[Date Prev][Date Next] [Thread Prev][Thread Next] [Date Index] [Thread Index]

Re: Sponsorship requirements and copyright files



On Wed, 25 Mar 2009 15:44:20 +1100 Ben Finney <ben+debian@benfinney.id.au> 
wrote:
>Scott Kitterman <debian@kitterman.com> writes:
>
>> On Wed, 25 Mar 2009 13:22:04 +1100 Ben Finney 
<ben+debian@benfinney.id.au> 
>> wrote:
>> ...
>> >Those who don't like the very *idea* of a machine-parseable format
>> >for .debian/copyright ? apparently exist, but I don't understand
>> >their position yet :-)
>> 
>> I'd be one of those.
>
>Thank you for your explanation; after reading it, I would not actually
>classify your position as stated above :-)
>
>> Whenever you add new structural rules on a file it creates more
>> things one needs to know, more things to get wrong, and more work.
>> This is inevitable.
>
>Yes.
>
>> To counter this, I see some very minor potential benefit. IANADD, so
>> I don't get a vote, but if I did, I'd be against it.
>
>Okay, so it's not that you're against having a machine-parseable
>format for the file, but that you don't yet see that the benefit
>outweighs the cost.

Yes.  I'm against adding requirements for work with a negative value.

>> The cost/benefit ratio of the proposal is certainly open to
>> reasonably varying opinions, so I don't expect arguing over
>> different perceptions to have a lot of benefit. I do think it's
>> worth (once) pointing out why I don't like the concept.
>
>As I understand your position, it's not the concept that you don't
>like, but your perception of the cost:benefit ratio. Is that a fair
>restatement?

Yes, but given that I see the potential benefit (so far) as close to nil, 
I'm not particularly expecting this to change.  There may be some great use 
case that really suprises me, but I'm not seeing it so far.

>> I'll convert my packages when it's required by policy.
>
>Okay. Certainly I would hope there will be demonstrable (as opposed to
>the merely potential) benefits to such a format, before anyone
>considers making it mandatory.

There are also inherent negatives to adding more strctural requirments.  
One that particularly concerns me about this one is that it adds to the 
mountain of stuff new developers have to learn.  It's hard enough to teach 
people to accurately get all the currently required elements into 
debian/copyright that I don't relish adding to it "No, that can't be a 
dash, it has to be a colon" types of issues.

In Debian (and Ubuntu) it's not rare for me to see enthusiastic new 
contributors get discouraged and give up over the existing mountain of 
requirements.  I don't like adding more without a really good reason (yes, 
I know this isn't required anytime soon, but to be truly effective it will 
have to be eventually).

I hope that clarifies my perspective.

Scott K


Reply to: