[OSGeo-Discuss] EU Cyber Resilience Act - potential impacts on open geospatial software?
Iván Sánchez Ortega
ivan at sanchezortega.es
Wed Jul 26 08:11:23 PDT 2023
El viernes, 21 de julio de 2023 23:20:22 (CEST) Adam Steer via Discuss
escribió:
> It would be great to hear your thoughts on the impact of the proposed
> legislation
TL;DR: Not such a big deal.
At least, not a big a deal than GitHub (i.e. Microsoft) and ASF seem to be
making.
A bit of context: I'm a EU citizen, I have fair amounts of experience reading
legalese, and I've read through the proposed Act as well as the Github and ASF
statements. I also speak as a solo developer/maintainer.
It's important to note that as of today (2023-07-26), revision 454 of the
proposal seems to cover the concerns about donations and developers employed
by corps.
It's also important to remember that FOSS licenses rely and build upon
intellectual property laws; they *hack* IP laws. In the same vein, it's
possible (and, I think, desireable) to hack the EU CRA, by adding a very
simple statement such as:
«If you will not, or can not, accept liabilities for using this program, then
you may not use this program under the terms of this license. In particular,
if you will not, or can not, comply with the obligations of the EU Cyber
Resilience Act, then you may not use this program under the terms of this
license.»
There seems to be a fear that there will be undue burdens placed on individual
mainainers. I don't think that'll be ever be a problem; but if it would ever
be, then I'd like OSGeo to ask the FSF to release an updated GPL v3.1 wih a
statement I wrote above.
It's a nice hack of the law that, if you think a FLOSS developer should be
liable, then that piece of FLOSS is not FLOSS for you anymore, and you should
treat it like closed-source software and pay the developer for it (and comply
with EU CRA anyway). Magic!
---
I do have a concern, and it's about Github's position. It does read as
Microsoft FUD: «But think of the POOR DEVELOPERS!».
My experience with GitHub for the last few years is that it has become an
instrument for Microsoft to shape and control the software supply chain. Bit
by bit, the burden of maintenance has shifted from the user to the developer
and, from my PoV, they're directly responsible for creating maintainer
burnout. Now they're getting told that entities which benefit monetarily from
software should be responsible for vulns, and Microsoft doesn't like being
told that.
---
I disagree with ASF's position about the EU CRA applying to The Commons. After
all, my interpretation is that The Commons is very, very different from The
Market; and the CRA only applies to The Market.
I need to add a "citation needed" bit to this sentence in the ASF text:
> [...] the policy makers have made it crystal clear to the ASF that they
intend to have the CRA apply to open source foundations.
I'd really really like to know who made that clear to the ASF, and with what
words. Until then, my position is "The Commons is different from The Market".
If the EU CRA would really apply to FLOSS under the umbrella of a foundation
(i.e. OSGeo software), then my recommendation is an addendum to the FLOSS
license, as above. If (or when?) this becomes an issue, a FLOSS foundation can
draft updated licenses and pressure devs to use those updated licenses. It's
very much possible to shift the burden away from the developers and away from
the foundation(s) via license.
---
What's the EU CRA, anyway?
A human-readable summary of the obligations is:
- Use your brain when designing things like storing user credentials and the
like
- Try to minimize exploits (SQL injections, XSS, etc) and DDoS
- Have some way of logging stuff if it's helpful
- Have some way of offering updates
- Know your software dependencies
- Keep track of vulns (e.g. have a bug tracker)
- Have a way of contacting you
- Fix vulns ASAP
- Keep a changelog
- Keep technical documentation (i.e. a README file); write about
- what this is for,
- what dependencies it has, and
- how you've used your brain to try and avoid exploits
- Do some kind of audit and/or testing
- Report any vulns to ENISA (the EU counterpart to the US NIST)
AFAIK, OSGeo projects are doing most of this anyway, so the extra burden is
not so big - and remember, the burden applies only if you're making money with
the software.
---
Jody Garnett said:
> The economics of this are where I would like to know more. I hope we get a
Paul Ramsey keynote on this topic
The economics are going to be intesresting and, yes, please, more Paul Ramsey.
I predict that a number of bullshit jobs will be spawned around the EU CRA,
the same as happened with the EU cookie law. And instead of empowering users
and developers, the bullshitters will somehow create auditing services that
ultimately are not necessary and rely on fear and FUD. You know, much like the
freaking' obnoxious cookie popups are not needed at all *because you are
forcing an antipattern to maximize profits instead of trying to be nice to the
user*.
There's RHEL, and there will be bullshitter auditors, and there will be (I
guess) project forks just for EU CE mark. The later worries me, in terms of
actually improving upstream.
On other hand, the EU CRA doesn't apply to software in alpha (or beta, or
gamma) status, or otherwise "for testing purposes only". I predict that some
pieces of software will adopt a "all of our releases are for testing purposes
only" policy, as another workaround.
Seth G said:
> If OSGeo can find a way to capture some of this value by ensuring compliancy
and gathering funds from large organisations that use OSGeo projects, then
this could be seen as an opportunity rather than an impending disaster.
Well, then OSGeo might want to offer auditing services at better terms than
the bulshitter auditors, and establish mechanisms to put a CE mark on OSGeo
projects.
Not that OSGeo can be a target of the EU CRA, though; as ar as I'm concerned
it's a non-profit and therefore part of The Commons and not part of The
Market. It would just proxy EU CRA responsibilities from commercial users.
---
Jody Garnett said:
> I am going to stop writing
Yeah, me too.
--
Iván Sánchez Ortega <ivan at sanchezortega.es> https://ivan.sanchezortega.es
More information about the Discuss
mailing list