Luca Ferrari

> > Sounds good - a prototype of it will likely give a clearer indication of what is the best solution > > Ok, I think this will be done...

@T1t4m1un yes, sure, thanks!

@T1t4m1un thanks for the contribution, but I don't like the idea of using `printf` to print directly to the output. I agree that the output format is quite simple, but...

> Could you give me some examples of the format? I don't clearly understand the format you envisioned😶‍🌫️ See . I think the command should give this kind of output:...

I think it would be nice to have a floating tag `latest` (or something) that points to the latest stable release tag. In this case we can point the users...

I do agree in not using thread right here. Also, as a minor note, you should run crustify against the code.

Commit 189c5f7 silents all the warnings, and works with `uncrustify` 0.78. One problem I see, is that after a run of `uncrustify.sh` the multiline macros are changed as the diff...

Same issue happens in `pgagroal-admin`: ``` % pgagroal-admin user pgagroal-admin: unknown command or subcommand ``` while it should report something like *command user requires a subcommand*.

Yeah, open a new issue. At glance, I would remove `do_free` completely and `memcpy` the `password` into another `char*` so that `free` can be called freely (no pun intended). Seems...

> #### 1. Modify `parse_command` so it deals with `pgagroal_command_type` > > We could define each command as a struct containing the command string, the expected number of arguments, and...