User Tools

Site Tools



NavierStokes is a set of Python classes that allow you to bridge between social network accounts. These classes rely on a number of external tools to do the hard work of actually talking to networks. There is a master “executable” file that uses the supporting classes to bridge between your various accounts. See usage information below.

It employs “fuzzy text matching”, as well as a record of posts that have already been shared between networks, to try to prevent a post from being shared more than once to other networks (or back to the originating network). Fuzzy text matching is needed because different social networks encode or format the same information slightly differently. For instance, a post in HTML on will not be formatted in HTML on Twitter, and Twitter will shorten links, thus altering the text of the original post. Fuzzy text matching uses statistical methods to attempt to compute the probability that the message has already been shared on a network. Above a match threshold, the post will not be shared.

NavierStokes considers only posts made in the last hour when it scans the streams from different social networks. It won't re-post something that is 6 hours or 6 days old.


Copyright 2014, Stephen Jacob Sekula

Licensed under the Apache License, Version 2.0 (the “License”); you may not use this file except in compliance with the License. You may obtain a copy of the License at

Unless required by applicable law or agreed to in writing, software distributed under the License is distributed on an “AS IS” BASIS, WITHOUT WARRANTIES OR CONDITIONS OF ANY KIND, either express or implied. See the License for the specific language governing permissions and limitations under the License.

Download Latest Release

git clone
git checkout v1.0.1

Release notes:

    • Images posted to now have the To field correctly set
  • General exceptions handling and CURL usage improvements
  • Email any bugs you find to

Browse the code:


In general, here are the Python libraries needed to make this package operate:



  1. Check out the code from the git repository above. This automatically creates the navierstokes/ directory.
  2. You should create a configuration file for NavierStokes: ~/.navierstokes/navierstokes.cfg. The syntax for the file is explained below.
  3. If you intend to bridge between and other networks, you must use PyPump to register a client (e.g. NavierStokesApp) on Follow their instructions for getting the client credentials and tokens. Enter those into the

    file (see example below). PyPump docs:

  4. Make sure Cliaspora is available in the PATH environment variable, or Diaspora interactions will fail.

Example navierstokes.cfg file

Be sure to make this .cfg file only readable by you:

chmod 600 ~/.navierstokes/navierstokes.cfg
type: gnusocial
site: https://gnusocial.server.url
username: myname
sharelevel: Public

webfinger: user@pump.server
client_credentials: XXXXXXXXXXXXXXXXXXXXxx
sharelevel: All
noshare_keyword: noshare

type: diaspora
webfinger: user@diaspora.server
aspect: public
sharelevel: All
noshare_keyword: diaspiverse

type: twitter
sharelevel: Public
shortenurls: True

[my blog rss feed]
type: rss
max_message_age: 84600

service: shortenizer
serviceKey: pseudosecretKey
#service: ur1
#serviceKey: False

Note that “sharelevel” means at what level of publicity from other networks you want a notice shared to this one. I've set this, for now, the way I like it. If you set this to “Public”, ONLY notices that are public on other networks will go there. For instance, I only like to share things that are public on with Twitter. Things on Twitter are public by default, so they will ALWAYS be shared with other networks.

“shortenurls” presently enabled will take ALL URLs listed in the message text and shorten them via In the future, this will be a choice the user can make.

“max_message_age” defines the maximum age, in seconds, that a post can be before it is no longer considered for sharing. This is best used for RSS feeds where articles added to the feed may be older than the time at which they are added to the feed (e.g. sharing news articles in a reading list). This defaults to 3600.

The URL shortening section allows you to define a custom URL shortener. If unspecified, the default (presently, is used.

Running NavierStokes

Once you have written a .cfg file and setup account information in it (and, in the case of and Twitter, authenticated PyPump and t against those respective networks as clients), you can try executing NavierStokes manually:

python ./

If you get errors, try running in Debug Mode and see what you can learn:

python ./ -d

I run NavierStokes every 5 minutes using a CRON job:

*/5 * * * * bash -l -c 'python /path/to/navierstokes/ >> ${HOME}/.navierstokes/navierstokes.log 2>&1'

If you get any errors that are unrelated to passwords, logging into, report them to

A simple program to authenticate PyPump against your instance

For PyPump v0.5, this ought to work:

#!/usr/bin/env python

from pypump import PyPump, Client

client = Client(
    type="native", # Can be "native" or "web"

def simple_verifier(url):
    print 'Go to: ' + url
    return raw_input('Verifier: ') # they will get a code back

pump = PyPump(client=client, verifier_callback=simple_verifier)
client_credentials = pump.get_registration() # will return [<key>, <secret>, <expirey>]
client_tokens = pump.get_token() # [<token>, <secret>]

print "client_credentials: %s,%s" % (client_credentials[0],client_credentials[1])
print "client_tokens: %s,%s" % (client_tokens[0],client_tokens[1])

Bugs and Features


Since v0.5.0, it's become clear to me what the roadmap for this software is going to be.

  • Releases until we hit 0.9.0 are all to be considered “alpha” - try at your own risk, don't expect it to work, and please, please, please report bugs
  • Releases in v0.9.X will be the “beta” - considered to have been exercised enough and by enough people, with features added and bugs squashed as a result. The final feature set needed to bridge your posts from one network to another will be locked in during this period.
  • Releases in v1.X.X will be the stable versions resulting from the beta period
  • The goal for v2.X.X is a bit ambitious - maybe even unnecessary - but that would be to bridge conversations that result from your posts shared across networks. At the very least, people should know that an interesting conversation is resulting on another network.

I would be happy to just coast on v1.X.X until the federated networks implement APIs that allow automatic bridging (e.g. if GNU Social implements the API, it likely won't be necessary to bridge anymore with NavierStokes).

Other Information

navierstokes.txt · Last modified: 2016/04/16 11:53 by sekula