Allow origin is now to the http_origin because origin "*" will fail if
withCredentials is set in the request header

Revert previous commit as OpenSSL 1.0/1.1 are already used here and there. Instead make OpenSSH uses OpenSSL 1.0.

This reverts commit 6ca7789a.

Revert "component/openssl: Version up 1.1.1": OpenSSH 7.4p1 does not support OpenSSL 1.1 (FTBFS on erp5testnode).

This reverts commit 14d98aee.

The download URL for the .war was the same for all versions, being
updated evertime there's a new version.

Now plantuml team changed their release process to distribute each
release under a different URL, so we can just use the new URL.

https://github.com/plantuml/plantuml/issues/153

Since default openssl component is 1.1 and is not compatible with this
component, use openssl-1.0.

Since default openssl component is 1.1 and is not compatible with this
component, use openssl-1.0.

Since default openssl component is 1.1 and is not compatible with this
component, use openssl-1.0.

Since default openssl component is 1.1 and is not compatible with this
component, use openssl-1.0.

Since default openssl component is 1.1 and is not compatible with this
component, use openssl-1.0.

Since default openssl component is 1.1 and is not compatible with this
component, use openssl-1.0.

Since default openssl component is 1.1 and is not compatible with this
component, use openssl-1.0.

Keep old version in openssl-1.0 for not yet compatible software.

Depend on caucase, as it shall be installed before as it is dependency.

It does not require reference, as it works on locally provided parameter_dict

  The version 1.4.12 was broken so it required a re-release

  Traceback of the tests:

    Installing eggs.
While:
  Installing eggs.
  Getting distribution for 'urllib3>=1.18'.
Error: Picked: urllib3 = 1.24.1

  Using help script

  This change is to follow up recent refactorization and perl and CPAN Setup.

  This is to comply with obs package build.

  This is a workarround, perhaps a cleaner solution can be consider

  As PERL5LIB is defined based on destdir, the perl binaries during the build are unusable without it.

  This patch can be reverted later, if the shared build get fixed for this point

  This is required when prefix/location is customize during the package generation

Add msgpack-numpy egg so we can serialize binary numpy arrays. msgpack-python change its name to just msgpack.

Wendelin tests passing at

https://nexedi.erp5.net/test_result_module/20181126-7D4D3F45/view

@rafael  and / or @alain.takoudjou  - please merge in unless objections.

/reviewed-on !451

/reviewed-on !450

Always running SQL_CHANGE_MASTER breaks GTID setup (at least on 10.1.24)
as it resets MASTER_USE_GTID. So only run it when not using GTID.
Also, fold SQL_SET_GTID into the existing "if" bloc for simplicity.

Update jsonschema eggs and some dependent eggs (including setuptools - but it looks OK) so that we can validate draft7 of json schema.

Because jsonschema now bundles some schemas for json schemas, we don't need to include it here.

/reviewed-on !425

Also fix JSON schema, that was just invalid.

Now we have a schema that renders properly on
https://lab.nexedi.com/bk/rjs_json_form and is similar to the one from
erp5, so maybe one day we can have a convention on erp5 testnode to push
the URL of a selenium server as test-runner.server-url . In any case,
it's good for now to have consistency.

Most of this work comes from Boris Kocherov from https://www.raskon.org/
Thanks !

string.format must be a string according to
https://json-schema.org/understanding-json-schema/reference/string.html#format

We were using our own version of metaschema with a patch for this. Now
we are back to using official schema, but with a custom format
`internet-address` that should be equivalent to:

    "oneOf": [
      {
	"format": "hostname"
      },
      {
	"format": "ipv4"
      },
      {
	"format": "ipv6"
      }
    ],

The schema was not matching implementation (it was not even correct
schema)

Using
https://python-jsonschema.readthedocs.io/en/latest/validate/#versioned-validators
instead of a local copy of json schema

( not needed for the current work on jsonschema draft 7 support, but
done at the same time, to merge together two unrelated changes which
cause recompilation so that we recompile only once )