mort

What's everyone's view on the #EUPL compared to the #GPL? I read thru it and my impression is that the EUPL achieves most of the same things as the GPL but with much less convoluted language, and written by EU lawyers rather than #FOSS activists which instills confidence.

This part feels a bit weird tho: "The Licensee must cause any Derivative Work to carry prominent notices stating that the Work has been modified and the date of modification." What's "prominent"?
#law #OpenSource #programming

Oh yay #clangformat updated and is now even worse

(No this can't be fixed, give me a .clang-format file which fixes this particular case and I guarantee it'll break 10 other cases)

#cpp #cplusplus #programming #clang #llvm

I found a sound asset I like on #Pixabay, and the license seems cool based on the summary: https://pixabay.com/service/license-summary/ -- you can use the asset in derivative works, modify it, no restricitons on commercial use, etc.

However ... that page is just a summary of the license. Where in the world do they put the actual license? I can't find a link to it anywhere? The terms of service also contains a summary of the license but also doesn't link to the actual license text... What am I missing? #gamedev #law

Oh my god every time I have to use a new version of #Python I find that it broke something, this time I found thaht Python 3.12 has removed the built-in 'imp' module which broke the version of waf (a build system) used by the version of Jack that Yocto Kirkstone uses

Python is literally the most unstable programming language I have had the misfortune of using, they constnatly remove or rename widely used functionality

#programming #yocto #openembedded

Has scientists figured out yet how to get clang-format to not bizarrely try to align chained methods? #cpp #clang #llvm #clangformat #cplusplus #programming

Packing components snugly together is the most satisfying part of PCB design #electronics #pcb #pcb_design

Hmm #AsahiLinux doesn't want to boot today, that's unfortunate.. the fedora boot spinner just spins for a bit, then stops spinning, then the laptop reboots into macOS #fedoraasahi #linux

It's actually wild that regardless of how much time and effort you spend trying to prevent clang-format from aligning stuff, you'll still stumble upon cases which it absolutely insists on aligning. You'd think writing a non-aligning formatter was easier than writing an aligning one?

Worse: it occasionally aligns with tabs if you use tabs, even when you tell it to align with spaces

#ClangFormat #clang #llvm #cpp #programming

Hey fellow #embeddedlinux people! Have you ever wanted to generate a canonical representation of your devicetrees? Or diff two devicetrees? How about adding back labels to a decompiled devicetree? I've polished up and released a little tool I made for these things called DTCanon: https://github.com/mortie/dtcanon

This has been useful for me in some situations. Maybe it'll be useful for you too. It's #FOSS of course.

#yocto #linux #devicetree #opensource #programming

You may not like it but this is what the ultimate small factor PC case looks like

(Yes there is dust, these parts have been sitting unused for many years now and I don't have my compressed air with me) #pcbuilding #pcbuild #smallformfactor

I'm not hating this #json #json5 #programming

Things I would change in the #devicetree syntax if I was god emperor of the world (or at least the embedded Linux part of it):

* Make semicolons after closing curly braces optional (ideally they'd be prohibited but that'd be a breaking change and I'd be a benevolent god emperor)
* Make status = "enabled" a synonym to "ok" or "okay", to match status = "disabled"

These two small changes would remove 99% of the mistakes I make when doing devicetree stuff

#linux #embedded #embeddedlinux #yocto

So #GNOME crashes when I close the Minecraft window now, cool

I tried using the crash report feature but after spending a while writing a detailed description of what was going on the problem reporting app just told me it's a duplicate, I think the duplicate detector is wrong but that doesn't matter, it didn't create the issue and instead deleted by description so that's cool, I'll definitely be reporting more bugs /s

#GNOME46 #fedora #fedora40

Does any #yocto / #openembedded / #linux / #embeddedlinux people know what could possibly cause this situation:

U-boot loads and starts the kernel no problem, the kernel goes through its normal boot process, everything looks fine, it finds all the partitions in flash, and then .. it does nothing

It's as if /sbin/init just hangs? Without even printing error messages?

I'm gonna try to get xxsh (https://github.com/mortie/xxsh) into the initrd somehow and explore but if anyone knows anything, do tell

Hey so I wrote a messagepack parser and serializer: https://github.com/mortie/msgstream #cpp #programming #messagepack

An interesting thing about #msgpack: its *type system* doesn't differentiate between unsigned and signed integers, but the *format* does, and I don't understand why.

The format makes space for 1-byte, 2-byte, 4-byte and 8-byte signed integers, *and* 1-byte, 2-byte, 4-byte and 8-byte unsigned integers, even though those options mostly overlap. To support the same range of integers, you'd just need the 4 signed integers and one 8-byte unsigned integer format. You could then add 5/6/7B integers!

Okay can someone just make a rustfmt for C++ already, I'm extremely tired of trying to configure broken formatters

I'm just trying to OR together some bits and #uncrustify is making that extremely difficult

If anyone knows how I can make this not absolutely terrible please do tell me because I'm out of ideas (and please don't tell me to use #clangformat, that's even worse)

#cpp #programming

Is there any detailed info on why that whole #nginx / #freenginx fork happened? I've tried to dig a little, and the only thing I can is that apparently F5 wanted to assign a CVE to an exploitable security bug in a feature in nginx but that Dounin didn't want the CVE assigned because Dounin thinks CVEs shouldn't be used for experimental features? That seems very very weird to me, surely there's more?

does any #CPP people know how to get #ClangFormat to stop doing this BS, it looks so horrible

#clang #llvm #programming #c

I guess I'll learn how to use #virtualenv and use Python 3.9? (or should I use #pyenv? #pyvenv? #venv? #virtualenvwrapper? #pipenv? Why are there so many of these things??)

So I noticed the lack of prominent #JSON5 parsers for C++, and made one of my own: https://github.com/mortie/json5cpp

The goal is to be simple, robust, and a very small dependency for projects which already use JsonCpp.

Thoughts?

#programming #cpp #json

Darn you #ClangFormat !

#llvm #cpp #clang

It seems literally impossible to get #ClangFormat to not align stuff.

Don't believe me? Take this piece of code and find a clang-format config which doesn't align stuff:

auto getWhateverStmt = db_.stmt(
"SELECT * FROM mytable "
"WHERE k = 2 "
"ORDER BY createdAt").bind();

Feel free to play with it in https://zed0.co.uk/clang-format-configurator/. I literally can not find a way to make it not produce something absolutely ridiculous where the SQL is pushed all the way to the right.

#llvm #cpp #c #programming