11:22 <farkuhar> braewoods: the pkgadd speedup you envisioned, by avoiding a second pass through the archive, has attracted at least one earlier attempt: https://lists.crux.nu/pipermail/crux-devel/2019-June/004107.html

11:22 <beerman> lol, thats actually braewoods himself ;)

11:25 <farkuhar> it's too bad that thread didn't attract more attention, then. it might have encouraged an earlier release of braewoods' rewrite.

11:29 <beerman> it's the norm, there is barely a crowd interested and capable enough, but you know that already

11:31 <braewoods> farkuhar: yea, that was something i had worked on. by caching the necessary data you could speed it up in exchange for more RAM usage relative to the size of the archive.

11:32 <braewoods> given how plentiful ram is today it doesn't seem like a bad idea to do such a tradeoff

11:35 <farkuhar> braewoods: the caching idea makes a lot of sense, and it doesn't require us to change the way binary packages are built. Although there might still be advantages to having a packing list at the beginning of every pkg.tar.gz

11:45 <braewoods> it's something that would have to be introduced later most likely.

11:45 <braewoods> We would need support for it in pkgadd first, etc.

11:46 <braewoods> probably would require a transition period where the original behavior works still

14:12 <beerman> jae

14:13 <beerman> meh.. jaeger did you turn off postgresql about a week ago? Oo

14:13 <beerman> or was that when var was swapping over?

14:19 <beerman> in any way, i just cleared another 14G of data

14:19 <beerman> well, /var/log/old/.