I have had windows and Linux on separate hard drives for the past 7+ years with no problems with upgrades. If they are on the same hard drive all bets are off though.
Visual studio is a pretty nice development environment. However, when compiling c++ software it routinely takes much longer than compiling with make on Linux on the same hardware (dual boot). Close to 10x longer! This is with cross-platform cmake projects. Anyone else experience this?
I think sublime merge is pretty close to the optimal git GUI, https://www.sublimemerge.com/. It uses standard git commands and terminology and allows you to add custom git commands to the GUI. Unlike Sourcetree is cross platform for Windows/Linux/Mac and includes a merge tool. I think command line git and Sublime Merge git translate back and pretty easily.
I generate the QR code and then add some text to make a kind of poster and print it out. Ios and android users can then just use their camera app to automatically connect to the wifi.
I think wxWidgets is fantastic. I use it in my CMake project. I have a git submodule with the wxWidgets source. I just set the options I want and add_subdirectory(). I statically link it it into my binary for easy distribution, because the license is more permissive than LGPL in that regard. Thanks to the wxWidgets collaborators for all their hard work!
Sublime Merge is just much easier to use than VSCode git integration. Yes VSCode can do everything, but Sublime Merge is just a much better experience.
I really like Sublime Text, but I have mostly moved on to VScode. I will probably buy a license just to support their software, especially Sublime Merge. For me the Sublime Merge git client is an essential.
I think sublime is a fantastic text editor. But it just doesn't cut it for me as a development environment. Getting productive in vscode is easy. It just works. Though Sublime hq's sublime merge git client is fantastic. I don't really care about sublime text anymore, but I'm a devoted user of sublime merge.
I wholeheartedly agree with the sentiment of this article. In addition to this kind of policy being elitist I also think it is racist. Those attending higher education tend to be from more well off backgrounds and white. This is where the whole affirmative action thing started, under representation of minorities in higher education. It is a snub to those who could not attend college due to their socioeconomic background. There are many struggling people who never attended college and don't have student loans, what about them?
EDIT: I would directly benefit from this as I have student loans, but if I could make the decision to forgive student loans like this I would say no.
That might be an ok solution for consumer devices, but the kernel drivers I'm talking about are for custom setups that will only be used internally by me. There would never be any reason to upstream them. Constantly upgrading the kernel just creates a lot of work.
I really like that the AMD drivers are open source and up streamed into the kernel. However, that doesn't solve all the use cases. If I want to use newer hardware on an older distribution (RHEL, DEBIAN, Ubuntu LTS, etc) the upstreamed drivers will be too old. Distributing the driver like Nvidia or Radeon Software for Linux will still be necessary.
I think that Numpy with Numba for routines that aren't easily vectorized and or need if statements is just about the perfect combination. https://numba.pydata.org/. Numba functions are very close to C/C++ performance.
EDIT: You can write loops explicitly and have OpenMP style for-loops and they are fast.
Numba I think is the perfect tool for numeric code. Much easier to work with than Cython or native modules. You can also use numba to write Cuda functions.
I really like wxWidgets. It is very convenient to add the wxWidgets repo as a git submodule and add it to your cmake build as a subdirectory. The licensing is also great. I highly recommend it! A fantastic project
For me, flatpaks are a better user experience in just about every way. The one feature they are missing is that there is no equivalent to the snap --classic confinement. That means that for applications like vscode the flatpak experience is very poor. You cannot easily access command line tools from the editor. For instance using the anaconda python distribution from the vscode flatpak is a pain.