[Arm-netbook] I Have An Possible Chance To Meet An ARM Senior Manager

[luke.leighton]

On Sat, Mar 16, 2013 at 11:02 PM, Alexander Stephen Thomas Ross
<maillist_arm-netbook at aross.me> wrote:

> Don't hope for free graphics from arm (there working on there own gpu.)
> or with arms help. If your prepared to buy it currently then that's good
> enough for them. Rickard MS is right. Complete, out right rejection of
> proprietary/non-free junk is the only way to go.

 alexander - thank you.  bit of an eye-opener for you, huh?  we see
the evidence of these peoples' blind faith in their position, from
their lack of understanding and having never been exposed to
*successful* free software projects... because they can't be bothered
to look.

 the classic one - if i had known that the conversation would have
gone this way - the classic one to mention is the fact that the
valve/steam team got together with the intel 3D team after both found
that they were having problems with their respective GPL-licensed code
bases; one of the team members wrote it up publicly as "the most
productive work meeting they'd ever had".

 apart from anything, the engineers could arrange the meeting
*informally* - without ever having to sign any NDAs or request
permission at Director level - on short notice.

 what this guy hasn't thought through is that it requires vast
resources and finances to create hardware.  it's typically an 18 to 36
month development cycle for any one product, and one mistake is
*seriously* costly.  $5m down the drain and a 6 to 9 month delay is
not uncommon.  make a second mistake, and you've missed the window of
opportunity for the *entire* product.

 as far as actually selling the chips, when it's such a damn risk for
any company to enter the market - the software is like...
insignificant by comparison!

 and if any company *was* to develop a 3D product, they sure as hell
wouldn't start by making it fully-hardware-compatible with someone
else's software, because that wouldn't give them an edge: they'd be at
least 2 to 3 years *behind* what the company they "copied" had
produced.

 instead they would try to design something innovative, wouldn't they?
 like icubecorp.  they're designing a hybrid CPU-GPU combination,
which hasn't really been done before.

 so, mr marketing person: if you're reading this - your strategy is
guaranteed to annoy people and to drive them *away* from your
products.  you're annoying the users, you're annoying the software
developers, you're annoying the hardware developers and you're
annoying the product designers.

 only the people who do not know the full picture are sticking with
you, but if they get burned enough times and an alternative is
presented to them, they will *walk away*.  especially if that
alternative is offered at a lower cost or with a faster
time-to-market.

 is that clear enough that your strategy is going to affect ARM's bottom line?

l.