Hacker News new | ask | show | jobs
by haunter 2109 days ago
>the age-old argument that free software still is not accessible or easy to use

Personally I'd say most of the time the problem is the lack of documentation, help, manuals, guides etc.

I had so many times when I had a problem with a software under Linux and basically you have to spend hours in Google trying to find that stackexchange/reddit/askubuntu thread where you might find your answer, or not.

7 comments

Poor documentation practice is pervasive in the free software world. Much of the issue, I believe, is the rise of sites like Stackoverflow. Documentation these days seems to consist of the bare minimum--docs autogenerated from method/function signatures with maybe one or two trivial, useless examples--and an expectation that people will "tinker" with the software and post/find answers on these sites. Google and Stackoverflow, in other words, are the documentation of choice.

I understand: documentation is hard. It's tedious and it's difficult to get right for all the likely/major audiences involved. That's not a good excuse, though. The amount of fad-driven, cargo-cult coding and the like has just exploded in the last few years, much of it essentially copy-pasta from sites like Stackoverflow or whatever a Google search throws up.

Not in my experience. I’ve spent a lot of time reading documentation since switching to a Linux desktop about two years ago. Of course, lots of software is poorly documented, but I’ve learned far more from the systemd man pages than I ever learned from Windows knowledge base articles. And even when documentation is completely lacking (as it is for a lot of proprietary software), you can look at the source.
I think the problem is not the lack of documentation, it is that people don't know what documentation is.

The iPhone 10 or 11 or wherever we are, does...

...not ship with a manual! Have you ever noticed that? In fact, documentation is the reason why people struggle with Linux, not the lack thereof. iOS tries to make things obvious. But obvious means obvious for some people and I assume they maximise for the set of people looking for a smartphone and perhaps more money than average, with some vague I idea of what kind of UI "looks current".

In fact, there is a story about a somewhat famous programmer going on sabbatical with only an installation of one of the BSD variants and from the documentation was able to make commits to the code... Maybe someone remembers who it was? It was mentioned on HN comments.

Having said that, yes, it is difficult to read documentation! And that's why people struggle with Linux, because it is a massive time investment.

Sounds like John Carmack. I seem to recall him doing one of his famous "lock myself in a hotel room and don't come out until I've hacked something cool" things, this time with setting up and using an OpenBSD workstation.

And yes, one time I was fooling with NetBSD and went from zero to hello-world kernel device driver with just the man pages. So it's not surprising Carmack would be able to do the same with OpenBSD.

Right you are! Here is the HN link: https://news.ycombinator.com/item?id=23224584
Developer documentation and user documentation are two things. Also, programmdr commited to make commits will put way more effort into figuring things out and into experimenting then is reasonable for user who has completely different job.
> I had so many times when I had a problem with a software under Linux and basically you have to spend hours in Google trying to find that stackexchange/reddit/askubuntu thread where you might find your answer, or not.

In my experience, the only way that differs from proprietary software is that developers of free software are less likely to ban you or threaten legal action for drawing attention to said problem.

This works into my other comment about how training is provided for big box purchases. No one in any form of academia has the time or money to develop their own training materials - the tool is supposed to benefit the students, anything that isn’t directly doing that isn’t a priority.
I agree with you in principal, but offer a small counterexample:

I was a student worker at a university with a documentation / training dept. We maintained a wiki used by faculty and staff. We encouraged adoption with other groups, and built out a library of articles on tech resources, like how to use screen recording software to make and edit videos for the classroom.

We also had dedicated hours in one of the computer labs, where one of us would be present to help anyone with questions on using Blackboard, wiki, etc.

So in some cases they are trying, but to your original point, we only had 3-4 people, mostly part time trying to provide training and documentation. I think we made a difference, but there was a lot more that could be done we didn't have resources for.

Windows has that problem, but also had a millions companies trying to buy up the top search slots.
Don't tar all free software with the brush of Linux software doco, though. This is not a universal.
> Personally I'd say most of the time the problem is the lack of documentation, help, manuals, guides etc.

This is a very “blue pill” way of looking at technology. Like when a huge company releases a SaaS product, and it observed a high bounce rate from the first few days of trial usage of your new Educational Software, blah blah blah, the metrics, blah blah, improve documentation to improve the metric, because that logically seems to be related to the bounce rate, and look there’s this Net Promoter Score survey that says people didn’t use the stuff we put in there or they seem confused so of course onboarding is what we’re missing...

Linux is in disguise the most used operating system in the world, it’s the worst example to bring up because you’re talking about real pain points. But this whole idea of “pain points” and the metrics and the onboarding and documentation stuff, this is just a bunch of post hoc rationalizing business speak, a form of astrology in product development that would have been 200% wrong about Linux’s success and will continue to be wrong about all sorts of software. And then you’ll jump into pedantry, like “oh I meant specifically software under Linux not Linux itself,” or whatever.

Listen it doesn’t matter, documentation doesn’t really matter, the people writing documentation for unsuccessful or useless stuff will not make it useful by writing documentation. Documentation can’t fix what’s wrong with a piece of broken software. However it is an enteprise product, it is a piece of utterly worthless differentiation that salespeople can go out and sell and buyers can go buy, and the money changed hands, so of course everyone is out there talking about how important it is.

Don’t use software that’s hard for you to use! No one is forcing you to write systemd units, iptables commands, command like arguments for ffpmeg etc. And yet, it must piss you off so much that there are people basically selling this free software in wrappers. You’re willing to try at least, which is good, and it’s not coming from a place of valuing your time or from lack of knowledge or accessibility - it’s from the ethos of not being ripped off, of giving the people doing the valuable thing (writing the software not the docs) most of the value, and not some commercial halfwit middleman. The ethos of being revolted about giving money to some giant company, so that some VP who doesn’t do anything can chauffeur his kids to their pod in the repurposed yoga studio on 1 Main Street of his suburban home’s town.

This is what free software for education is about. When the time isn’t super important, when the money is either not spent or sucked up by cronies - certainly not spent efficiently - you want a piece of software that is compatible with the ethos of what education is really about.

Education, more than anything else, is fundamentally opposed to the enriching of middlemen. In software middlemen tend to slap documentation on free stuff.