wallunit

This is actually rather a reason to avoid Python in my opinion. You don't want pip to pollute your system with untracked files. There are tools like virtualenv to contain your Python dependencies but this isn't by default, and pip is generally rather primitive compared to npm.

"Penalizing solo travelers" is a hell of a spin on quantity discounts. If this isn't click bait what is?

I used zopflipng in the past to optimize PNG images. It made sense since there was no better alternative to store lossless image data than the PNG format at the given time in the given environment. Zopfli is awesome when you are locked in on deflate compression. I feel like if the npm folks would want to optimize for smaller package size a better strategy would be switching to some more effective text compression (e.g. bzip2, xz). That would result into a larger file size reduction than 5% for a smaller CPU time increase compared to Zopfli. You would need to come up with some migration strategy though as this change isn't per-se backwards compatible, but that seems manageable if you are in control of the tooling and infrastructure.

That's awesome. I'm looking forward to any feedback. Not sure for how much longer I will keep monitoring this thread. But you can reach me either by filing issues on GitHub, or if you could at least drop me an email at [email protected] and share your feedback I would much appreciate it.

Honestly, when I first started working on mypass, I wasn't aware of "pass".

Anyway, one issue I have with "pass" is that it leaks metadata, as it uses the file system to organize different credentials, while only the files storing the credentials itself are encrypted.

Also "pass" uses GPG for encryption, which can provide additional security if you store your private key on an external drive or smartcard, and take additional measures to make it more difficult to obtain access to your password store and private key at the same time. But in the more common setup where the password store is stored along the private key on the same device, cracking your GPG passphrase will require less computation than cracking a passphrase using PBKDF2 with 256,000 iterations like used for key derivation in mypass.

I'm the author of mypass. First of all, I'm very sorry to drop support (temporarily at least) for macOS and Windows with the latest release. It was a necessity for adopting SQLCipher (without available bindings for Python 3). But I reached the point where rolling (and encrypting) my own data structures is no longer a reasonable option (in particular with added support for multi-context credentials). On the other hand, the latest release adds support for FreeBSD (in addition to Linux).

Unlike most Show HN posts this isn't a new software, but one that I initially created back in 2014, and ended up spending a fair amount of time on recently, leading to its 2.0 version which I released today. In the beginning it was just a hack scratching a personal itch of mine. But I think mypass has matured a lot since then, and I'd love to get some feedback from other potential users.

Is that even true? In practice rather than downloading a single package you'd download/update a bunch of packages over the same connection, and an attacker would only see the accumulated size, right?

No, I never used Facebook (or WhatsApp or Instagram) in the first place. ;)

Awesome, I'm looking forward to see whether it is working on your Acer laptop.

You can use anything that is exposed by the Linux kernel under /sys/class/leds/*. This includes the keyboard backlight of ThinkPads. As for other vendors, it depends on whether the keyboard backlight (or any other LED) can be controlled in software, and if so whether it is supported by any kernel driver.

Thanks, man. It's almost frustrating how few attention my projects that are targeted to Linux users get nowadays on HN. That said, I appreciate your appreciation, very much.

If you try it out, let me know how well it works. So far I only had a chance to test it on a ThinkPad X230 and X250, and I would be interested to know how well it works on other models, potentially even with other vendors. Though, I haven't seen any non-ThinkPad laptop yet that has LEDs that can be controlled through the Linux kernel.

Also in case you are using Firefox on Debian/Ubuntu, note that blinklight requires at least Firefox 50, which is not yet in Debian stable/testing, but Firefox 52 can be installed from experimental.