Feedback to the Dot story

[Aaron J. Seigo]

On Thursday, September 5, 2013 12:57:49 Thomas Pfeiffer wrote:
> On 05.09.2013 12:26, Aaron J. Seigo wrote:
> But if you give it to some ordinary consumers (even with an introduction
> of key features, without that they're totally lost and I have data to
> prove that as well) and give them some tasks to complete, they get lost
> and they get disappointed, even though they, too, like the general concept.

I’m going to keep repeating this until you get it, Thomas:

there are things PA does not do great at. i can come up with tasks that will 
stump people or show case inefficiencies or even outright bugs, too.

there are things PA does very well. i can come up with tasks that highlight 
that.

so what do we do? wait until there are only tasks that a majority of random 
test subjects do well on?

that’s a fantasy land concept and is impossible to achieve: by chasing that as 
the requirement to say “here is our product” we will never get a product out. 
with an eternally beta product due to chasing perfection, we will never grow 
demand and support in a way that will get us to more polished results.

seriously, go back and use the first production versions of iOS or Android. 
they are as polished as they are *today* because they took the approach i’m 
suggesting *then*. they still have tons of warts, bugs and stupidities baked 
in ... yet they are the bar by which all else is measured right now. this in 
spite of the fact that i can also come up with all sorts of tasks that stump 
people on their current Android devices that really shouldn't. 

amazing, right?

> To me, this is not a "polished product", this is still an (advanced)
> early adopter / developer version.

i understand you feel this way (i noted so in my earlier email, even). this is 
not your call to make, however.

> And since we target early adopters
> and devs, why do we call it "polished product”?

we target early adopters and devs because there are no devices that it comes 
pre-installed on.

> > real products ship. real products are not perfect. yet somehow people use
> > them and even fall in love with them.
> 
> Release early, release often, I know. That's what all FOSS should do and
> most do. But e.g. Simon or KTp give their - already pretty mature -
> products 0.X version numbers and say they'll call them 1.0 when they
> find nothing really missing anymore (that doesn't mean they cannot be
> improved anymore, though).
> I know this isn't about version numbers, but I've never heard David call

yes, it is not about version numbers.

that said ... Simon is quite mature; KTp still has a ways to go. regardless, 
one of the hurdles KDE projects face is this perfectionism bleeding over into 
public communication.

we can, and do, have all sorts of valid criticisms of the products produced.

for most people, that doesn’t matter one bit. and so when we go out and tell 
people how *bad* our software is (which is exactly what your messaging 
translates to in people’s minds) they just don’t use the software at all or if 
they do they focus on the negatives we’ve  primed them for.

the result is that they do not use software that would actually work *just 
fine* for them and even make them happy.

why? because we’ve pointed out what it does not do well in an attempt to over 
our ass instead of highlighting what it does do well and focusing people on 
that.

> I don't mean to say that we did bad work, because we didn't. We did
> great work, but we still have more great work to do before we have a
> "polished product". PA4 is great, but still very "rough" in a whole lot
> of ways.

shall i list all the bugs i run into on a daily basis on my (not very old) 
Android phone?

shall i share the pain of watching my brother-in-law use his brand new top-of-
the-line Android phone while he complains about this or that as he fiddles 
about with it?

you are trying to live in a world of perfection that doesn’t exist, and that 
mindset is an existential threat to the  project.

-- 
Aaron J. Seigo
-------------- next part --------------
A non-text attachment was scrubbed...
Name: signature.asc
Type: application/pgp-signature
Size: 198 bytes
Desc: This is a digitally signed message part.
URL: <http://mail.kde.org/pipermail/active/attachments/20130905/21b128c3/attachment.sig>