On Sat, Aug 02, 2025 at 02:55:37PM +0000, Ihor Radchenko wrote: > <to...@tuxteam.de> writes: > > >> P.S. I am confused why multiple people are asking to disable cache. > > > > Perhaps because there is a need? > > >> Org > >> has been creating ltximg directory since forever and nobody ever asked > >> to "disable" that one. > > > > This might have several reasons: > > > > - Not everyone is doing LaTeX with Org. Those possibly won't ever > > see ltximg > > Then, setting org-element-cache-persistent will be good enough.
Until some other "caching client" creeps up. > > - Image directories aren't (strictly speaking) caches: Org has to > > have a file to be able to display an image > > - The Org persisten "cache" as it is now mushes everything in some > > directory "out there". Image files can be whithin a project directory > > structure (more akin to intermediate artifacts in a Make/build > > setup [1]. > > > I respect Org's decisions, but in this case I dislike them so much > > as to use That Other Crowbar described elsewhere in this list to > > get rid of it. > > Patches adding org-persist an ability to maintain caches in > project/current directory will be welcome. That's fine -- but note that I just offered *one* example. Playing whac-a-mole with every single reason someone doesn't want Emacs to dump uncontrolled stuff to the file system doesn't look (to me) like the right thing to do. Thus the crowbar :) I still don't see why a cache has to be persistent *beyond* the need to speed up things across sessions. For those not needing that speedup, it should be a no-brainer to disable it. Everything else has, to me, an "architecture smell". But that's probably me :) Cheers -- t
signature.asc
Description: PGP signature
