Re: fish names

Re: fish names

[Pavel Machek]

> > If the opinion of mainline linux is "we will not take your code because
> > we don't like what you named your hardware", okay, so be it.
> 
> We are perfectly capable of running sed over the patches... but you
> should be also capable of just running sed over your trees, and that
> would be better for anyone.

Just... having misnamed support for dream in mainline is way better
than no support at all.

But, consider this:

a) everyone knowing what trout means knows what dream means.

b) about 10 people total outside of google know what trout means.

c) proportion of non-google people working on the code will only
increase, because google is now working on nexus seven or something...

...so -- what is better name?

And what is the cost really? The patches need to be cleaned up,
anyway, so...?
									Pavel
-- 
(english) http://www.livejournal.com/~pavelmachek
(cesky, pictures) http://atrey.karlin.mff.cuni.cz/~pavel/picture/horses/blog.html

Re: fish names

[Bill Gatliff]

Pavel Machek wrote:
> And what is the cost really? The patches need to be cleaned up,anyway, so...?
>   

A higher priority is getting everyone (Google, OEMs, and the community)
to settle on a common tree, from which we all base our ongoing work. 
Then we won't have this ongoing need to clean up anything.

The path of least effort towards that goal is to stick with the names we
already have for the existing code.  At some point in the future, when
the code in question isn't so volatile, then we can mutually consider
whether renaming creates more pain than it eliminates.  Since by that
time we'll all be comfortable with the mapping between the fish and
commercial product names, I doubt there will be any motivation to do the
additional work.

I really don't have a problem with the current, aquatic-vertebrate-based
convention so long as once a trade name is announced, we note that
somewhere in the source code so that LXR et. al can find it.  We're all
familiar with variable name abstractions like "'x' is the valve position
before the error term is added", I don't see why we can't deal with
platform names in the same way.  Google calls it mahimahi, HTC calls it
Nexus One.

<devil's advocate>
If we do a code rename, should it be "nexusone", "nexus_one",
"nexus-one", "nexusOne", "NexusOne", or what?
</devil's advocate>

As for new platforms, the one who writes the first line of code always
gets to pick the name.  That shouldn't be news to anyone here.


b.g.

-- 
Bill Gatliff
Embedded systems training and consulting
http://billgatliff.com
bgat@billgatliff.com