Re: Advice Workplace that Forces Non-Free Software

[fischersfritz]

Dear Crista,

Let us separate two concerns: Your own freedom and your support
of others' freedom.

I suspect you can largely maintain your own software freedom at work.
I personally don't mind using proprietary software for work, provided
that I don't make a remarkable effort to learn the software, that it
run on a dedicated computer and, of course, that I get paid enough.

As for contributions to other peoples' freedoms, I think it would be
very helpful if you simply talked about freedom to user useRs outside
of work. At present I tend to be the only one in any large room of useRs
to voice such concerns.

Also, if you come up with an interesting software idea, I suspect it
will be more useful to release it to the public rather than to work.
If it is going to be very useful, it will likely still be useful in many
years, so you can wait to create it until you retire or get fired.



I think the choice of option 1 versus option 2 has little direct impact
on your freedom and on the freedoms of others, as it seems you will have
little effect on your workplace regardless of the option. This is partly
because it seems you have little power in your hierarchical workplace.

I suggest going with option 1 because I think it will make your work
easier. In order to maintain your own freedom, instead of convincing
others to use free software, simply use it yourself. You can't avoid
running the proprietary software, but you can avoid learning it,
and you can also avoid using any work resources for non-work activity.

Also, part of your job is necessarily to read and understand the
licenses of the software you are using. Since proprietary licensing is
much varied compared to free software licensing, this can take a long
time, and you should expect your employer to give you that time, even
if it delays whatever project you are going to use the software for.

As for your specific software issues,

  Would it be okay to produce research posters in Beamer, convert them
  with Inkscape to a high-resolution raster image, and then copy that
  to PowerPoint? Or maybe you can use a vector format, so it would look
  like you really made it in PowerPoint. It won't be okay if you ask for
  permission, of course, but if you didn't, would anyone notice?

  You can write up research reports in LaTeX but render the result
  to Word. Again, don't say you're doing this, but would it work?

  Use git but don't tell anyone you're using it. Nobody needs to know.
  I think you can install git without administrator rights on windows,
  and it comes with bash. You certainly can install fossil, since fossil
  is a single executable.
  
Since your use of free software probably makes you especially
productive, I suggest additionally that you not tell anyone at work
about how you use free software at work to make your job easier.
It is important also that you pretend your work takes you almost as
long as it would take anyone else. I find that non-hackers appreciate
hard work, so it is better if your work seems difficult.



In case you do choose to go with option 2, I suggest asking on r-help in
some other R forum, as this is a common concern among useRs.

Note that many useRs these days are oblivious to the software freedom
considerations and to R's status as a GNU project, so many are likely
not to understand some of your concerns. I was very frustrated that
UseR! 2020 was announced by email only after the submission deadline.
If anyone submitted, I have to imagine it was announced by other means
beforehand.

I suggested r-help because I find that those useRs who care about
freedom tend to prefer email and official R foundation venues over other
R discussion forums. You might try asking rOpenSci collaborators, but
many of them eschew freedom in favor of supposed benefits of GitHub,
RStudio, and Twitter.

Since the RStudio IDE is free software, I elaborate on my concerns about
RStudio. Indeed, some of RStudio's products are proprietary, but the
bigger issue for me is the expectation that its users and contributors
run other proprietary software. For example, patches are often accepted
by GitHub only. Also, while the licensing of contributed packages is
generally free, lock-in is often created by the complexity of
dependencies, which I tend to have trouble installing on free operating
systems. I imagine they work just fine on Mac and Ubuntu.



I apply option 1 in the workplace. I maintain my own freedoms by keeping
work and life very separate.

  I have a very proprietary computer provided by the company, but I use
  it and only it for work.

  If I am to do anything complicated, I prefer to do it in free
  software, because I don't want to waste time learning proprietary
  software. It happens that I generally already know how to do whatever
  I want to do with free software; because it is free, I have been
  studying it most of my life, so I already know it very well.

  When I was requested to install apps on my phone, I first responded
  in confusion because I had not been issued a phone. Then when
  I learned that I was expected to have my own mobile phone, I explained
  that my mobile phone didn't have apps, so I received a hardware
  two-factor authentication token.

  Lately there have been instructions to join Gloom meetings. I join
  these by phone, without creating an account. I was recently instructed
  to install the Klaxoon "app" but was not explicitly told to do it on a
  non-work computer, so I initiated a process to get it installed on a
  work computer. This involved requesting an exception to the firewall.
  I am intrigued that the company considers these softwares unsafe
  to run on company computers but acceptable to run on other computers.
  When I find a convenient opportunity, I will bring up the concerns
  that our Chief GNUsance has been preparing.
  http://www.stallman.org/zoom.html

There are some software choices as well.

  I expect information to be exchanged in Microsoft file formats, but
  I mostly modify them with free software.

  My colleagues mostly don't know about version control. If they know
  about it, they know only GitHub, not git, and they don't really know
  how to use it. So I don't tell my colleagues that I use version
  control. I personally use fossil, but git's good too.

  I eventually did get administrator rights to my work computer,
  but before that, I noticed that some things were configured better
  on RStudio than on normal R, particularly the packages that are
  developed by RStudio; they don't seem to care to support non-RStudio
  editors. Anyway, the point is that I previously did not manage
  to install LaTex, so I made do with markdown, and the Word output
  format was working in RStudio, so I copied the configuration from
  there. Now that I have administrator rights, I can install Cygwin,
  which makes things sooooooooooo much easier and also faster. If I
  think I will have this job for a while, maybe I will try setting up
  a free operating system in a virtual machine.

All of this is may seem like lot of pointless trouble, and that is
because it is work. If it were easy, people would feel like they did
not deserve their compensation. I personally would rather not work,
but I still put up with it in exchange for money.



I do something sort of like option 2 as well. When I am around other
useRs, I agitate aggressively for free software. I don't meet other
useRs at work.

Around other useRs I also agitate about my view that the contemporary
ubiquitous interest in statistics is in fact a religion that worships
novel statistics as a diety that will save humanity. I believe that
proprietary licensing contributes to the sensation of novelty that is
part of the faith.

In my current work, for example, we overfit models in order to prove
relationships random noise, and my employer considers the resulting
fitted models to be proprietary information. The norm is to have faith
in data is seen as the source of all truth, and this exercise is
a ritual that blesses our particular project. In some cases it is
important that the relationship match other beliefs, but usually all
that matters is that data be involved somehow.

I believe that the principle of freedom of religion gives us the choice
whether to believe in this diety of proprietary data stored on other
people's computers. However, I consider that the norms are tending
to the removal of this freedom, to a situation where one must accept
the faith in nebulous data in order to be employed.

Competent useRs tend to like hearing my view of statistics as a new
religion. Maybe I should relate that to software licensing in order
to promote awareness of the benefits of free software.



Anyway, this is getting me thinking, maybe it would be interesting
to prepare materials for useRs on the topic of software freedom.
While it seems hoarders have managed to promote proprietary software
among useRs over the past few years, I believe there is still much
respect of free software among useRs because of its R's academic
tradition. I think this is especially strong among those who began
using R before RStudio.

That is to say, rather than trying to promote free software in your
workplace, I think it could be more interesting to promote free software
among useRs.



I respectfully remain your servant,
Fischers Fritz