Hi Joerg,
On Wed, 2020-12-23 at 11:25 +0100, Jörg Hoh wrote:
> I would love to get some feedback on
> https://github.com/apache/sling-org-apache-sling-servlets-post/pull/7
> , so
> we can get a fix for my problem committed soon.
I had a single comment on the conceptual level, it would be good to
I would love to get some feedback on
https://github.com/apache/sling-org-apache-sling-servlets-post/pull/7, so
we can get a fix for my problem committed soon.
Thanks,
Jörg
Am Do., 17. Dez. 2020 um 12:54 Uhr schrieb Jörg Hoh :
> Thanks Bertrand, I implemented your feedback.
>
> As my initial
Thanks Bertrand, I implemented your feedback.
As my initial implementation was deemed to simple and not covering relevant
cases, I created a new PR at
https://github.com/apache/sling-org-apache-sling-servlets-post/pull/7 which
tries to return a different status code based on various JCR
Hi,
I'm coming late to this discussion, with a smallish thing..
On Wed, Nov 11, 2020 at 1:10 PM Jörg Hoh wrote:
> ...Having that in mind, I would nevertheless argue to switch the behavior of
> the SlingPostServlet to return a 405 "Method not allowed" in the case of a
> PersistenceError [2]...
Hi.
Strongly agree this should not be a 5xx response. A 4xx makes sense.
- Andrei
On Wed, 11 Nov 2020 at 14:25, Carsten Ziegeler wrote:
> Hi,
>
> agreed - in many cases we're return a 500 where a 4xx would be more
> appropriate. As you mention, the post servlet can't decide which case it
>
Hi,
agreed - in many cases we're return a 500 where a 4xx would be more
appropriate. As you mention, the post servlet can't decide which case it
is, so we have to map all exceptions either to 500 (like today) or to
4xx. In both cases, we have a false positives.
But I agree, that it's more
Hi all,
while trying to assess a number of internal server errors, I came across
the behavior that the SlingPostServlet is always returning a 500 if it has
been invoked and wasn't able to write to the repository because of missing
write permissions [2]. See
