[Fwd: Re: Edubuntu: Gallium vs. Kalzium]

[Edward Holcroft]

Oliver Grawert wrote:
> when did i say i'd be "reconsidering my opinion" ??
>   
Read your own mail for goodness sake. You said, and I quote:

"but that mail attack from the kde side
against me personally (about a tech problem we try to solve) starts
really making me reconsider my opinion here."

But let's move on from this, it's no longer a useful converstaion. More importantly:

>> ... on the other hand, how about the DVD/multiple CD version of Edubuntu 
>> that some of us have been calling for since day one of this list. Then 
>> there'd be space for so many more great apps. That's what this argument 
>> should bring about in my view.
>>     
> multiple CDs wont be shipped through shipit simply due to costs. our
> majority of users lives in areas where you neither have the bandwith to
> download a DVD nor find DVD writers (or even readers)... do you really
> want exclude these people from edubuntu ? 
>   
In my view Oli, the exclusion is already happening as a result of the 
current approach. I do think that this is a worthwhile debate to get 
into and I am pleased you're discussing it. Hopefully it will move 
beyond discussion and be transformed into action while it's still a 
relevant issue. I live and work in Africa and have deployed many 
hundreds of Linux thin clients in schools in several African countries. 
As such, I can say with authority that by publishing Edubuntu on one CD 
you are doing exactly what you hope to avoid: excluding people. While 
the single CD approach has certain merits relating to your own 
distribution process, any belief that this is a universal model, is 
based on a flawed understanding of how Linux is best distributed in 
Africa. It is a bigger problem for us in Africa to deal with a single CD 
with massive online repositories that people will never be able to 
access, than to overcome the hurdle of a once-off big download whenever 
a new release comes out.

In response to the shipit issue: you can keep the shipit option as a 
single CD option to keep costs down. Frankly, I think you should drop 
shipit altogether and use the money to develop the multiple CD version. 
All you need is have links on your website to the myriad of entities in 
Africa that will copy Linux distros for anyone for free. And they'll 
probably get it faster than by shipit.

In response to the issue of those folk without DVD drives: You can offer 
a multiple-CD version for these people. Edubuntu will not be the first 
distro to offer multiple CD as well as DVD options. Just by the way, one 
would be hard pressed (yes, even here in Africa where it seems we don't 
know what's best for ourselves) to find a new Pentium 4 box which will 
be deployed as an LTSP server that does not ship with a DVD drive.

In response to the bandwidth issue: yes this is a challenge BUT all it 
takes is one person to grab the disks off the web and they can be copied 
locally and they will multiply like bunny rabbits. Copies distributed in 
this way spread like wildfire - just ask Microsoft :-). This is the 
primary mechanism by which distros make it to rural African villages 
without the luxury of shipit. Furthermore I am deploying school labs in 
areas that would not even have access to Internet for placing a shipit 
order. The vast majority of schools in Africa are in this position. 
While on the face of it this may seem contradictory, if you unpack the 
issues, you can see that it is in fact the very reason why we need to 
see Edubuntu offer a bigger CD/DVD-based repository. Until there is 
fast, cheap broadband connectivity generally accessible in rural African 
villages I cannot seriously consider deploying a single-CD distro. We 
want to give schools a bigger selection of apps as part of the initial 
offering than what is possible on one disk.

Is it possible to have a multiple CD version that will install a basic 
working system off one disk, but the more disks you have, the more apps 
you can add? One in which the installation process asks which disks you 
have available. Of course it is. Others are doing it. This is what we 
need in Africa for now and probably the next few years. I suspect the 
same is true of many other underdeveloped nations.

A bitter irony for me, is being forced to beg someone sitting in Europe 
to help meet our needs in Africa, when the *buntu Linux concept was 
conceptualised and paid for by a South African. Maybe I should have 
majored in computer science instead of English Literature, then perhaps 
I could do what so desperately needs to be done here ;-)

cheers
ed
PS Apologia:
Please don't get me wrong, I think Edubuntu is poised for great things, 
it's just that there are better ways of dealing with certain grassroots 
issues, especially in rural Africa.