On 29 Jan 2003, Kirk Reiser wrote: > Nicolas Pitre <nico@cam.org> writes: > > > With initrd and the upcoming initial ramfs about to become mandatory in > > Linux 2.6 or 3.0 there is absolutely no reason for such support to live in > > the kernel. This, plus the current trand about all speech synthesis made > > into software through a sound card is yet more reason for a user space > > solution. > > > > Unfortunately, the speakup developers just don't seem to "see the light". > > The speakup developer saw the light a long time ago and refuses to > make second class citizens out of blinks. There is no reason in the > world why a blind person shouldn't have speech from boot up if it's > possible. It's way more valuable for a blind person to have a much greater choice of speech synths, including much cheaper ones, than having speech output from the very milisecond the kernel starts using printk(). IMHO persisting with a kernel solution is what actually makes second class citizens in the end. > That is not to say thaspeakup does not have flaws. I haven't seen > developers crawling out of the wood work to help though. I have seen > a lot of blind people so used to being taken by the hand and helped > that they cannot even imagine being independant. Sad but true. Let me say that I'm blind. I'm also one of the principal ARM Linux developers. That's my day job. Try Google with my name if you wish. I'm therefore pretty knoledgeable with Linux kernel development, and I do that all day long on Linux with both braille and speech support. I also wrote a good chunk of BRLTTY. I'm currently working on a software French speech synth for Linux in my very little free time I have left and I intend to give it away for free too. This being said I can't agree less with what you said above, but I wanted to make sure you don't put me in the same category and that you truly consider what I say as coming from an experienced blind user and developer. > Speakup will probably have a userspace component at some point to allow > the use of devices like the Dectalk PC and software synthesis but it > will never relegate blinks to the back seat of computing. I really really don't understand what makes you say that. Please would you care to enlighten me on why you're picking that stance when it's question of user space solutions? In my opinion a fully user space Speakup would have much more benefits and can be AS RELIABLE as the kernel solution when it's time to give access to early kernel stuff and I'd like to pursue a technical discussion with you on this issue. > [Good and pissed at folks that moan but can't take their thumb out > of their asses to help] Really? Nicolas