david_draco

His formal background is in theoretical physics not astronomy, although he has been at Harvard's astronomy department for decades now.

Bugzilla 3.0 in 2007 was a significant improvement in usability.

Maybe an "Acid test" for Unicode would help? These pages seem to go into that direction: https://www.kermitproject.org/utf8.html https://web.archive.org/web/20160306060703/http://www.inter-...

Placing a fuzzy tester like "hypothesis.strategies.characters" into the CI may also be revealing.

Perhaps a good way forward is to come up with a positive incentives, like already exists with the various badges for CI success, test coverage, documentation, valid HTML5, etc. There could be a service that verifies that PGP key can be fetched and the latest pypi version verified. To avoid putting the burden of hosting on you, perhaps you could provide a python script which takes the package name and allows people to self-validate?

You are mistaken to think Stallman is primarily a software person. He is a leader of a social movement, with social goals (freedom from the shackles of proprietary software) achieved by strategic investment in software development and legal battles over the decades. From this lens, it is quite his lane.

"No one is going to ban Google Analytics" -- no need to explicitly ban individual products, the rules are written down in GDPR

I use OpenGPG for signing and writing emails, they are plaintext encrypted. https://support.mozilla.org/en-US/kb/openpgp-thunderbird-how... https://www.linuxbabe.com/security/encrypt-emails-gpg-thunde...

Enigmail functionality is now built-in, it's quite nice!

Unfortunately the web keeps changing. Firefox is trying hard to remain relevant and needs to be flexible with their code-base. So Thunderbird does not live in a stable world.

"The Cause of Depression" assumes there is a single cause, which may not be the case.

> the idea that you'd leverage open protocols but try to police them like a walled garden feels backwards.

Just because it is technically possible, does not mean it is socially a good idea. Members and server owners should decide what can and cannot be done with their data. If you want a Mastodon server without limitations for experimentation, you can create one, and people can join if they consent.

Would it be possible to use GPT model infrastructure to compute the likelihood that a given text was created by that GPT model?

This is pretty nice: https://www.modernwiki.app/

Maybe machine learning techniques for translating text could produce unstripped binaries based on stripped binaries, given a large training set?

Did this with Gentoo for a decade, even across two computer changes. boot into a rescue disk, set up grub, then for each partition do ssh oldcomputer "gzip</dev/sda5"|zcat>/dev/sda5, then resize the fs.

I see two previews, one next to the file name, and a large one on the right when clicking on a image file name. So:

25 years, now with preview images in file picker :)

If in distribution discussions Linux is name for the operating system and shell, downplaying the role of GNU, then it is also fair game to say here: Linux does not have a stable ABI because glibc changed.

Someone should sign up for all the mailing lists with a email address used nowhere else and track the cross-mailings. Maybe a bubble babble hash of the company name as the email prefix, and a big mailer like gmail or protonmail as the server. When the email is leaked and the company does not inform the user, report the company via GDPR.

"profile view challenge" coming in 3, 2, 1 ...

Usually you have a landing page, and then you enter stuff there, and finally you get output. That gives you about 1 minute before returning first results.

If the user is faster, you can show something like cloudflare does when you visit through Tor.

On later submissions you can reuse the reputation from the cookie.

The suggestions posted here (give a quote, say no, say you are not interested) are part of a larger, useful guideline:

Respond by specifying clearly what would be required for this feature to land in your code.

This can be money, this can be someone stepping up to do it and maintain it, this can be technical requirements under which you can see this go into your framework. Yes, users opening issues are demanding, but as a github repo owner you can ask for anything as well and put up barriers that the user needs to solve to see it happen. They may be motivated too.

Specifying clearly the problems and what is needed to overcome them leads the user to appreciate you more. It will sharpen your thinking on why exactly you are saying no. And sometimes you will be surprised by someone actually making you happy by meeting your requirements.

Have a "CAPTCHA" that gives the IP reputation for some time (cookie+IP=key), but instead of a CAPTCHA make the web page / browser solve and submit a BOINC task from a randomly picked science project. No user interaction needed, it has the benefits of "paying by computation" of cryptocurrencies without the tracing, and if bots solve the problem efficiently, it's good for science.

Is the "right answer" a classification based on US English with background knowledge of US cultural background, or is the goal to build a global sentiment data set?

You and OP ("the indians labelers don't know how to do it correctly") seem to want the former, so it would be good to state that goal upfront.

Therefore, one could easily generate lightning with equal destructive power? Why is it not used yet in war (airplane to ground) or for entertainment (airplane to airplane)?

Have a look at notmuch

The most convincing arguments for not having a blog, however, cannot be found in a blog post.

I also wondered why the author set it up to echo the existing comments, instead of for example building a bot that asks people for evidence supporting their claims, or tells people they appreciate them, or whatever else one could possibly think of for a bot to do.

You could argue that bots are expected by the hosters and people there, and they are happy to be experimented with (ok, not sure what's the evidence for that argument), that's one thing, but another is releasing the model freely and not being bothered by ethical concerns.

The world is a better place by creating and running a vile bot. Start a social experiment with users without their consent. Lets empower others by making it as easy as possible to auto-generate vile content. It's okay to pollute the /pol/ board, nobody needs to know after what was bot-contributed and what was not. Let me get fame and money from posting videos about this. -- the author

I wonder if that person ever took a CS class on ethics and the connection of computing and society, their compass is way off. But keep smiling into the camera.

"Gerrymandering is universal in any political system that has geographic units that demand representation." A unstated assumption here is a winner-takes-all election system, which is common in the US and UK, but not other countries, where parliament is proportional, and coalitions are common. It works.

A more interesting question would be how to build incentives in a winner-takes-all election system towards reducing gerrymandering, and to discuss what it would take to change away from a winner-takes-all geographical election system.