TIL: ancient Romans used a concrete that was far better, more resistant, less prone to earthquakes, it costs less and has a better environmental footprint of what we use today.


Economists like to say that the economy reigns over everything, but they tend to always forget something: "it's the politics, stupid!" (quasi-quote).

COVID and the hearings about the Big 4 are showing just this: politics has left economy alone for a few decades, but now it's taking control and power back at the source.


Against IDEs:
- Having something remembering you that you're missing the 12th arg to that function is useful, but doesn't change the fact that having 12 args is the problem
- "How can you run this?" "Just push run on the interface" "I have to run it in another environment" "Ah, ok. Then I don't know"
- Jumping through definitions and usage is very useful, but as previously stated doesn't solve the problem of having a spaghetti architecture

Conclusion: IDEs are useful, but they won't save you from sinning

Hey kids, you might not believe this, but when I was a kid we were losing our sleep not surfing the web or social media, but trying to install software on our PCs!!!!

P.S.: if you really are very very young, "installing software" means making a program readily usable on your machine. Yes, no internet involved, everything is local and works (hopefully) as intended!

That feeling when you launch a training job with scikit-learn on a server and it everything fails seemily randomly because you don't have anymore space on the disk.

Then you suddenly remember: multiprocessing...

I just learned from my eye specialist that the retina has a one-to-one mapping to the points that we then see, and if you mess with it the brain will keep seeing the points as mapped in the old way, without recovery.

So this means that somewhere in our brain there's a lookup table for our retinas, and I think this is pretty cool!

I found this McCarthy interview (infoq.com/interviews/mccarthy-) where he bashes Ruby not being up to speed with LISP, and this other story where he supposedly does the same with Python (smuglispweeny.blogspot.com/200) and they've become my favorite things ever!

Since I used to work for a bank I feel uneasy whenever I look at investment/savings plans.

I end up thinking: what if the person managing my money ends up being like some of my coworkers? And I feel so scared about this...

I just spent 2 full hours trying to understand why some Python code wouldn't just go past a certain point.

I was looking for breaks, continuations, passing, etc, but I couldn't find nothing.

In the end I saw it. In the middle of a random function in library code. I just want to die very slowly and change career asap... (answer in picture)

Is it really a matter of functional vs OOP wanting functions/methods with at most 5 arguments, static data configurations, passing around at most 1 or 2 data structures and then having IO confined only at the edges of any program/app?

I think these are things that we all want!

So let's stop fighting/discussing and let's start teaching these things to people instead of classes or lambda calculus, those can come later and at that point won't matter much!

I recently had the chance to try catboost.ai once again after a couple of years, and I have to say that I'm positively impressed by it.

The last time I compared them, github.com/microsoft/LightGBM was at a better stage, but it seems that Microsoft has almost stopped development on lightgbm, while Yandex kept investing in catboost

Can we all just say that chaining us to scrum or kanban or whatever just defeats the purpose of agile?

It's clear to me that agile basically says: this is a creative job, devs are the ones who sould be doing creative and critical thinking and they need the time and resources to do that when they need to.

That's it, all the rest is just fluff

I know this is from 2017, but political theory doesn't really 'age': getpocket.com/explore/item/why

Apparently this is another American who just rediscovered hot water, he basically reinvented the 'elite theory' by Gaetano Mosca (wikiwand.com/en/Gaetano_Mosca), or if you like them better Vilfredo Pareto and even Machiavelli could be considered an elitist.

Nothing new under the Sun, as usual.

2 coworkers engorging cookies and laughing loudly are just disturbing all the other coworkers, they aren't collaborating.

Show thread

This might account for another 15% in productivity.

I'll stop here, but I could go on and on with other benefits that eventually will impact positively productivity, and it's clear to me that giving so much weight to just one thing - collaboration - it's just wrong and stupid.

Moreover, people seem to think that casual interactions = collaboration, but that's not true, to make it work collaboration must be a structured and deliberate process.

Show thread

Now let's say that this wonderful collaboration accounts for an outrageous amount of productivity: 30% (more than that would be just admitting that every company wfh shouldn't exist).

By wfh I'll spend less money and time on transport, so I'll be richer, happier and healthier. Let's say that this accounts for at least a 15% improvement in productivity.

Let's add that by wfh I'll have less disruptive interruptions during my work, and that I'll have more time to focus on the things that matter

Show thread

I'm tired... I keep hearing and reading: "the office stems collaboration, if we keep wfh we'll lose that!".

First off: who cares. If you aren't able to perform your job by yourself it's very likely that you don't actually have a job, but what you do is one of those bs jobs made up just to make someone else look better or more powerful or whatever.

Second: let's say we do actually miss on collaboration (I don't believe it, but that's not the point here)...


Show more
Mastodon for Tech Folks

This Mastodon instance is for people interested in technology. Discussions aren't limited to technology, because tech folks shouldn't be limited to technology either!