There is no way I'm going to entrust my data to a vibe coded tool. Sorry man, you seem proud of it but it's a nonstarter for me.

> I'm not even super proud of it

well, scrap that

What does this offer over using rclone? If it’s not significant functionality, why not make pull requests there?

> It's fully written in Go

> I am not a Go developer

what a time to be alive

This wasn't a particularly unusual thing to say even before AI.

It's funny to me that this is how JavaScript (and PHP?) got its bad rep (before Node forced it into a tailored suit). You don't have to understand it, just copy and paste this and put it on your site.

Proton Drive already has a CLI built with linux users in mind by the Proton AG team...

https://proton.me/business/drive/cli

https://proton.me/blog/proton-drive-cli

Thanks for this. Been with Proton a while and they're not great at letting their customers know of new products. Either that or they are too successful at filtering their own marketing email.

To the OP: Proton is privacy first. I use their tools for a reason, trusting file transfer to a third party would never happen even with an open code base.

It's fine if the tool is for yourself, great for your CV and well done for shipping something. HN can at times be a cruel place, think of this as a learning experience.

> great for your CV

Caveat I am not a SWE. Is there any value in having AI developed tools on your CV?

Before AI, it at most showed an interest in coding outside of work. There's a 'passion' element that might be interesting to recruiters but there's not much more to it than that.

Look at the amount of AI-required jobs advertised these days. If you really like driving agents, someone will absolutely pay you to do it.

Yes - because most software will be developed this way in the coming years.

They released it just few days ago (maybe 2 weeks or so). I personally got a mail in my inbox for the release, but maybe it's because I have left feedback about it or something. A very welcome news nevertheless!

got the email on october first, just yesterday i switched my restic backups from hetzner storagebox to proton drive. works like a charm

Can you comment on any speed difference?

Why let things like "this thing already exists" get in the way of reinventing the wheel with AI?

>Be kind. Don't be snarky. Converse curiously; don't cross-examine. Edit out swipes.

https://news.ycombinator.com/newsguidelines.html

A command line access tool is very different from a mounting tool.

I wish it was possible to pipe into it tho.

e.g tar -cf - * | <some unwieldly proton drive cmd>

I guess this is not possible due to the way the encryption works?

Its crap, it can’t do sync, only upload and download

From https://proton.me/blog/proton-drive-cli :

"Sign-in happens through your browser — no password on the command line."

which I can confirm having tried it myself. It's hard to imagine what the proton devs were thinking. Did the case of using a command line tool on a headless or remote server never occur to them?

The browser doesn't need to be running on the same machine.

I don't really care what you do with AI on your own computer for yourself but top tier irony sharing this around with this post on your blog: https://blog.lsantos.dev/en/backlog-10-slopware-the-end-of-w...

Considering you ignore the other post on the subject, it's clear he came to terms with it. People's opinions evolve over time, and calling someone out for having an evolving viewpoint is anti-intellectual. I think it's perfectly fine to have an opinion one year and have something else the next year, and not feel the need to delete your old writings.

I don't know, I find growth and learning important.

Why are you opposed to changing your mind?

I don't know if the OP is opposed - maybe they just want an explanation? I mean saying something is ironic isn't really a harsh insult or a bash.

The post is talking about vibe coding not AI in general. It's about people who don't even read the code and ship unfinished React sites with performance issues and insecure backends. This project is fully client side, written in a fast language and only connects to Proton servers through their SDK so any risk is minimal.

Last time I tried ProtonDrive(years ago), the speeds were complete dogshit, I could barely surpass 10 MB/s on a half gig fibre line. Anyone used it recently, and if so has it improved?

I’m not a customer and don’t know but IIRC they’ve been rewriting the storage API or whatever it is and it should resolve that exact problem.

Not sure, but that would explain how Google photos are always in sync and proton drive is always backing up my photos with 50-70 items to go (with no way to see which pictures/videos are backed up). Often their VPN will prevent even their own sevices from working and I have to reconnect manually to make it work. Never had these issues with mullvad.

I'm generally not very happy with proton's services or apps. When this year of my payment is over I'll be switching away.

Yeah, I'm also switching away. They're very sluggish on features. I went through all the rigamarole of setting up my own domain for the email service. Specifically with the idea that I could them automatically filter all my mail into folders by service. But, as it turns out, Protonmail has very limited support for sieve extensions. The crucial one is being able to create mailboxes from sieve scripts. So I can sort emails into folders, but I'd have to go create a bunch of folders, and make a new one every time I sign up for something. I'll be switching to fastmail, I think.

Though, if ProtonDrive became fast finally, I might still be convinced to pay a smaller fee for PD plus their VPN, which is completely fine for my purposes anyway(mostly just torrenting).

I think they are spreading themselves a bit too thin by trying to have an answer to every Google feature. I would have stayed if the core stuff worked. Would have been a lot better if they made some privacy focused API other apps could hook into and other providers could copy, and then we could create an ecosystem that replaces Google.

Good on OP for being transparent with how it was built

What’s wrong with `rclone mount`?

Wait, you can mount your Proton Drive with rclone? Hah, never knew. It's in Beta apparently though: https://rclone.org/protondrive/

Isn't rclone copying everything rsync style (at least for Proton) ?

it was very slow, and I had to re-authenticate often. I want something like dropbox daemon, which mounts drive and syncs it. Set up once and forget. Since I have not found anything like this for Proton I am still using Dropbox for some synced data.

Tried doing this with Proton Drive one time. It broke in several different ways simultaneously. Doesn't seem to be ready yet, but maybe some day it will be good for this.

This is great! I recently started working on a privacy-first MCP server for proton mail, calendar and drive, written in rust that supports optional pseudonomization of identifiers in the data: https://github.com/matt-w-horn/protonctl

Supports mac and linux, but the linux part could use a bit more work.

If Proton got their devs together for a two week internal hackathon, they could build a very simple Proton client for Linux that would match functionality of Dropbox and quickly gain a lot of new customers. And make existing customers very happy thus creating free publicity to attract even more customers. Alas, their priorities are trying to upsell unnecessary products to their existing customers and not providing products they actually want. And I am saying all this as a paying Proton customer.

I work in this space with my oss work on Filestash, and its sync client https://github.com/mickael-kerjean/fdrive. A proper sync client is nowhere near a two week hackathon unless you vibe code it and don't have much care. I've been on it for 6 months and it's far from finished. Linux alone has a fragmentation problem. The sync engine can be shared, but the integration can't: tray, status, file manager overlays and notifications all work differently on GNOME, KDE, COSMIC, Omarchy, Sway. Electron looks out of place everywhere, so you end up maintaining one frontend per desktop, plus packaging for each distro. Then there's the sync itself, which is a long tail of edge cases, often platform specific. Last week I optimised the rm -rf case, FUSE hands you one unlink per file, so a folder with 10k files becomes 10k API calls unless you do clever things. Today, I've tried to optimise thumbnails, problem is the file manager opens every file in a folder just to render previews, so browsing a directory of videos can easily get uggly as well, let's not talk about the state of atomic save which if you are not clever here too will create multiple files on your server and waste bandwiths. There are many dozens of these edge cases just on linux and that's before you tackle window, mac, android, how do you handle delta sync ....

My "two week" estimate was too sarcastic. Thank you for detailed explanation about all the problems and challenges that comes up when building a file sync client. I really admire your attention to detail while trying to correctly integrate the sync engine in all different file clients.

Cool.

I'm using gdrive under Linux and it's a pain. Takes 10 seconds to open a folder.

I have a proton subscription and would love to move my stuff there if it works reasonably faster under Linux than gdrive

Very cool I like it!

It's like Proton is going through their fat Elvis stage. They can ride on their name for a while but who knows how long.

i heard proton shuts down accounts just for using tempmail for verifications and refuses to help after, calling it a TOS violation. Also they use zendesk as their ticket system while claiming absolute privacy. And do not even mention they do not even accept XMR as a payment while claiming privacy. It is just another corporation selling privacy in its most profitable form.

good job. I was there once but gave up..