Ville Voutilainen

In C++, class VolumeObserver : android::os::CallbackInterface // same as 'implements android.os.Handler.Callback' class ContentObserver : public android::database::ContentObserverInterface // same as 'extends android.database.ContentObserver I can Kiss The Sky.

Wheeeeee qt.io/blog/qt-jenny-…

Now we're seriously cooking. git.qt.io/vivoutil/libun…

@jfbastien Language! Such phrasing is hardly appropriate for pretty much any forum!

@CplusplusVille Sounds like Gentoo!

I had a dream: AInux • Packages are prompts • Install = generate + test + sandbox locally • Output = unique binary tuned to you No apt. Just prompts. Shipping tarballs? Shipping intent!

It's live godbolt.org/z/ME3EbnEsh

I can haz an implementation of wg21.link/p2034 at github.com/villevoutilain…

@ChShersh A very nice thing about C++ Ranges is that I don't have to be a fully-qualified FP person to use them.

Just learned about in C++ I'm impressed by what mainstream FP has achieved. It's also interesting to compare C++ and OCaml here because C++ managed to do some things better. 1. In C++, views are lazy out of the box. In OCaml, you need to wrap lists into Seq first. 2. In C++, you can pass pointer to member directly to filter or map by fields. In OCaml, a lambda is required. 3. In C++, you still get all the benefits of super fast code. 4. Pipe operator is EXACTLY like in Bash, so there's some cross-language consistency. OCaml still looks slightly more elegant in some places. Type inference requires less boilerplate, composable 'iter' API with 'filter' and 'map' instead of a separate for-loop, no imports, and automatic function currying. But I'm surprised that C++ turned out to be not so much bigger than idiomatic OCaml. The problem is that it's becoming harder to justify using FP languages, especially when you already have a huge legacy codebase in C++

It now has documentation and an example doc-snapshots.qt.io/qt6-6.10/qtjen…

I am in possession of a selfie with @pdimov2. His existence has been anecdotally confirmed by eye witness testimony. 😀 What, however, has not been established is evidence countering the claims that he is an artificial intelligence singularity, i.e. an AI smarter than humans. 🤣

@BarryRevzin The section title at #get_bases_classes_t-and-get_data_members_t" target="_blank" rel="nofollow noopener">brevzin.github.io/c++/2025/06/09… seems like a typo, should be get_base_classes_t instead of get_bases_classes_t?

It occurred to me recently that I could actually implement the Reflection TS on top of the C++26 reflection design (p2996). So I thought it would make an interesting #cpp blog post to compare the type-based to value-based models for reflection: brevzin.github.io/c++/2025/06/09…

..or rather Qt 6.12, because Other Things To Do Before That..

QtStdExec now supports cancellation. Which means that closing the funny test applications now stop the running async work, instead of just cleanly waiting for its completion. Woo hoo! It's feature complete, and needs only tests/examples/docs for inclusion into Qt (6.10 or 6.11).

Well here's another interesting toy onlinegdb.com

@jfbastien cmake has CMakeCache.txt, which contains both 'cmake' and 'ache'.

`make` and `cmake` have `ache` in common. Coincidence?

Run any Sender (and by extension, any wait of any QObject signal) as a stackful coroutine: #L223-238" target="_blank" rel="nofollow noopener">git.qt.io/vivoutil/libun…

@jfbastien Dude, we had .pro files earlier, and have since nuked them in favor of CMakeLists.txt files.

C++’s legacy is C. A hard aspect is its file-based build model. It tried to change with modules, but arguably isn’t there. Now, C++ is considering fixes to issues including security and will introduce an enterprise ready solution: - pro files - semipro files - free tier files

It's on Maven Central now: central.sonatype.com/artifact/org.q…

@LauraLoomer @elonmusk You do know that your speech on twitter isn't protected from twitter itself, right? All of the free speech protections protect you from the government, not from individual enterprises, so you have no particular protected right to say anything on twitter.

.@elonmusk why are you labeling 75% of the replies on my account as “probable spam”? This is an abuse of power & everyone in MAGA needs to call this behavior out. This is not what we voted for. This is totalitarian censorship and retaliatory behavior over protected free speech.

@jfbastien @PatriceRoy1 That's a curious definition of "nobody has bothered to do the work", when such work shipped in GCC 14. :P

@PatriceRoy1 @CplusplusVille There’s significant room for improvement of implementations without breaking ABI and without adding new language features. But nobody has bothered to do the work. Lots of talk though.

Fascinating. databasearchitects.blogspot.com/2024/12/c-exce…