We all have those tasks in our applications that are not common enough to warrant a project (and another deploy) to make an admin panel, but also are painful to do, because they are tricky or time-consuming. So I'm working on an internal tools building platform to quickly wrangle some admin interface do automate them.
This is currently just a prototype I'm developing, but I'm honestly eager to receive some feedback, as I'm not completely sure about the direction.
I'm building a internal tools engine partially inspired by Interval (RIP).
It will be completely open source from the start and support many languages other than JS.
It is in the very early stage of development, but it would be nice to know if anyone other than me is interested, and if you have any tips on what bothered you in other similar tools or things you would want on it.
This. I used to be very skeptical of anything Blue Origin. But after the CEO change they appear to have changed their attitude for the better.
They are not on SpaceX level, but they are growing recently and I think this test, even with the many problems or things I didn't like (SpaceX spoiled us), it was positive.
You must hear this a lot, but.. do you have any tips on how to break into the field?
Currently I'm a backend engineer, but my initial interest to dive into development was operating systems and networking.
I recently have been improving my knowledge doing some systems related projects and trying to contribute a bit to open source, but getting a role working with that still seems so unreachable not having any formal experience...
Security consultancy companies need to always point out something that needs to be changed, even though it is not really important to show themselves as useful.
And executives don't have enough tech knowledge to discern between security measures that are actually effective or not, so to avoid risks they just make their tech teams implement it because the consultancy said it should be done
Had a similar situation in my current job, and unfortunately it is not something worth picking a fight with senior leadership for.
Ironically most of these companies allow access from Web Browsers (which are completely controlled by the client).
It is. They realized that writing a modern browser (i.e. one that for example uses multiple cores efficiently by doing layout/rendering/etc in parallel) is almost impossible in C++. To the point where creating a whole new language just to solve the problem would be a smaller undertaking. Which says something about the scope of this problem. And I really do think they are right.
> writing a modern browser (i.e. one that for example uses multiple cores efficiently by doing layout/rendering/etc in parallel) is almost impossible in C++
C++ also happens to be the only language in which that has been accomplished, so far. (The Rust components are relatively smaller, about 11% according to sibling comments.)
Why is it almost impossible? I'm a little out of practice with C++ but I thought the modern C++ features were considered pretty solid in regards to memory safety and the like?
c++ isn't memory safe and will probably never be. that being said, it's not more or less possible to create an efficient browser engine for multicore systems with c++ than any other language. programming language is not the thing that makes the task difficult.
> They realized that writing a modern browser (i.e. one that for example uses multiple cores efficiently by doing layout/rendering/etc in parallel) is almost impossible in C++
And yet Chrome does it just fine, and Ladybird can render some websites properly that Servo cannot, curious.
No, Chrome doesn't eg. do style computation with a parallel algorithm like Firefox does since it adpoted Servo's style resolution engine. See how faster it is: https://nolanlawson.github.io/css-talk-2022/#70
Servo's rendering issues are a different class of bugs unrelated to performance: they are still working on the layout parts.
Performance is overrated and irrevelant. It's simple force - not a skill. I can't do in more performat Firefox, neither it doesn't matter how performant Chromium, things that was possible in Firefox with old extensions. Not having such abilities, the only thing left to Google to talk about was.. performace - and crippling powers of Firefox. Rust is a new skill, not more of force.
Another skill, inconvinient, forbidden and forgotten ( * * * that I wish Ladybird will have * * * ): plugins - so I have A Browser For All Documents that can play every video format regardless of web video codec support or licencing etc. (with vlc, mplayer, whatever).
Neither Javascript shall be needed to see anything on the Web - if not being essential for tracking and ads
- but it's you who have to pay for it with that threadrippers (and no access to source data) ? Yes, it happend people are rich enough to belive that.
Making things simple and possible is the most important skill.
Check out this comment: https://news.ycombinator.com/item?id=40862781. While it's certainly possible to write the same parallel algorithm in C++, it wasn't done in Firefox before Rust, nor has it been done in Chrome.
But you can implement the same algorithm in c++, you can even implement it in assembly, or brainfuck!
I do not care about c++ or rust or any of the language wars stuff. All I wanted to do was point out that the initial statement about disregarding performance was ridiculous and also point out that majority of the performance critical libraries are written in C++.When it comes to numerics its a mix of C++, C, asm, and fortran.
Use your brain not mine, read and repeat. You can't measure performace of something if it's not possible to do that at all first. Like.. of your understanding till you do - then you can work on it (to make it less noisy and that, not how fast, is what for others matter).
This is currently just a prototype I'm developing, but I'm honestly eager to receive some feedback, as I'm not completely sure about the direction.
https://github.com/luizgfranca/synthreon