Re: [hybi] Yet more additional WebSocket Close Error Codes

[Julian Reschke <[e-mail]>]

On 2012-05-18 23:53, Jason Duell wrote:
>>> Chrome is not calling onerror for this, so we have a difference here.
>>> The language in the spec isn't really clear if this covers the
>>> connection-never-happened case.
>>
>> I think it is;
>
> So you think the behavior for "server not available" is clear in the
> spec?  Which flavor (call onerror or not)?

I think it is clear in that it talks about a response code, so does not 
deal with the case you are referring to.

>> but I agree that the behavior for the other case should be
>> specified at all.
>
> I couldn't quite parse this either.   Is that support (or the
> opposite) for a code for "too many websockets open", or something
> else?

It's support for defining the API behavior for connection problems.

>> Anyway; I think this should be raised as a WSAPI problem.
>
> Yes, most of the wordsmithing needs to happen there  (and as soon as
> their mail server starts accepting my mail I'll have a post about this
> there).   But the list of close codes (including pseudo-codes that
> never go on the net) has so far been maintained more on this list, so
> I figured I'd ask here too.

Agreed.

Best regards, Julian