In the Italian dubbing they manage to get both wrong by translating them respectively "a base di carbone" and "a base di silicone" which actually mean "coal-based" and "silicone-based".
Infuriating :)
609 karma · joined June 15, 2015
In the Italian dubbing they manage to get both wrong by translating them respectively "a base di carbone" and "a base di silicone" which actually mean "coal-based" and "silicone-based".
Infuriating :)
So much cognitive load.
Just configure Vim to put everything you delete in your clipboard and use a clipboard manager like diodon[1] with 1000 items.
Of course works also outside of Vim, you just have to copy the part you might want to reuse before deleting/modifying.
In Italian weddings is customary to give a "bomboniera" to guests, a small party favour as a souvenir and thanks for their attendance.
Usually it consists of some formal bullshit like a silver spoon or small ceramic thing. My wife and I wanted something a bit more fun and Munari's "supplemento al dizionario italiano" become ours.
[1]https://www.amazon.it/italiano-Supplement-dictionary-Supplem...
IPhone is unusable in this regard. Switching from one language to the other is just not cutting it. That's simply not how polyglot people communicate.
The reason I switched back to my Samsung Galaxy is because I changed job and will need to use it for work, and frankly the Jelly is just too small to be efficient.
It is otherwise an amazing little phone.
There's an easily searchable index of B-Corps: https://www.bcorporation.net/en-us/find-a-b-corp
https://www.reddit.com/r/law/comments/3pv8bh/comment/cw9t52d...
https://business.gov.nl/starting-your-business/choosing-a-bu...
https://web.archive.org/web/20220401182358/https://www.busin...
- miniconda (~2GB)
- a final RUN chown -R statement (~750GB)
We reduced the image size and relative Spark cluster considerably by playing around with dependencies in order to stick with plain pip and using COPY --chown.
I also recommend [dive](https://github.com/wagoodman/dive) analyse what contributes to each layer.
I guess I need to go out more.
I guess it means you'll have to go back! Next time go to Florence and order a Fiorentina Steak with a Montepulciano. That's going to be 50€ for two. And, as you said, the memory will last forever.
On a personal, nostalgic, note: running LFS on your laptop in my university was the ultimate Linux nerd cred. A step below, immediately following, there was Slackware.
The reason I feel confident enough working with Linux today is because I started back in the early '00 using Slackware, which recently had a new release[1].
Don't get me wrong, compiling your own kernel, solving shared libraries compilation hell, messing up your init.d and starting over, leaving your Pentium 4 compile for several hours was a massive waste of my 20s. But, boy, did I enjoy it!
“You should, in science, believe logic and arguments, carefully drawn, and not authorities.” [1]
[1] https://twitter.com/anammostarac/status/1495594139865731074
[1]https://www.bbc.com/travel/article/20210225-why-the-dutch-lo...
It worries me that it is actually going to work.
A news source these days doesn't need to be accurate. It is sufficient to be well-known.
- communicate your progress more often. The status of the project should never be a surprise, especially when things are not going perfectly well. Write it down, send slack updates, be proactive in communication.
- being proactive includes asking for feedback and for help when needed.
- avoid doing with the goal of being praised. You know when you're doing well, you know when you're learning. That's all that matters really. Especially in your first job. Your goal should be self-reliance.
- avoid attaching your sense of self worth to the code you write. You'll write great and shit code. That doesn't reflect on who you are. You'll be horrified at the code you wrote 2 years ago, but you will also be proud of how far you got.
- you are good enough
Take it easy my friend. If the goal of mountain climbing was to get to the top then people would summit the Everest on a helicopter. Enjoy every step of the ascent instead. Slow down and enjoy the view sometimes.
And fuck your boss.
I wouldn't be surprised if a freelancer would charge 100$ per hour to do this kind of work spanning multiple work days.
Upon reading these words, he suddenly awakened to enlightenment.
In reality, increasing debt to a certain point also increases risk, which in turn increases return on equity.
I am also a freelance consultant but in the cloud infra space. I appreciate the freedom it gives me and the insight on the operations of running a business are invaluable.
As a software engineering I always looked suspiciously at accounting and finance, but I have to admit I've been wrong. I discovered a practical and theoretical elegance that is very pleasing for someone like me.
I ended up enrolling in an MBA because I wanted to learn more about the business side of things. I had a few courses about corporate finance, valuation and accounting which filled many gaps I had.
I still don't trust myself running my own finances completely, so I do pay an accountant to take care of the details. But I am very happy I know financially what is what when deciding where to take my little business.
Knowledge is power, especially when it comes to finance.
I might be romanticizing it in my memory a bit, but I remember it fondly.
We would run ejabberd servers and could easily know which nodes were up and talk/send commands to them.
The scale and quantity of software being developed is just too much for traditional Linux distributions approaches. I personally rely on Debian to know that most of the packages I use are reasonably secure. But often I need to sacrifice flexibility and bleeding-edgedness.
And it couldn't be otherwise. There are 339,267 projects on PyPi.org. I bet a good percentage of these are riddled with security flaws. But people use them anyway. How many distro maintainers would you need to handle this workload? Is it worth it? Does anyone care?
It looks like people care much more about experimentation and speed of development than stability, security or coherence and cleanliness of the solution. The author seems to feel this is wrong from an engineering (or even moral?) perspective.
I am also uncomfortable when I have to deal with Python and I have my own favourite few tools... but perhaps if Python users really wanted a single solution then it would already exist.
If you are instead convinced that there is this need and no adequate solution, then congratulations, you just found a gap in the market. Go on and do better than everyone else before you. Relevant XKCD comic is already in the article...
Make the overlay of your dreams!
https://www.sciencedaily.com/releases/2014/09/140924113744.h...