TAC ranges are different in between LTE and NR.

It must be "ue", not "uesim" to match software-type in the software
release.

* master:
  software/end-to-end-testing: add test_ors.py
  software/end-to-end-testing: add test_lopcomm.py
  component/python3: version up python3.11.8
  stack/slapos: version up pyparsing (3.1.1 on py3, 2.4.7 on py2)
  stack/slapos: version up defusedxml 0.7.1
  component/pygolang: remove some duplicate versions
  *: normalize versions of pexpect
  stack/slapos: move httplib2 version pin in slapos.cfg and update to 0.22.0
  component/ZODB: update ZODB5 5.8.1
  software/gitlab: run git-backup test during software step
  software/gitlab: version up gilab-ce
  software/gitlab: remove outdated comment

Original SR had defaults as

mme.gtp: 127.0.1.100
enb.gtp: 127.0.1.1

With this both can listen at the same port and talk to each other.

But if we set gtp addr to be the same for mme and enb, then enb listen
will fail as the listening address is already taken by mme.

Fix it with restoring explicit enb gtp loopback default.

rx_delay is meaningful to have as the default, as that is RU-model
specific and goes as T2a+T3a-Toffset.

tx_delay is less meaningful to have as the default: it is T12 + T2a,
where T2a is RU-model specific and T12 depends on the length of the
fiber link. But I keep the previous default because it might be handy to
have something as the default.

Also restore previous values for sunwave n_antenna_* as it was there in
old ors-amarisoft SR.

The code already behaved in the "updated" way with

    slaplte.jinja2:{%-     do cell.setdefault('root_sequence_index', 1 + 203*(cell.cell_type == 'lte') + n)  %}

See merge request nexedi/slapos!1531

See merge request nexedi/slapos!1529

These are already defined in stack/slapos.cfg

jupyter-py2 used an old version not py3 compatible
pygolang/test used the same version as in stack/slapos.cfg