illusion (Re: tap-to-click feedback)

Paul Fox pgf at
Thu Apr 15 15:31:46 EDT 2010

tuukka wrote:
 > Further, the deployment teams *cannot* represent the children. They 
 > should strive to, but there is no substitute for going out on the field 
 > and seeing with your own trained eyes.

while your point is well taken, the fact remains that no one but
the deployment teams is in a position to represent the children. 
i'm not sure how big an organization you imagine OLPC to be, but
it's not nearly big enough to spend enough time in the field to
allow replacing feedback from the deployment teams.  (which isn't
to say that we don't make field visits.  we do.)

 > > There have actually been very few reports of problems with this
 > > touchpad behavior. We don't know why that is, and I don't think you do,
 > > either - although we can both speculate all day and fill the list
 > > archives. We do get plenty of reports of other problems, so I don't
 > > think there's a fundamental failure in our ability to get input from the
 > > field. We can always hope to do better, of course.
 > We can't be sure and we could discuss, but we start to get the feeling 
 > that we do have a better idea than you do, and yet you're not willing to 
 > present your side and discuss.

let's remember that this conversation started with a discussion
of how we might go about changing the tap-to-click behavior. 
again, OLPC has limited resources, and we certainly don't want to
make mistakes in applying them.  it would be a real shame if we
went to the effort of removing tap-to-click only to find that
there was a large group of users that was perfectly happy with
it, and who, in fact, like it.  my personal opinion is that the
existence of such a group is unlikely, but ed is right to say
that we should be sure.

 > We haven't engaged here before, and maybe we still shouldn't have. All 

on the contrary -- we of course welcome your (constructive)
suggestions, and, in this case, your feedback directly from the
field.  the system (such as it is) seems to be working.

 > If you re-read our message, we mention some imperfections in how this 
 > issue has been handled from the beginning in the OLPC project. Please 

truly, in terms of mistakes made by OLPC, this one is probably
pretty low on the severity list.  richard has addressed this, but
in case it wasn't clear:  the touchpad hardware change was
partially involuntary on the part of OLPC -- it was largely due
to a supplier issue, but it had the side-effect of eliminating
the huge issues we were having with the previous pad.  the new
h/w happened to coincide with a major s/w release -- in fact, i
believe that it turned out to be the last major s/w release OLPC
did for the XO-1.  at the time, enabling the synaptics driver in
the kernel (which would have let us disable tap-to-click) caused
other problems which were far, far worse, so we decided not to do
so.  this seemed reasonable since the _only_ downside of not
doing so was the tap-to-click issue.  given the huge problems
we'd been having (and still have!) with the previous touchpad, i
think tap-to-click seemed very minor by comparison.

in any case, i guess we should put your votes in the "please
disable it" column.  thanks for letting us know.  :-)

 paul fox, pgf at

More information about the Devel mailing list