Open-sourcing Sonar, a new extensible debugging tool
code.facebook.com
code.facebook.com
Although I must admit, I'm disappointed that they didn't use react-native-web to build this app in a cross-platform way. I agree Electron is a good choice for the current landscape, but I'm hopeful that we will be able to move past it. If Sonar was implemented with React Native's layout API, it would be easily portable to projects like react-native-macos, React Native on Windows, or projects like Proton Native, once it matures and stabilizes.
Of course Emil knows this, because he is one of the primary maintainers of Yoga, the cross-platform flexbox implementation in React Native. This is one of the primary technologies that will allow us to move away from Electron one day, if people would just stop using CSS!
Like I get that people don't like Electron because of it's resource heavy nature but it's been a boon to us Linux users who have finally been getting a lot of cross platform apps and I truly appreciate that.
getting a Mac is generally not an option - it's too expensive just for a few GUI apps (and overall pushes the bar of development quiet a bit higher) and in all honesty, it's not great for the developer ecosystem to stick to one platform either (new Macbook with the touch-bar is a good reminder of that).
I am currently fortunate enough to own and work on OS X, but I do remember when I was in university with a Lenovo T420 running Ubuntu 12.04 having issues with quite a lot of tools when I was just getting started.
If the app was built with RNs layout API, you could continue supporting Electron via React Native Web. But there are awesome upcoming projects like Proton Native that support Linux UI properly on top of libUI.
There was a react-native-ubuntu experiment that unfortunately died down because it forked RN, rather than the Windows approach of using real RN releases.
There are quite a few options without packing a browser engine with the application.
If that is the only way to get applications on Linux, then we should just all move to ChromeOS and dump the Linux desktop attempts as is.
That's a bad example. Gerrit is not a new name for Reitveld, it's a totally new implementation (with similar inspiration). And both of those were publicly released under those names.
Your underlying point is quite true, though.
I am familiar with most of the other tools mentioned here (e.g. SonarJava) from my time doing Java development outside of Facebook and never had a collision in my head because this project is not actually a library or tool for Java. It's a standalone Electron app that works on iOS and Android for debugging purposes, the only Java part is incidental in the Android app (and could be Kotlin or C++ or whatever you want).
I'm not actually interesting in debugging mobile apps, so it would have been nice if that was mentioned at the start. Now I had to wait for the third section, not very helpfully titled "An open source architecture", to find any clues.
> With Sonar, engineers have a highly flexible, intuitive way to inspect and understand the structure and behavior of their iOS and Android application
...and the top of the post is tagged as "Mobile", "iOS", and "Android".
> I had to wait for the third section, not very helpfully titled "An open source architecture", to find any clues.
Python, C, Java, Go, Swift, Ruby, Assembly, Basic, JavaScript ...
It would have been chaos. Obviously none would like that, would they ?
Also, many of those you name collide with words that have no particular collision within the industry, so while you and I might think "Swift" is a non-ideal choice, it's not the same as the problem people are speaking of here.
JavaScript gets a pass because it was intentional. Not wise, as recruiting emails point out to me painfully regularly, but nonetheless intentional.
Go is the major and blink-tag-in-neon exception. I have no idea how naming something that is not only one of the english words with the most definitions but ALSO a game that has a higher-than-average overlap with people in the industry AND one that has been a significant target of the industry survived. I can only assume that someone underestimated the concern or I'm overestimating the actual pain.
Nowadays, however, there's no reason to repeat the Firebird debacle and fail to do a simple web search within your industry. Particularly if you're with a large company. Doing so earlier in your process is better, but even a last-minute name change can avoid generating a lot of ill will and confusion.