Are "hard links" important or a nice to have in a filesystem?
I'm trying to assess possible feature compromises for a #decentralized / #p2p based #filesystem that can be mounted on multiple machines, so any views on the usefulness of hard links with examples would be appreciated.
One such compromise here is merging of changes made from different devices, that will be much harder with hard linking.
Here is how use awk with a regular expression to match lines containing the word "nginx" and the first fireld must be "ii". In other words, it will show all installed nginx package names on your Debian or Ubuntu Linux:
@nixCraft awk and send are true powerhouse tools of the #posix world. Makes me want to revisit my bathroom technology tutorial series using #ObjectVoice.
Anyone familiar with Syncer, a virtual disk written in Rust?
It's a massive virtual disk with #POSIX interface that can be mounted locally using #FUSE.
Written in #Rust, the architecture and backend mean it could be adapted to a decentralised backend, something I looked into in the past and an revisiting.
I'm wondering if anyone else is interested in this potential, and if anyone knows the author or second contributor.
❝Today, thanks to Android and ChromeOS, Linux is an important end-user operating system. But, before Linux, there were important Unix desktops, although most of them never made it. …❞
@rnkn If it’s useful and portable to #Windows users as well I would prefer #Perl. Not everyone can or wants to install #WSL or some other compatibility layer.
Also, it is very difficult to avoid #bash-isms or dependencies on non-#POSIX tools creeping into a #shell script, especially if you accept others’ contributions. (“Everybody has #GNU coreutils!”) There are tools and tests to mitigate this, but then you might want to automate them… and #Perl has way better testing tools.
Maybe-contentious claim: for technically-oriented people, no user interface has ever surpassed the Unix shell.
I remember "discovering" it in 1987, 9 years after I'd started writing code in FORTRAN and BASIC. The old-timers showed me how I could do things in minutes that took weeks on other systems, like generate a full sorted concordance of a text in a single pipeline of shell commands.
(If I were ever forced to use a Mac, that's where I'd spend a lot of my time.)
When writing C++, I consistently find myself wanting to use std::string_view, because I want to take a view into a string. Then, as time goes on, I eventually want to pass that string to something which eventually has to interact with a #C (usually #POSIX) API. At that point, I need a 0-terminated string. string_view isn't zero terminated, so I revert back to using a const char * instead of a string_view.
Why oh why isn't there a std::c_string_view which contains a zero-terminated string
Today I learned how to split a variable with multiple lines for a Bash/ZSH script. I was using the previously mentioned mdb-query package, and I have a variable named tables with the different tables, separated by returns. When I fried to do for table in $tables; do echo $table; done, it did not split the lines.
I got the right output with
while IFS= read -r table; do $table; done <<< $tables
“Argument list too long” is such an archaic error.
Sorry, your computer can’t run this program, because somewhere there’s a buffer limited to one millionth of your RAM size. #posix
Watching some of the vids coming out of this year's Perl and Raku Conference. Ingy döt Net's talk on his work on Lingy and YAMLScript was pretty amazing.
Lingy is a #clojure hosted on #perl - Perl code can call Lingy code can call Perl code. The idea of having a Lingy VM with JIT is very appealing.
YAMLScript is a YAML-style dialect of Lingy (or any Clojure implementation), for those who don't like LISP style.
@bololacertus@overeducatedredneck@profoundlynerdy@fuzzix Put another way, checkbashisms exists because the only way to get #shell scripters to stick to strict #POSIX syntax and userland programs is to automatically reject deviations with extreme prejudice.
And please don’t get me started on #Python portability. Y’all can’t even settle on how to manage the virtual environments that every individual Python-based tool requires.
[meta] thanks for not camel casing your post titles
I wasn’t sure how to express my gratitude....