[rfc-i] v3imp #8 Fragment tagging on sourcecode

pkyzivat at alum.mit.edu (Paul Kyzivat) Fri, 23 January 2015 18:37 UTC

From: "pkyzivat at alum.mit.edu"
Date: Fri, 23 Jan 2015 13:37:48 -0500
Subject: [rfc-i] v3imp #8 Fragment tagging on sourcecode
In-Reply-To: <20150123181608.GJ2350@localhost>
References: <54C20F92.4090400@seantek.com> <54C232FC.1000604@gmx.de> <54C275BC.1040905@alum.mit.edu> <20150123175511.GI2350@localhost> <54C28E3F.4040901@alum.mit.edu> <20150123181608.GJ2350@localhost>
Message-ID: <54C294FC.5000204@alum.mit.edu>

On 1/23/15 1:16 PM, Nico Williams wrote:
> On Fri, Jan 23, 2015 at 01:09:03PM -0500, Paul Kyzivat wrote:
>> On 1/23/15 12:55 PM, Nico Williams wrote:
>>> Keeping the canonical module scattered is not an answer.
>>
>> Different strokes for different folks. This *can* work. It would
>
> Yes, it *can* work.  But when you make large changes you have to scatter
> them all over your I-D.  That's double-plus-unfun.
>
>> work much better if you could automatically extract the collection
>> of fragments when you want to process it.
>
> That's easy enough if you're willing to do what RFC5403 does.

Sure, but without a standard way to do it you need a custom extractor 
for every draft. In the case of ABNF, which is widely used in drafts, it 
would be highly desirable to have a single tool that could verify the 
ABNF in any draft.

Also, the tagging in 5403 diminishes the readability of the draft. 
Consider how much less readable RFC5234 would be if you did that.

	Thanks,
	Paul