[Date Prev][Date Next][Thread Prev][Thread Next][Date Index][Thread Index]
Re: Using 'libfaketime' for reproducible builds
From: |
Jonas Hahnfeld |
Subject: |
Re: Using 'libfaketime' for reproducible builds |
Date: |
Sun, 27 Dec 2020 21:58:44 +0100 |
User-agent: |
Evolution 3.38.2 |
Am Sonntag, dem 27.12.2020 um 18:02 +0100 schrieb Werner LEMBERG:
> Before investing more time into it I wonder whether the use of
> 'libfaketime' would be a valid solution for creating reproducible
> builds.
My 2 cents: The widely accepted solution is SOURCE_DATE_EPOCH and if
there is anything in LilyPond itself that inserts unreproducible data
(into PostScript code), that should be fixed. Intercepting syscalls (or
whatever the library does, I didn't check) doesn't sound like the right
approach outside of testing reproducibility.
The larger "issue" with this topic seems to be LilyPond's dependencies,
in particular Ghostscript. A contribution to add support for above
variable was closed as WONTFIX:
https://bugs.ghostscript.com/show_bug.cgi?id=696765 I think that's a
pity, but nothing we can change as a "consumer" of library functions.
Jonas
signature.asc
Description: This is a digitally signed message part