Maverynthia

Local F.O.E. SPROUT in your dungeon

A tiny Brussels Sprout


🌱Labyrinth of Blue
maverynthia.com/

arborelia
@arborelia

The Software Heritage Archive is a project dedicated to creating a historical record of "important" pieces of software.

I made a piece of software called ftfy that fixes mojibake (Unicode mistakes) in text. It's in use in a lot of places, and I found out it was in the Software Heritage Archive. Then, I found out that nearly everywhere you find it in the archive, it's credited to my deadname. I only thought to look for this because I had heard of another trans woman's struggle trying to update her name in the archive.

I already fixed my name in my code. I updated the README and the copyright notice, and I ran git-filter-repo to rewrite the git history so it had always said my correct name, including in commits. This is a thing you can do.

I contacted the Software Heritage Archive (SWH) and asked them to change it. This started a process of over two years of me trying everything I can to get them to update my name and stop deadnaming me. They've said many different things on the topic, but the answer remains: no, they won't.

The reason is because they're a bunch of privileged cis people who have never comprehended the idea of data hurting somebody, but the other reason is a fucking blockchain.


You want to change your name? Blockchain says no

When I first contacted them, politely, they didn't even respond to me. What happened instead was that their content policy changed from a standard EU content policy to one that is very defensive about their ability to archive personal data forever with no exceptions.

A list of numbered points about their compliance with GDPR (RGPD in France, where they are located) suddenly grew this very very long 9th point, an unhinged, transphobic ramble about how personal data like names cannot be modified. Feel free to skim it.

9. You have the right to access your data. In accordance with article 21.6 of the RGPD, the right to object is not applicable, as the legal basis of Software Heritage is the execution of a mission of public interest, the long-term preservation of software source codes, carried out by Inria which is a public entity. Since one of the purposes of the data processing is scientific research, it is not possible to respond favorably to requests for deletion, rectification and limitation of processing for data identifying the contributors to the development of the source codes stored in the archive. Indeed, the identity of the contributor (first name, last name or pseudonym) is an integral part of the metadata which is essential to trace a software development, and is used in the calculation of the cryptographic key which allows both to link the various versions between them and to guarantee the integrity of these versions. Any modification would have repercussions on the whole chain and would invalidate all the keys of all the successive versions of the development, compromising the integrity of the archived contents, which are collected from public sources. The personal data contained in the development history cannot be modified, contrary to the personal data associated with the user identifiers specific to certain collaborative development platforms, which can be modified by the user. If you have any questions about the processing of your data in this scheme, you can contact the DPO

Did you catch that? The reason they can't just change references to your deadname, like any decent person would, is vaguely because of "integrity" (the common refrain of transphobic cowards), but more specifically because they made a fucking immutable blockchain out of it.

I should be clear: I'm not accusing them of running a cryptocurrency scheme. This is solely the more generalized form of "blockchain", where you create a big unchangeable append-only data structure for a purpose other than financial fraud. Which is what they did with people's code, and data about them such as their names.

Hold on, is it really fair to call them transphobic? They just want to, uh, deadname you for all of history

This would be some great dark humor, except that I've heard it too many times before, so it's not funny anymore.

The part where they try to explain to me that they're right and I'm wrong

I was on their content policy page because I was trying to figure out what to do about the fact that their "Data Protection Officer" straight up ignored my first e-mail, and I saw the transphobic rant. You can guess what happened next: I fucking exploded at them on Twitter.

This got a faster reaction than anything else I'd seen in the process. The rant against changing names in the content policy disappeared again. One of the developers went to my DMs to tell me condescendingly that I'd understood it wrong. It wasn't a blockchain, he explained, it's just a Merkle tree, like Git, because we copied your Git repository.

I screenshotted their documentation that said "blockchain". I pointed out that I understood their documentation, and that they weren't just copying repositories, they were making their own chain where they would assign their own immutable cryptographic IDs to people's code.

He said, but we have to do that, what if someone wanted to cite your code in a paper? You can't just change a citation.

I pointed out that you can just change a citation if it's wrong, and that I've gotten citations that deadname me changed. I pointed out that people cite my code under my correct name through Zenodo, not through whatever cryptographic bullshit SWH was dreaming up.

He said, well anyway, I don't see why this is our problem. You can't even change your name in your own repository! Imagine if people could just change history ha ha

I pointed him at my Git repository that I had already mutated using git-filter-repo.

He stopped replying.

What came next

  • I made a GDPR demand against them, because they're in France.

  • Their Data Protection Officer, Anne Combe, finally showed up to the job, mostly to lie to me and refuse to acknowledge my GDPR requests.

  • Several months later, they put a name change policy in the spot where the transphobic rant had been in their content policy.

Oh, so everything's okay now, right? They gave in and made a policy that lets people change names?

No. It's fake. You can try to follow their policy, but you just go in some sort of trans people database and they don't change the name they visibly call you on the website. Did you know transphobes can lie?

That's enough for part 1 and I've barely covered the last two years. In part 2, I go from merely being angry, to being legally angry in French.


You must log in to comment.

in reply to @arborelia's post:

why. why the fuck did they make it a blockchain?? if you despise trans people want immutable data that badly just don't allow mutations on your database tables?? the fuck???????

One of the more agonizing things at my last job was how frequently cis people would just say shit like "people's names shouldn't ever change, and we can make a case by case exception if necessary due to marriage"

There is some guy that won a lawsuit against his bank because they refused to update his name in their system, claiming their systen was too old to support diatrics. Maybe involving a lawyer might help things moving along?

It's so easy to say "involve a lawyer" but how? Who? I am not actually rich enough to have a personal lawyer on retainer, and I expect lawyers capable of bringing suit in France are very hard to find and expensive

If you are not in france, it might be hard yeah :( but I don't think every single lawyer is expensive. They do work on divorces and landlord being asshole for exemple. In this case, I was hoping that simply having a lawyer sending a letter/email would be enough to make things move in the right direction.

as a deeply GDPRpilled person with genders this shit makes me especially incandescent, jesu christi. the weakest link of the regulation imo is also one of its most admirable elements, that it enshrines good faith and...... lots of people and organisations don't fucking care! i'm more used to seeing it from big corpos and less from archiving projects but i shouldn't really be surprised. i hope you get them, in french.

I'm so sorry to hear this. I actually recently started following you here because of ftfy, and I was so delighted that a trans woman had fixed my problems for me that I wanted to show my support :)

Sorry to hear it's been causing you grief...

God damn. ftfy is an amazing fucking piece of software. I've used it myself before and it's one of the rare pieces of software that simply does exactly the thing it says, every single time.

I really, really hope part 2 is a change in fortunes for the better.

I'd be interested in hearing more about this.

It doesn't apply to me, to be clear -- the software license I used is permissive. It is not what prevents them from copying my code, or copying other people's copies of it. Software licenses aren't the only consideration here, though.

Yeah most of my stuff is permissive too, but I also just have what one might call random crap out there too, with no license posted. Stuff I'd be happy to license if someone asked. So while taking a look at their project, I searched my own repos there and found it all copied wholesale. I guess that's what kids would call doing a piracy, unless GitHub has partnered with them in a way that their ToU now includes a rights grant to this project. Speaking of partners, I also see Hugging Face so being cavalier about scraping and mirroring and human dignity sounds about right..

okay that's exciting. I heard the same thing from someone on the fediverse, who mentioned that France doesn't even have a "fair use" concept, you just can't yoink someone's stuff even if you have what you consider a good reason.

This could be an approach that gets them to reconsider what they're doing

i always felt they were more than just a crane as they have a lot of purpose built stuff to make using those models easier (as well as modifying and running them) and from some anecdotal evidence it can be pretty hard to get them to take down training sets with your work in it.

Have you tried contacting the CNIL or, assuming you're in Europe, your local GDPR enforcement agency? Last time I had to, they were quick to respond, and they're not known to fuck around.

Ha. Been a while since I read the GDPR (and I'm not a lawyer) but my understanding is that you are not protected by anything under it as you are not a data subject. (Correct me if I'm wrong; I'm hoping not to fuck up my sleep schedule again by re-reading the entire thing tonight).

I think you may need some locals to complain to the CNIL for it to start barking. Assuming it does at all, because it's the INRIA and not some corp.
But I'm assuming you tried that already too?

So have you not asked locals about it? Unless you already did, I'm gonna start asking and poking around. There's a few ways I already see.

CNIL and Inria are the same people

I am not aware of that, - from my view those are two organisations both part of the Apparatus but operating very differently, - and I'd appreciate if you share what you mean in particular as I'm not too intimate with the internal politics at play in this context and I'm not yet sure how to figure it out (read: been a while since I last cared lol).

when I first reported Inria to CNIL for ignoring the GDPR, I found on CNIL's web page a news post about how they were partnering with Inria in some way

And from my experience and the experience of another trans woman who was trying to change her name, CNIL seems to give Inria the benefit of the doubt in all situations. Inria can tell them "oh we fixed it, the trans people are happy now" and CNIL will believe it and close the case, even if there is absolutely no truth to it

anyway, she heard more than I did. The status of my complaint to CNIL has been "Votre demande a été transmise au service de l'exercice des droits et des plaintes de la CNIL" since October 2022

Thanks.

they were partnering with Inria
CNIL seems to give Inria the benefit of the doubt

Yeah they are, and I understand they may; I wouldn't call them "the same people" but whatever.
I'm gonna see what I can do. I'm optimistic because all that I have in mind has absolutely nothing to do with being trans, but I will guarantee neither results nor timely updates (nor updates at all if this site goes down ;-;).