[wsfii-discuss] the cooperative way for India?

Sun Jul 9 10:05:39 UTC 2006

You are hitting good points Aaron, comments in line.

En/na Aaron Kaplan ha escrit:
> (...)
>   
>> Some of our conclusions were:
>> -A local community that just relies with the ability and availability of 
>> a few networks admins can run a network of just about a few dozens of 
>> routers
>> -By having some utilities, such as a database, wiki ant network console 
>> (cacti, nagios....), it can be up to a hundred interconnected nodes?
>>     
>
> Well, the current systems that I know of basically lack a scalable layer
> 2 (WLAN). So IMHO this goes first when it comes to scalability.
>   
Yes, the "L" of wLan means "Local". They are designed just for a few 
clients. Does not scale to something larger by his own nature.
But here I was not referring to wLan scalability, instead, I was talking 
on tools for having lots of wLans, and linked together in a single 
larger network by PtP links, etc.
If that's has to be done manually by expert net admins, those admins 
will become the bottleneck and jeopardize the goal of easily allow 
acquiring new connections from new users, and building links with the 
communities of everyones  neighborhood.
>   
>> -By building a database automatic IP provisioning system, maps, forums, 
>> built-in online network monitoring, configuration tools etc... that 
>> could bring us up to thousands?
>>
>>     
>
> Maybe this helps:
>
> We released our funkfeuer wireless mesh ISP software as GPL. 
>
>    http://redeemer.sf.net
>
>
> It includes: node registration, user admin, VOIP integration (users can
> self administer), a nice google map, smokeping, nagios integration, etc.
> the idea is basically that we explain to people what to do, how to do it
> and then give them an account on our ISP software. There they can
> register IPs, register VOIP extensions etc themselves.
> The best feature about it: you can easily write small drivers that will
> feed your other systems from the node/user Db. (coding credits go to
> Wolfgang Nagele)
>   
Cool. Yes. That's the point.
Quite similar in many areas with the apps we are using at guifi.net
The next point here has to be to forget about "easy drivers". Coding is 
only "easy" to coders. Network users/promoters don't know how to build 
drivers, in rare cases are programmers and just use what's available, 
therefore the best will be simply be able to agree with an open XML 
format, with "drivers" already built.
Something like this:
http://guifi.net/ca/guifi/nodexchange/2444/nodes
So, let's work on it ;)

> I know many of the people in the free networking and meshing community
> don't feel happy about a centralized approach as described above.
>   
BIG COMMENT HERE: I know about that feeling, but very often that's a big 
fatal error and common misunderstanding. Being able to interchange 
networks descriptions in XML, if someone wants to built in a innovative 
way, will always have a chance to implement it. 
But if simply builds the network in some way that only himself can 
manage/know, in the real world that's not an open network because is 
unable to easily inter operate with other networks, in other words, 
that's closed to himself or his community. I mean, can be based on open 
source, but could result a proprietary network. What's more 
"centralized" than admins taking decisions of what/how/when and without 
publishing their networks?
As an example, is like when someone says that is producing Open 
Software, and certainly is willing for, but has only a working copy in 
his well firewalled server at home and no CVS/SVN/tarball.
In the meantime, local wireless communities are still de-centralized by 
his own nature, you know, wLan (L=local) works only in a local 
environment, therefore, why the people have to be afraid? afraid of what?
The real challenge to scale is the ability to connect those still local 
networks or consolidate them in something larger, and the turnkeys, the 
methods and tools that allow this. In the real world, without those 
"enablers",... just sexy local wLans, never more. We can talk a lot, 
have tons of code, but do nothing.

> So...
> There is a very nice second project "AcDc" (which I am proud to be
> mentoring at the Google SoC)
>   
> http://www.reseaucitoyen.be/wiki/index.php/AcdcProposal_en
>
> It is a decentralized p2p DNS & captive portal system with XML beneath
> it for describing the configuration.
>   
Interesting. I'm not very familiar en ad hoc networks for the very last 
mile in urban areas. I do realize that distinct scenarios might have 
different solutions. That shows that different solutions to different 
scenarios have to be complementary.
To give you an example, an approach of "no node can be more important 
than another" can be true in a region-scale network like this:
http://guifi.net/ca/node/2444/view/map
(> than 500 km2, combining small to medium urban areas and rural areas, 
 > 800 radio devices, some of them having dedicated p2p links (we call 
them "supernodes"), other simply regular users or repeaters...)

BTW, can you provide the real world communities/sample servers to see 
the apps. you mentioned live?

>   
>> Wireless community is not just about socializing, wifi state-of-the-art 
>> technology is a huge challenge and it isn't at the hand of a few 
>> volunteers. There are never excedent on volunteers.
>>
>> Major problem I do envision now here is uniformity. Local communities 
>> are plural and self-centered (they should be, that is the basis). No way 
>> to share anything if there are no common practices. So what we can share?
>>
>> There have to be a motivation for sharing not only experiences, also 
>> resources. How many of us are really sharing resources across 
>> communities? But might be different approaches for solving technology 
>> problems (i.e. routing, hardware...) so.... what?
>>
>> What about at least having the ability to syndicate community networks 
>> by describing them in XML? So regardless of how you do run a given 
>> network, you still are able to see others, if someone finds out a good 
>> hardware configuration tool solution, by providing your network 
>> information to it you are just able to take advantage of it without 
>> disrupting other features that maybe you have and love...
>>
>>
>> Ramon.
>>
>>
>>  
>>
>>
>> En/na Ken DiPietro ha escrit:
>>
>>     
>>> Vickram (et al)
>>>
>>> I find this thread fascinating from a couple of different standpoints. 
>>> First and foremost is the dynamic we are discussing between, 
>>> government, private business and local communities. I applaud you 
>>> being able to clearly and concisely explain some of the reasons each 
>>> component of this structure is not capable of managing this work on 
>>> it's own.
>>>
>>> If that is the case, I would suggest that each one of these segments 
>>> must take responsibility for a piece of the job and this is how I 
>>> would suggest one possible solution could be implemented.
>>>
>>> Government -  The national government in partnership with local 
>>> governments create a training facility to educate a staff of young, 
>>> resilient workers capable of deploying this equipment even under the 
>>> most adverse local conditions. Additionally, government either 
>>> mandates the manufacture of this equipment at a subsidized price or 
>>> provides tax incentives to make the manufacturers of this equipment 
>>> eager to produce it. From what I can tell, the government will also 
>>> have to remove the mountains of red tape that it seems to thrive on to 
>>> make this project go forward as well as removing any tariffs that 
>>> might also add to the cost of this equipment.
>>>
>>> Business - Private sector must be motivated to build this equipment. I 
>>> would suggest that this could be done by using a combination of tax 
>>> incentives weighed against the threat of allowing the importation of 
>>> goods should the local businesses not meet the demand in a reasonable 
>>> time period. Specifications for interoperability as well as 
>>> environmental hardening need to be mandated. Another possibility would 
>>> be for the government to subsidize the training of qualified employees 
>>> to manufacture this equipment.
>>>
>>> Local communities - This is where the demand is generated. One of the 
>>> ways these communities can pay for this connectivity is to supply 
>>> labor to the manufacturers as well as to the installation staff. These 
>>> trained people could return to their communities to keep the network 
>>> operational after a set period of time and it would be up to the 
>>> community as well as government to pay for the workers (or a portion 
>>> of their salary/expenses) during this time.
>>>
>>> Please note - many of these communities will be able to work together 
>>> as this connectivity will be supplied by passing through one community 
>>> on the way to delivering to another community further down the line. 
>>> It will be critical that these groups can work together or the segment 
>>> of the network will eventually collapse.
>>>
>>> I realize this outline is full of holes and is not intended to be a 
>>> boilerplate for how this project should be rolled out but rather a 
>>> staring point for discussion.
>>>
>>> Respectfully,
>>>
>>> Ken DiPietro
>>>
>>> New-ISP.net/NextGenCommunications.net
>>> Wireless solutions - not concessions.
>>> http://www.nextgencommunications.net
>>> 1044 National Highway LaVale MD 21502
>>> Tel# (301)789-2968 Cell (301)268-1154
>>>
>>>
>>> vvcrishna at radiophony.com wrote:
>>>
>>>       
>>>> Quoting Balaji G <balaji_g1947 at yahoo.com>:
>>>>
>>>>  
>>>>
>>>>         
>>>>> What should these mechanisms be ? I have my doubts on the scaling of
>>>>> community based approaches, but one must develop the area further, 
>>>>> as well
>>>>> as explore other mechanisms.  Hope is in strenghtening the energies of
>>>>> communities as against burdening them with another development task, 
>>>>> which
>>>>> is easier done by Government and moneyed industry.     
>>>>>           
>>>> Community based approaches, on the contrary, are probably the most 
>>>> powerful in
>>>> terms of scaling. Half the reason that mobile networks in rural areas 
>>>> are so
>>>> pathetic as that no urban executive wants to go spend months in the 
>>>> boondocks
>>>> wrangling with ticklish local issues of electricity, workers, spares, 
>>>> drinking
>>>> water, food etc.
>>>> That critical problem is erased when the work is community-driven. But
>>>> unfortunately, decades (centuries) of autocratic rule has damped down 
>>>> the
>>>> desire of most rural Indian groups to do things for themselves, which 
>>>> is why
>>>> the most genuinely successful ICT projects are in places where a 
>>>> significant
>>>> amount of preparatory work has been done, sometimes for years.
>>>> Depressing, perhaps, but no reason to lose hope or abandon this 
>>>> country's fate
>>>> solely to 'big' governments and (essentially greedy) private 
>>>> businesses to come
>>>> to the rescue.
>>>> Vickram
>>>>
>>>> _______________________________________________
>>>> wsfii-discuss mailing list
>>>> wsfii-discuss at lists.okfn.org
>>>> http://lists.okfn.org/mailman/listinfo/wsfii-discuss
>>>>
>>>>
>>>>
>>>>   
>>>>         
>>>
>>>       
>> _______________________________________________
>> wsfii-discuss mailing list
>> wsfii-discuss at lists.okfn.org
>> http://lists.okfn.org/mailman/listinfo/wsfii-discuss
>>
>>     
>
>
> _______________________________________________
> wsfii-discuss mailing list
> wsfii-discuss at lists.okfn.org
> http://lists.okfn.org/mailman/listinfo/wsfii-discuss
>
>
>   