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

Re: What’s the use for Standards-Version?

On Thu, Aug 13 2009, Charles Plessy wrote:

> Le Thu, Aug 13, 2009 at 02:31:53PM -0500, Manoj Srivastava a écrit :
>>         Hmm. As my memory serves me, the statement was:
>> Message-Id: <[🔎] 1250148888.23574.10.camel@shizuru>
>> >>>> Fine. It's been several years now that managing the standards
>> >>>> version has been an unconditional bump for me.
>>         Unconditional bump, I parsed, as merely bumping the version,
>>  without regards to actual status (since otherwise it would have been
>>  conditional on checking and making any relevant changes as required)
> I am really, really, really tired to see long threads that start
> because one interpretes the words of another on a worst-case basis and
> gets offended by his own catastrophe scenario.

        Firstly, all my emails were predicated on the premise that anyone
 who is not checking their packages for policy conformance and
 merely updating the S-V filed is not performing their dutes. I never
 directed that action to any individual, so you are against removing DD
 rights even if a DD is not performihng their duties? I do not thik I
 agree with that.

        For the record:
Message-ID: <[🔎] 87skfw2rdd.fsf@anzu.internal.golden-gryphon.com>
>>>>> If there are people not managing their Standards Versions fields
>>>>> like they are supposed to, and are too lazy or incompetent to keep
>>>>> track of a simple version, I suggest we start thinking about removing
>>>>> DD status from people based on a track record of incompetence.

Message-ID: <[🔎] 87ocqj3jmj.fsf@anzu.internal.golden-gryphon.com>
>>>>>        Lying to lintian and the rest of the world is the right
>>>>> behaviour? Since when?

>>>>>        The field is supposed to contain the version of the policy that
>>>>> was last checked for conformance by the developer. Checking policy
>>>>> changes to ensure your package conforms is oneof hte duties of a debian
>>>>> developer. If you want to not perform your duty, and wait until other
>>>>> people actually dosciver policy non-compliance and file bugs, at least
>>>>> one could be honest, not change the standards version, and add a
>>>>> lintian override so that there are no annoying warnings, and yet
>>>>> someone else looking at the package has a clearer idea when
>>>>> themaintainer performed the policy check, _that_ would be the right
>>>>> thing to dso.

Message-ID: <[🔎] 87k5173jdj.fsf@anzu.internal.golden-gryphon.com>
>>>>>  Declaring one is not going to "spend effort to keep packages in
>>>>> conformance to policy"  should help tip the balance too. I suspect that
>>>>> any such decision should take into consideration multiple factors, and
>>>>> it should probably not be defined by any single one, and that there
>>>>> should be warnings, etc before action is taken, and so on.

> It has been repeatedly said that the Policy is changed when most of the
> packages already conform to it. I prefer to interpret the above citation

        That is incorrect. That is a quote often misattributed to me,
 and it not what policy is. The genesis of that misinterpretation is
 that policy is not usually where one does design, and policy prefers to
 have concrete, deployed, and tested schemas before they are written
 into policy.

        But, fundamentally, policy is all about documenting the right
 thing to do, and to select one of several technically viable options
 when one needs to for integration.

        By no means is policy meant to document suboptimal practices, no
 matter how popular they are.

> differently: that Josselin is proactively making his packages up to
> date, and that when a new Policy is published there is nothing to
> change in the package.

        Looking at the whole set of email, I do not see how one can
 reach that conclusion. I do not think my thinking that packages had S-V
 updated without checks was a misinterpretation. There is more than one
 email involved in that thread, and I think if one looks through the
 whole discussion, it is pretty obvious what the meaning was.

        If you wish, you may ask me off line, and I'll detail what leads
 me to this conclusion.

        I stand by my conclusion.

> If you had any doubt, you could have asked him in private, or in

        I had no doubt, no.

> public if you had enough confidence that it would not have added noise
> with a conversation like ‘Are you sure you do your work correctly?
> Yes, I am sure, thank you’.

> This would have saved us your digressions about lies and
> expulsions. Please consider that this list is high traffic and try to
> limit your messages in quantity.

        You are assuming I do not consider the emails to be relevant,
 and my proposal merely sent out of some petty emotion: it was not. It
 was real, and I do think people who are not checking their package for
 policy conformance and otherwise neglecting their duties as a DD (like
 not tending to their packages or uploading new version or fixing bugs or
 voting for a couple of years should also have their DD rights removed.

        This is not a digression.  You do have a point about this thread
 probably will now be a better fit for -project, rather than -devel.

        Further discussion should probably be moved to that mailing

The Public is merely a multiplied "me." Mark Twain
Manoj Srivastava <srivasta@debian.org> <http://www.debian.org/~srivasta/>  
1024D/BF24424C print 4966 F272 D093 B493 410B  924B 21BA DABB BF24 424C

Reply to: