[olsr-dev] I'm thinking about putting a patched - unofficial - version of olsrd-0.4.10 online

[Dan Flett]

Hi electra,

I'll throw my support behind this idea!  You're right - getting the right 
patches and knowing the *real* optimised config-file settings is like 
finding your way through a moving labrynth.  This info needs to be 
centralised somewhere.  Ideally, olsr.org would be the place - but recent, 
real-world documentation is hard to find there.  Also, a lot of guys I talk 
to in Australia find it hard to find *any* useful information about OLSR in 
English at all.  They go to the freifunk site and are turned off by all the 
Deutsch.  Personally, I'm seriously considering learning German to get a 
better understanding - if I can find the time - but not everyone here is so 
dedicated.

Cheers,

Dan


>From: elektra <(spam-protected)>
>Reply-To: OLSR development <(spam-protected)>
>To: OLSR development <(spam-protected)>
>Subject: [olsr-dev] I'm thinking about putting a patched - unofficial - 
>version of olsrd-0.4.10 online Date: Sat, 13 Jan 2007 16:18:22 +0100
>Hi -
>
>first of all: Please correct me if I'm wrong in anything stated here.
>
>I wanted an up to date version of olsrd to build into a upcoming release of 
>Meshlinux. I learned recently that many important patches didn't make it 
>into CVS yet.  Sven-Ola has a collection of patches at 
>http://212.222.128.68/sven-ola/nylon/packages/olsrd/files/
>
>Since it is cumbersome to apply them one by one in the correct order I 
>asked for a patched version of the code.
>
>I think it is high time that an up-to-date resource 'ready to compile' is 
>published somewhere - at the moment getting a reasonable and stable working 
>version of olsrd is something that only an insider can achieve. To 
>exaggerate a bit:  you have to be in a certain club in Berlin at night on a 
>certain day and drink beer with the right guys to get the URL where you 
>find the pieces of a puzzle... Not to mention that you have to be aware 
>first of all that you need exactly this code, because it:
>
>* consumes a fraction of cpu-load
>* doesn't crash if it has to calculate more than 31 hops between two 
>destinations - believe it or not, Freifunk really ran into this problem...
>* LinkQualityDijkstraLimit switch works
>
>to mention a few.
>
>This code is tested in Berlin and many other cities that have meshes 
>working 24/7 since it is included in the Freifunk Firmware. A vague 
>estimation claims that between 1000 and 5000 people use the mesh in Berlin 
>every day for their communication.  The Berlin Freifunk mesh has 400+ nodes 
>- running mostly on of-the-shelf hardware for 40-60 Euro per unit. There is 
>no central administration - individual people administrate their own 
>routers. This is the situation here - other Freifunk networks (Leipzig for 
>example) are closing up to this figures.
>
>I have seen the big excitement and enthusiasm in Dharamsala/India that 
>people have about their - yet small - mesh. And I learned that they were 
>not able to configure it properly. The participants of the WSFII/AirJaldi 
>Summit faced instability of the internet connection - since the Tibetan 
>Technology Centre didn't know which configuration file to use and that they 
>had to enable FishEye to get rid of routing loops under payload.  Many 
>people from India/Pakistan/Nepal/Bangladesh (...) are looking forward to 
>roll out meshes to mitigate their lack of communication infrastructure. I 
>was approached by people that are in need for information and working code.
>
>I think it is high time to state: There is an enormous communication 
>problem.  Mesh development is important for many people on this planet - 
>especially the poor (telemedicine a.s.o.). People need a new release *now*.
>
>Governments - especially in developing countries - need a well working 
>open-source solution to provide mesh communication infrastructure.  The 
>government of India has launched an ambitious program to connect 600 000 
>villages in rural India. A part of the job can be done with wireless mesh 
>and the Indian Ministry of IT is interested in this.
>
>Ok - that was a long introduction. Here is why I wrote this: I'm thinking 
>about putting the patched source archive of olsrd on line somewhere. There 
>is no point that a well-configured and up-to-date working version of olsrd 
>is only available if you install Freifunk Firmware or Meshlinux. That would 
>be  a good marketing strategy of course.
>
>I don't want to start a new development branch/fork/whatever of olsrd. Or 
>become a maintainer. I want to avoid doing unnecessary work or pissing 
>anyone off. I just want to have a URL with up-to-date code that I can point 
>people to. And I - finally - want to get rid of RFC3626 in this version. I 
>have made a change to the archive on my harddrive: I altered the behavior 
>of the Makefile and naming of the daemon. So far olsrd-0.X.X always 
>installs by default a configuration file that is RFC3626-compliant. I 
>consider this is a bug, not a feature... By default the changed Makefile 
>will install LQ with FishExe - because it is the way to go. Since this 
>modified bit would now do what we discussed about 0.5.0 one year ago - I 
>would publish it as 0.5.0-unofficial...
>
>Please tell me you opinion.
>
>cu elektra
>
>
>
>
>
>
>
>
>_______________________________________________
>olsr-dev mailing list
>(spam-protected)
>https://www.olsr.org/mailman/listinfo/olsr-dev