Gene Collins wrote: >> I'm not qualified to comment on the technical merits of a kernel vs. a >> user-space solution, but I know that from a maintenance point of view we >> would prefer it. >> >> > Well, you haven't told us how Speakup adds to your mainttenance > problems. I mean the only thing extra you are doing is running a > checkout script, in order to patch speakup into the kernel. After that, > you are just doing a make oldconfig with the .config file from your most > recent kernel. The kernel is a complex beast. Anything you put into it can have unpredictable consequences on the whole. We ship the distro to millions of users every 6 months. Those versions all need to be maintained with updates for 18 months, and with LTS for 5 years. Any extra non-standard kernel code becomes a maintenance burden. If you have to merge in a security patch from kernel.org it helps if your kernel is as close as possible to what kernel.org was when you released. Obscure patches with light testing that you have applied earlier on will become a burden. > The problem with having only software synthesis on the system is that > when there is a problem, the software synthesis is often the first > thing to stop functioning. Again, that's a bug and a community like this one would do well to help find and resolve those. That would help a much larger, less technical group in the long term. That is also my main focus. I had hoped this highly skilled community would help me with that. > But when blind folks ask for a solution like speakup and a hardware > synthesizer which will function even during a kernel panic sometimes, > sighted folks seem to have real trouble understanding the need for > reliable access. > That is completely false. The way that speakup made its way into ubuntu originally was that a visually impaired member of our community suggested it for inclusion. He drew up a proper specification for it which we discussed at the Ubuntu Development Summit in Paris. We held a workshop session about speakup where I and our kernel team was present. There was no lack of understanding for the importance of accessibility anywhere in this process. (we regularly invite users and developers with visual impairment to our dev summits, in November we have a a whole mini-conference about this). We agreed that we would include speakup in our default kernel for Edgy, but the maintainer expressed concerns about future maintainability already at that point. We agreed that we would engage with the speakup community and offer help with getting it into mainline. I did that, but nothing much came of it. For gutsy we have changed our policy to actively reduce the number of non-mainline patches. Speakup is clearly not on track for inclusion at this point and so falls under that policy. > I don't recall seeing any requests here from Ubuntu folks for help with > maintenance issues for Speakup. http://speech.braille.uwo.ca/pipermail/speakup/2006-October/040394.html http://speech.braille.uwo.ca/pipermail/speakup/2006-October/040435.html http://speech.braille.uwo.ca/pipermail/speakup/2006-October/040639.html http://speech.braille.uwo.ca/pipermail/speakup/2006-November/040979.html http://speech.braille.uwo.ca/pipermail/speakup/2006-November/041049.html Lots of debate. Little or no direct involvement from the speakup community in ubuntu to help improve our support for it or help maintain or move it forward. > It appears to me that what we have here > is a group of sighted folks who don't use the technology making > decissions about what blind folks need or don't need, want or don't > want. In my view, that's inappropriate. > You are not doing yourself any favours by making this into a blind vs. sighted issue. The blind/vi Linux community is quite numerous and could have a massive impact on progress in general accessibility on the platform and could help other weaker groups in the process. But to do that you have to get organised into a community structure and engage constructively with other parts of the free software community. That's how things get done in the free software world. This list seems to be doing a god job at providing peer-help using the existing tools, but I see little innovation and reaching out to other projects going on. I don't see a roadmap or advocacy campaigns for example. > I make my living as a computer consultant. It will be to bad if I'm > forced to recomend that folks stay clear of Ubuntu because of it's lack > of support for text console accessibility, but if that's the choice I'm > force to, then that's what I'll do. I hope the folks managing Ubuntu > will reconsider their decission, but if not, all I can do is steer clear > of distros that insist on causing me grieff. This is what I mean by 'not engaging constructively with the wider community'. Henrik
