My Computer

Andreas Hartmetz ahartmetz at
Mon Nov 26 23:02:20 GMT 2007

Am Montag 26 November 2007 21:45:22 schrieb Sebastian Kuegler:
> On Monday 26 November 2007 21:37:15 Andreas Hartmetz wrote:
> > > - My Computer actually has usability testing behind it.
> >
> > It is customary in scientific circles to present the precise setup, the
> > method of measurement and the raw results of an experiment if it is
> > supposed to prove something. Otherwise this amounts to little more than
> > "We know the truth and you don't. You lose.". There might be something in
> > the raw data that helps to make this discussion a bit more useful.
> So the logic here is
> " we don't have any data proving Places so we'll replace My Computer
> because you don't have any data but pretend to do so. "
> Just FYI, this data has been presented during aKademy 2006, and I wouldn't
> know why to lie about the results.
The report from aKademy 2006 says nothing about wording *at all*. It looks 
like a presentation that was made to impress the audience.
Have you seen how performance issues are discussed on the Linux kernel mailing 
list? The exact test setup and testing methods are explained and lots of 
tables with numbers are given. That is how it works.

> This is bordering ridiculous (on the opposite side of the river).

You can't argue for withholding data. That is very, very wrong.
Most of all, I expect people on this mailing list to be cooperative. So far 
there has been nothing from coolo - he is just ignoring this and you 
dutifully step in to end the discussion (because discussions waste time [I 
agree]) using the sledgehammer approach.
Not so fast.
Coolo, where's the data? Please?

> And don't forget, this is not a scientific circle, this is a meritocratic
> project _in_string_freeze_.

The scientific method is the basis of a meritocracy and it has had a little 
success recently. If you *do* perform experiments it's the only way to go.

At that point I don't care much about "My Computer" or "Gnorp" but about 
unfalsifiable statements derived from "internal" data that nobody can check. 
That's not how friendly (yes, friendly! Dammit!) cooperation works.

More information about the kde-core-devel mailing list