Handle errors during remote repository update more gracefully

An temporary error like this fails the entire update:

  File "/home/user/vcs/repomaker/maker/tasks.py", line 51, in update_remote_repo
    remote_repo.update_index()
  File "/home/user/vcs/repomaker/maker/models/repository.py", line 372, in update_index
    self._update(repo_index, update_apps)  # also saves at the end
  File "/home/user/vcs/repomaker/maker/models/repository.py", line 410, in _update
    self._update_apps(repo_index['apps'], repo_index['packages'])
  File "/home/user/vcs/repomaker/maker/models/repository.py", line 442, in _update_apps
    changed = app_obj.update_from_json(app)  # this also saves the app_obj
  File "/home/user/vcs/repomaker/maker/models/app.py", line 258, in update_from_json
    self._update_icon(app['icon'])
  File "/home/user/vcs/repomaker/maker/models/app.py", line 273, in _update_icon
    icon, etag = net.http_get(url, self.icon_etag)
  File "/home/user/.local/lib/python3.5/site-packages/fdroidserver/net.py", line 61, in http_get
    r.raise_for_status()
  File "/home/user/.local/lib/python3.5/site-packages/requests/models.py", line 929, in raise_for_status
    raise HTTPError(http_error_msg, response=self)
requests.exceptions.HTTPError: 502 Server Error: Proxy Error for url: https://f-droid.org/repo/icons-640/com.sam.hex.16.png

Since the ETag and update timestamp have been set already, the next update doesn't continue where the last one left off.