Re: Should the IETF be condoning, even promoting, BOM pollution?

[Adam Roach <adam@xxxxxxxxxxx>]

On 9/18/17 2:17 PM, Carsten Bormann wrote:

On Sep 18, 2017, at 19:50, Adam Roach <adam@xxxxxxxxxxx> wrote:

 Note the HTTP server provides no content encoding headers so it's up to the app to detect.

But that is not the way HTTP works.

I'd request that you be more careful with attribution; those are Dave's words. More importantly, context matters: he was discussing a technical aspect of how he was trying to game the browsers into acting as if they were loading the files from local store when performing bulk testing across multiple operating systems.


The overarching point is that Dave performed some pretty extensive tests with a large number of applications, and found (of those tested):

1. Applications that work both with and without a BOM: a significant minority
   
2. Applications that fail regardless of whether a BOM is present: again, a significant minority
   
3. Applications that work with a BOM and fail without one: a clear majority
   
4. Applications that fail with a BOM and work without one: Zero.
   

You have some theoretical arguments here that suppose the fourth category is the most important one, while applications in the third category should be ignored. That's some pretty bizarro-world reasoning, given the collected data. To make a compelling argument, I think you would need to explain why your notion of aesthetic purity should supersede practical impact on users.

/a