For small real-time #audio thing i’m working for my own entertainment i stick to good old #C and #PortAudio.
Working on Rust would mostly getting into unsafe mode and cursing that there is no “first-class” way to check each and every dynamic memory allocation success. Finally, the audio libraries are quite bad when considering ones actually implemented with Rust (e.g. CPAL), not just bindings to e..g PortAudio.
So where I would not like to Rust bother, it would come to poke me with. a stick, and where I would need its help, it totally ignores the issue.
I’d like Rust more if people would stop describing it as a memory safe language.There is no such thing as memory safe language. There’s only languages with weaker and stronger checks/guarantees, and usually only for a subset of overall memory management.
In the case of Rust it can handle dangling references but ignores memory usage and out-of-memory conditions completely.
Changing user account name in #Linux-based operating systems is not hard, but can at least be inconvenient dance to do for various reasons. Just to give an example, sometimes root account needs to be enabled for password login temporarily so that home directory name can be changed without issues.
It would be super nice if this could be done similarly as with TPM chips and machine over keys (aka MOK managed with mokutil
). I.e. there would be a way to set out a request for username change for logged in local account and upon next reboot there would be a query for new account name, and this process would be taken care of renaming username, group and home directory.
User name and group name are trivial because they don’t really change any identifiers associated with the user but AFAIK usermod completely denies home directory name change for a logged in account.
@ljs apparently it has this:
-o FILE
Export all necessary information to FILE instead of opening the browser interface. If
FILE is "-", the data is written to standard output. See the examples section below
for some handy use cases.
meaning that probably the maintainer does not get upset of submitting patches enabling e.g. JSON output json-c, I’d suppose…