Optimising MConfig

[psibi]

We currently have defined MConfig which is bunch of IORefs. But looking at the code - it seems that it's not really required. What we are doing currently can be just done via normal records and pure function. Is there any specific reason we are using IORef here ?

Again, I would like to refactor this - but would like to know before working if there is any specific reason for using around 15 plus IORef for doing it ?

[slotThe]

Looking at the code I get the same impression as you---the IORefs aren't really needed here. I would be in favour of refactoring this.

[jaor]

On Sat, Jun 06 2020, slotThe wrote: Looking at the code I get the same impression as you---the IORefs aren't really needed here. I would be in favour of refactoring this.

+1 (maybe there was some plan of making xmobar auto-reconfigure itself, maybe via a signal, or monitoring its config, and this was related, but i don't see an immediate reason either)