Bad info on RTAI
mantegazza at aero.polimi.it
Thu Oct 21 19:57:36 CEST 2004
Jonathan Corbet wrote:
> Paolo Mantegazza <mantegazza at aero.polimi.it> wrote:
>>In this case nonetheless, by stating that what is considered just a
>>"not to use any more" legacy only mode of operation is the RTAI only
>>working mode, it really took side with the worst FUD mouthing about
> Hmm, if that's the worst FUD you have to cope with, consider yourselves
Maybe I'm but, believe me, hiding that RTAI means user space, is the
best way to expand FUD. The reason is simple: the real freedom for most
users is there.
True hard real time, i.e. many Khz, is a niche and higher real time
performances, i.e. working in kernel space directly, a niche of a
niche, clashing against likely better DSPs power.
> My purpose in bringing RTAI into that article was to be sure that LWN
> readers were aware that the project is out there, that it provides a
> rather harder realtime guarantee than the "make lots of scary changes to
> the kernel locking model" approach, and that the project is moving
> forward. I was trying to do a favor for RTAI (and my readers); I
> certainly was not trying to FUD anybody.
> Obviously, I screwed things up, based on an understanding of RTAI which
> rather out of date. For that, I offer my apologies.
On the personal side I want to recall the following: "To my knowledge,
LWN has always been a fair source of information and it never denied
publishing any RTAI related annoucement I forwarded to them. In this
case nonetheless ... ".
So I've no doubt of what you say above. In fact admitting that you
"screwed things up" means we are saying the same. That's why I counted
up to ten and ended in just advicing people on this list.
So your apologies are even not needed because I keep believing what I've
believed always: you are doing a valuable work for the freedom software
> I'll find a way to set the record straight. To that end, I'll ask a
> favor: is there anybody out there who would be interested in writing an
> article painting a solid picture of where RTAI stands now, and where it
> is going? I can try to do that myself, but, if somebody involved with
> the project would like to take this on, the result is more likely to be
> a complete article without the need for another round of corrections
> afterward. LWN can pay (a small amount, in devalued US dollars) for the
> privilege of running it.
What suggested above is a good point. On my side it clashes against the
fact that I'm very very lazy at writing. Writing is just for when I've
nothing better to do, but I'll promise to think about it.
More information about the Rtai