[Ardour-Users] new bug/feature tracker features

Paul Davis paul at linuxaudiosystems.com
Sat Jul 12 11:05:35 PDT 2008


On Sat, 2008-07-12 at 18:46 +0000, John Emmas wrote:
> Hi Paul,
> 
> One drawback in using the existing mantis DB is that it doesn't
> (currently) distinguish between feature requests and feature proposals.

Feature proposal => cost != 0.00

(this works unless you're also interested in FP's that have zero cost
attached to them)

Feature requests => yes, good point.

To be honest, I should not have been so adamant about no-separate-DB. I
haven't really ruled it out yet. Finding the line between "feature
request" and "bug report" can be hard sometimes. "It claims to do what I
want, but does it wrong" is not so different to many users from "It
doesn't do what I want".

> Therefore any dev who wants to set a non-zero value for a particular feature
> would need to be very conscientious if he wanted to track down all the
> relevant requests.  I can definitely envisage this being very offputting (in
> fact there aren't any non-zero cost fields at present, as far as I can see).

Correct.

> However you implement it, my instinct is that there needs to be some central
> location where Ardour users (and devs) can easily view the current proposals
> without needing to trawl through dozens (possibly even hundreds) of feature
> requests.

I would plan to add such a thing. The next several weeks will see me
more focused on this sort of thing.

> Another suggestion I'd like to offer is that any developer who puts forward
> a proposal should be able to set a minimum sponsorship amount as well as a
> cost.  For example, if 1,000 users all pledged a small amount (say $5) by
> the time Paypal has taken its cut and the developer's bank have had their
> cut, the poor old developer might be left with a lot less than he'd
> expected.  Apart from that, if a feature's realistically worth $40, no-one's
> going to pledge $40 if everyone else is only pledging $5.

yes, the way a feature priced at $XX gets paid for is hard. the whole
thing is hard. a feature may realistically need $5000 of work. does that
mean that its reasonable to ask $5k for it? probably not. some of the
cost should be borne by future users. how much? is $3k too much? $2k?
$1k?

this completely ignores the hard issues about how to price things in a
global context. when i work hard for a month i expect/need to make an
amount that people in other parts of the world might expect/need to make
in a year. globalization is a bitch and a giver, and probably more than
that.

--p





More information about the Ardour-Users mailing list