative solutions like
>> internal file servers.
>>
>> At the end, a general question, how would DIY ISPs
>> (https://www.diyisp.org/) be categorized according to this draft? To
>> me is unclear.
>>
>> Some more projects to look into, and think how they relate to this draft:
>>
>> https://rhizomatica.org/
>> http://www.servalproject.org/
>> http://villagetelco.org/
>>
>>
>> Mitar
>>
>> --
>> http://mitar.tnode.com/
>> https://twitter.com/mitar_m
>>
>> _______________________________________________
>> gaia mailing list
>> [e-mail]
>> https://www.irtf.org/mailman/listinfo/gaia
>
>
>
>
> --
> Arjuna Sathiaseelan
> Personal: http://www.cl.cam.ac.uk/~as2330/
> N4D Lab: http://www.cl.cam.ac.uk/~as2330/n4d



-- 
http://mitar.tnode.com/
https://twitter.com/mitar_m