users@jersey.java.net

[Jersey] Re: Discussion about re-opening a bug: JERSEY-2942

From: Markus Karg <karg_at_quipsy.de>
Date: Mon, 12 Oct 2015 14:21:48 +0200

Graham,

thank you for your kind input, but see, pointing out that others did bad things in the past should not be an excuse to also do bad things.

As a community representative I really do share your view and I hate that point of the spec as you do, really. But I also have in mind all those application vendors that follow the specification as-is (as weird and as counterintuitive as it is). If we change the spec, ALL those applications will definitively fail. If we keep the spec as-is, only SOME applications will fail – particularly those not tested against the RI (hence: not developed against the spec but against a particular product that violates the spec). As the idea of an RI definitively is (among others) to point out such application design “faults”, the latter is the smaller group, unfortunately.

So I need to say that I do not see a real chance for a specification change (again, feel free to open a Jersey JIRA ticket for a product-specific change). If we love it or not, for the sake of backwards compatibility (one of Java’s holy grails) we MUST NOT touch that part of the spec.

-Markus


Von: Graham Leggett [mailto:minfrin_at_sharp.fm]
Gesendet: Montag, 12. Oktober 2015 13:56
An: users_at_jersey.java.net
Betreff: [Jersey] Re: Discussion about re-opening a bug: JERSEY-2942

On 12 Oct 2015, at 1:48 PM, Grzesiek <gregory.d3_at_gmail.com<mailto:gregory.d3_at_gmail.com>> wrote:

as no spec must break backwards behaviour ever


Current unintuitive/ strange/ buggy (choose whatever you want) behaviour in Jersey should definitely be changed.

Certainly, it must be changed - without breaking backwards compatibility.

Regards,
Graham