Don't touch this! Seven types of open source to dance away from

Any recommendations based on good practice?

"Potemkin villages"

That list pretty much describes Redhat...

Overall openwashing is a good thing.

... regardless of how we may feel about the practice.

One of the major outstanding problems with todays technology is the closed-source code driving pieces from individual chips to complex mechanical systems.

Simply having the code visible publicly allows the technology using it to be maintained in the long term. Perhapse commercially from what started as openwashed code. Perhapse replaced by properly FOSS projects later down the line after the commercial support evaporates.

As our world gets ever more mechanised by the IoT trends this openwashing behaviour plays a vital role in sustainability.

Adrian

I tried to read your article but I'm afraid I can't make head or tail of it. May I respectfully ask what are your open source credentials?

>>Licensing being the thorny issue that it is, this point is probably inevitable if still

>>disappointing. Open of course means that an unlicensed version of software

>>must always exist, for free.

In the USA, as soon as you write something, including software, its copyright belongs to the author unless/until it is assigned or licensed or specifically released into the public domain. Software that has no license you can find is a problem, as rights to it belong to its source (perhaps most often the author) and don't belong to anybody else. If someone uses it without getting a license to it, she/he stands the chance the copyright holder could claim compensation for its use.

A second matter is there is no requirement that "free" software, "free" as in "freedom," be available for "free," as in "free beer." Using "free" to mean two different things is a source of confusion in some cases. Some license somewhere might prohibit charging, but the definitions published by the Free Software Foundation do not.

Avoiding projects without copyright assignment... well, a mixed blessing.

Copyright assignment discourages contributions from all but those with the most serious need to scratch their itches, because it erects a barrier - at the minimum it's extra paperwork, at the other end, you may need to get executive approval from your boss to release the copyright on "your work".

It also creates paranoia. If you assign the copyright for your work to a company, nothing stops them re-licensing that code however they like. They can take the code private and profit from your efforts. No-one likes to be taken for a mug.

The converse, projects composed of mixed copyright code find it harder to change license - which means they are more likely to remain as open as they were before.

When LibreOffice forked from OpenOffice, it took most of the core contributors with it - and ditched copyright assignment. It's by far the more vibrant project.

Hogwash

What a load of hogwash. Any GPL licensed project that is sponsored by a company is simply a way for companies to make money off of free software development. The users rights are not considered in the equation. Do you really think RedHat allows commits that would advantage a rival such as IBM or Oracle without some gain for themselves?

The GPL ensures that nobody can take the software and innovate it and gain a competitive advantage, thereby giving the sponsoring company a huge pool of development with no risk. Mr. Stallman is the only person that still believes in the socialist bent of open source. If a company is involved, the pursuit of money is the only reason it exists.

What's a rubric without subjective, ill-defined measures?

Here's my list:

1. The proof of the pudding is in the eating.

I'm interested in code that contains something useful. If I pick up an open-source project, I'd better be damned ready to support it myself, to the extent that it remains valuable to me.

That's how we did things in the PS (Pre-Stallman) era, and they worked Just Fine.

Damn it, where's jake when I need a fellow curmudgeon?

RedHat EULA

I think you will find that every Red Hat commercial distribution has a EULA preventing redistribution as is. Red Hat commercial distributions are neither free as in beer nor free as in freedom. Sure you can get the source but just try and build it without a world of pain.

Measure of 'goodness'

The trap this immediately falls into is making the assumption that being open source is some measure of the saintly measure of a project or company. You may as well check whether the owners have made charitable donations or take in sick animals (no, not developers).

The intent behind open sourcing a project, and the actual end effect doesn't sit on some single continuum between 'evil and cynical' and 'advancing the cause of mankind' any more than the actual software itself is purely saintly or nasty.

When you engage with any project, open source or otherwise, the question has to be whether doing so will meet your goals - and you must recognise that your goals and the owner's goals may be many and varied and wildly different. Not good or bad, just more or less aligned. A single corporate committer may be quite acceptable if you simply need their current release to perform a task in a nice stable environment. Equally, a project may be of no use if its' large and active community wish to introduce breaking changes or pursue new developments that don't sit with your specific use case. Some projects open source a component so small that it's useless in isolation - like a new type of bolt for building an oil rig. Others open source the entire world safe in the knowledge that the chances of you being able to replicate a functional environment is nil - like being given the plans for a steam train with a note saying that building the rail network is a task left up to the developer. None of these are necessarily signs of good or bad intent - just different ways projects may be run.

The only consistent warning sign I've come across are those projects where the owner is enthusiastic to point out that the project is good simply because it's open source. Suddenly I get the strong whiff of snake oil.

Criteria

When I start using a project, I will also simply look to see if it's been out for a while, and appears to be well developed. I look to see if the docs are good enough to figure out how to use it, and try it to see if it does what I want it to do. Some projects are pretty inactive because they are mature -- look at some of the GNU utils, they are simple, focused utilities that already do everything they claim to do, and have had the bugs worked out for years. Of course, if you are planning to use these utilities online (like to provide some service to the internet at large), you better make sure that project is active enough so if security holes pop up they are dealt with.