luoc

Particularly, it hinders the migration of pipeline hungry diver populations!

I mean, this is not exactly at the scale of Atlantropa[1] but also not too far off. And analogously to the Atlantropa project, the potential side effects are not even closely understood (and the article leaves me with the taste they haven't even tried to). Cutting of the baltic sea from the north sea? Sounds irresponsible to me...

[1] https://en.wikipedia.org/wiki/Atlantropa

Thanks for the link :)

Thanks for the insight! Saw this in a documentary and it's been a while...

Industry's response, of course, will be plants that epigenetically load proprietary genetic material at runtime...

I remember, there was a case where a farmer grew patented plants on his field. They spread to his neighbours field who then got sued by the patent holding company for not licensing that plant. My memory is a bit fuzzy here but you get the idea.

Now, how about reversing that? Think of a GPL-like licensed plant that makes all derived plants also inheriting that license. Legally poisoning the binary distribution format DNA of all inferred works. Wait a few decades and, with a little bit of evolutionary luck, wake up in a Stallman garden.

I'm just a naive dude with no clue about genetics or even law. Just a thought that came to my mind :^)

Dit heest Pfannkuchen

But how do I get that all caps scifi font to my reader? It's just not the same...

My bet: the people behind Telegram got massive stakes in this cyber coin and now (ab)use their power to attract some of their users into this ecosystem. I guess the intersection of users with "advanced privacy needs" and those open to any sort of crypto scheme is significant.

It's getting more exotic but I wouldn't say this niche doesn't exist either

There is nothing wrong with this sort of hobby. Give it a shot :D

What's the advantage of this over Hetzner Storage Box[0]? Seems to be the better option if I don't go into petabyte scale...

[0] https://www.hetzner.com/de/storage/storage-box

It's ridiculously easy, especially if you use SQLite. The only downside IMO is that they take the Bitwarden frontend as is which expects the full feature set to be available. For instance, if you disable email 2fa via the admin interface (the only component vaultwarden add to the frontend), trying to set it up in the app fails silently.

100% agree and I think the same issues apply. For our search algorithms we learned how to handle local minima etc.. I just wonder if those solutions apply to data driven companies since the iteration speed and failure cost is so much higher. I'm afraid good intuition will outperform this on the long run.

Only the lack of slurs reminded me I'm not browsing /g/ in the moment

Anyone here managed to like Odoo? It's the first ERP system I work with and the experience is sobering so far.

- no useful documentation aka read the code bro, it's open source

- "FIXME don't ever do this" comments all over the code base

- no update routine: download source of enterprise edition, patch your private repo, feed it in your CD/CI. I'm pretty sure vulnerabilities get patched quickly all over the world ;)

- weird design decisions like undocumented semantics for identifiers (variables, module names etc.)

- Python/JS/XML code living in the database (i.e. need the production DB to develop/test)

- performance

- addons needed for serious business use cases are not open source

- code quality of third party addons is a nightmare

Maybe this is still harmless compared to its competitors, I don't know...

I'd go for the decorator approach for its explicitness. It reads like plain English without having a deeper understanding of the concepts. Also, it does not pollute the implementors namespace. Even if one argues for the naming convention solution, one has to admit that it causes confusion for most people stumbling over it for the first time.

Why would one mimic matplotlib's API besides the fact that everyone knows it?

I use matplotlib a lot but I keep googling the same things again and again, which is my subjective indicator that something in the design is wrong. How to set log scales again? Do I apply this method to the figure or axis object? Setting a different backend (like agg) after pyplot was importet? Also, nested kwargs make it impossible for my IDE to give meaningful hints on parameters, not even talking about typed interfaces so far. Global state!

I mean, this library solves a truly complex problem and yes, I get things done but none of the hours working with it was close to be pleasant.

I wish some day someone is brave enough to enter new lands, leaving this matlab style of thinking behind.

I remember asking myself exactly the same question some time ago. There is third party support for PostgreSQL [1] and MongoDB [2] which allows (in theory) for collaboration as the database may be hosted anywhere. However, I haven't tested this and it's far from trivial to be set up securely for most people I guess. Also, I don't know if you can configure some sort of permission management in such a setup which is a must if you don't trust your collaborators.

[1] https://gramps-project.org/wiki/index.php/PostgreSQL [2] https://gramps-project.org/wiki/index.php/MongoDB

Process mining is a field related to data science. Basically you take an event log, put it into PM algorithms (e.g. heuristic miner or inductive miner) and you get a model of the underlaying process eventually (think of it as a flow chart). On top of that you can run analysis that look for bottlenecks, for example.

I'm not quite happy with the ecosystem though. Most tools are either cloud based like Celonis or organically grown Java bloatware (PROM, Disco). PM4Py also looks like a Java project that was tweaked until it was syntactically correct Python. However, it's probably the most usable free tool for PM out there. And to be fair, it's a quite young project also, docs looked much worse half a year ago.

Edit: added PM tools

Well, if you want to extend your note collection you run in a problem, unless you use decimals for page numbering.