[c-nsp] Output drops on 2960

[Saku Ytti]

On 10 February 2016 at 09:45, Antoine Monnier
<mrantoinemonnier at gmail.com> wrote:

Hey,

> I am wondering what is the driver for this
> running-for-latest-feature-while-code-is-unusable approach?
> surely customers are not asking for that, are they? should customer
> requirements not be the main drivers?

At least CSCO and JNPR are very focused on immediate returns. I
suspect that very small number of companies buy significant percentage
of the units.
Often when I read release notes I see feature and I just cannot see
many companies needing it, and probably they don't, but many units
might.

This means that these companies get to decide where programmers time
goes. This approach may not be that different how other companies work
in other industries, and it may be hard to criticise it, what is
better approach?
I naively think, that if I'd be vendor, I'd have much more formal
process on receiving enhancement requests from my customers. Like
JTAC/CTAC cases, but formal cases for submitting feature request. I'd
even consider making them public, so customers can vote up/down on
them. Then I could keep my eye on features that are most requested.
But how do you justify this system, if you can already fully book
developers time from customers saying 'we'll buy 5000 boxes if you
support X in next 6 months', why should you spend time on anything
else?

Some of the most vocal anti-csco comments here are made by people who
consistently and heavily invest in proprietary csco-technologies, so
from vendor's POV, everything is going great, accountants don't put
nasty words on balance sheets.

-- 
  ++ytti