Journal

Free software at work: use it, and fund it

When a free tool does the job, I prefer it. I have set some up in-house and for clients: Mattermost for team messaging, Nextcloud for files, Jitsi Meet for video calls. This note s...

Article3 tagged threads

When a free tool does the job, I prefer it. I have set some up in-house and for clients: Mattermost for team messaging, Nextcloud for files, Jitsi Meet for video calls.

This note says where the habit comes from, what really slows things down, and what I think is still missing: paying for these tools.

Where it comes from

I worked for Aremedia, a public-health nonprofit. Field teams collected sensitive data there, and someone had to know where that data went. A free tool was the only one that left that decision to the nonprofit. The case is told in A self-hosted survey tool for a public-health nonprofit.

I kept one simple idea from it: choosing a free tool is not first a matter of principle. It is a way to stay in charge of your data, your budget and your way out.

Three tools, three needs

Tool What it is for What it replaces
Mattermost Team messaging Slack
Nextcloud Shared files Google Drive
Jitsi Meet Video calls An occasional need for video, for some

I did not roll them out everywhere at once. Each one answered a precise need: replacing a service already in place, or covering a use nobody had a tool for yet.

What slows things down

Installing is not the long part. Adoption is.

As long as part of the team stays on the old tool, the switch has not happened. Conversations are in two places, so are the files, and two systems are kept running instead of one. An adoption lag therefore delays everything else: the old service can only be stopped the day nobody depends on it.

What is missing: paying

In the cases I have known, the company paid nothing to the projects it used every day.

I think that is a mistake, and an easy one to fix. A company whose daily work rests on free software has suppliers it does not pay. Funding them is not a gesture: it is maintaining what the work depends on.

What I would propose:

  1. List the free software the business really depends on.
  2. Set aside a yearly budget, for example a share of what is not spent on licences.
  3. Choose the form: a recurring donation, a support contract with the publisher, or contribution time.
  4. Write it into the company's CSR policy when it has one, with the amounts.

The fourth point matters as much as the others. A line in a CSR report is followed from one year to the next. A good intention is not.

What I take from it

  • A free tool is chosen for a need, not for a conviction.
  • The real date of a change of tool is the day the old one stops.
  • Using free software without funding it is counting on others for it to still exist in five years.