[ckan-dev] Problems with CKAN 2.2.1 (possible BC break?)

Fri Oct 10 15:25:06 UTC 2014

Hi Stefan,

On 10 October 2014 13:10, Stefan Oderbolz <stefan.oderbolz at liip.ch> wrote:
> ValidationError: {u'extras': [{}, {}, {}, {}, {}, {'key': ['There is a
> schema field with the same name']}, {}, {}, {}, {}, {}]}
>
> We could trace it back to the following merge request:
> https://github.com/ckan/ckan/pull/1894

You are correct, the extras schema changed slightly as part of #1894
and the ValidationError you are getting is part of it.

> We have extra fields that are named like fields in the root schema, which
> apparently is not possible anymore. So we couldn't yet get data that was
> harvested with 2.2.1 into CKAN.
It is dangerous to have extra fields named as the same ones on the
root schema, as validation could be bypassed entirely on the root ones
when submitting extra ones. That's why we felt that it was worth
preventing it via a validator.

> It's very well possible that we don't use this as intended, but still I
> think this violates the "patch releases don't contain backwards incompatible
> changes". What do you think?
In hindsight this change might have been to disruptive and the
potential security issues didn't outweigh the backwards
incompatibilities. I'd be happy to revert it from the patch release if
others agree, but would be good to know other core devs opinions?

> The datasets that were harvested with v2.2 have now suddenly page loading
> times of  ~10s or more (before ~1s). This might be another problem (?).
I think this is completely unrelated to #1984, could you do some
profiling or investigate a bit more about which part of the code is
taking too long to run?

Thanks,

Adrià

>
> Best Regards Stefan
>
> --
> Liip AG  // Limmatstrasse 183 //  CH-8005 Zürich
> Tel +41 43 500 39 80 // GnuPG 0x7B588C67 // www.liip.ch
>
> _______________________________________________
> ckan-dev mailing list
> ckan-dev at lists.okfn.org
> https://lists.okfn.org/mailman/listinfo/ckan-dev
> Unsubscribe: https://lists.okfn.org/mailman/options/ckan-dev
>