This is basically the same problem as in 1b291415
(component/ghostscript: Workaround for slaprunner paths with double
slashs, 2021-06-02)

See merge request nexedi/slapos!1027

See merge request nexedi/slapos!931

The resiliency pull-backup-server status feed file `pbs-status-rss`
is generated by a cron running every 5 minutes. But there is a promise
that requires it to check the notifier status, meaning instanciation
fails until this file is available.

To allow the promise to succeed immediately, we generate the
feed a first time directly from buildout.

The resiliency notifier status feed file `notifier-status-rss`
is generated by a cron running every 5 minutes. But there is a
promise that requires it to check the notifier status, meaning
instanciation fails until this file is available.

To allow the promise to succeed immediately, we generate the
feed a first time directly from buildout.

See merge request nexedi/slapos!1041

we need to be explicit for extras.
Ping egg.

With multiple security fixes

See merge request nexedi/slapos!992

Follow-up for commit 98e7a29a (component/trafficserver: Version up to
9.0.2).

Drop patch from pull-request 7577, as it has since been merged upstream.

Zodburi 2.5.0 added support for demo: URI scheme:
https://github.com/Pylons/zodburi/blob/master/CHANGES.rst#250-2021-05-12

- due to a thinko in c170fe3e (ZODB: Provide ZODB4-wc2 which backports MVCC approach from ZODB5)
  4-wc2 revision was set on a ZEO(egg) section instead of ZEO(repository)
  section. As the result it was ZEO4 - not ZEO4-wc2 - that was installed.

  This was leading to hangs in tests with wcfs waiting for ZODB server to
  update it after a commit, but no notification was coming from ZEO at all.

  For the refrence: the reason ZEO4-wc2 exists at all is the backport patches
  that teach ZEO to send invalidations even on empty transactions
  (nexedi/ZEO@bf80d23d), and to include
  just-created objects into invalidations as well (nexedi/ZEO@f2fae122).
  Those changes are needed for WCFS to be always notificated on every
  transaction and with full set of changed objects.

- there was a typo: it was `revisions` instead of `revision`.

-> Fix both issues. Now it is ZEO4-wc2 that is installed for real when `ZODB.major = 4-wc2`.

The promise is commented out because it fails inside a test suite
due to port conflicts with the testnode of the test suite.

`srv/testnode/abc/inst/test0-0`  -> `t/abc/i/0`

Replace `softwaretype` recipe with `switch-softwaretype`

This is similar to d67f382f (wendelin.core: Fix test build 2)

Always reference to in-tree eggs via ${<name>:egg} instead of just name. This fixes e.g. the following build failure:

    Installing wendelin.core-python.
    While:
      Installing wendelin.core-python.
      Base installation request: 'wendelin.core[test]', 'pygolang[pyx.build]', 'neoppod[tests]', 'ZEO[test]'
      Getting distribution for 'ZEO[test]'.
    Error: Picked: ZEO = 5.2.3

Not sure why it worked before.

Neoppod is a bit special - leave it as is for now.

This teaches nxdtest to detect leaked processes and to clean them up
automatically instead of letting tests stuck because of them. See
nexedi/nxdtest@0ad45a9c and
nexedi/nxdtest!9 for details.

New dependencies of nxdtest are:

- psutil (already covered by stack/slapos.cfg),
- pytest-timeout, and
- setproctitle

/reviewed-by @jerome
/reviewed-on nexedi/slapos!1035

These comments are not updated automatically and very hard to update manually,
so it's better not to have them.

Users are often providing wrong values which are rejected later. So give the
clear information to the user.

This also adds a headless-chromium component and enables the http_sub
module in the Nginx component. Some basic tests are included. See merge
request nexedi/slapos!1014 for more details and discussion.