On 6 November 2015 at 23:35, Vincent Bernat wrote:
|  ❦  6 novembre 2015 14:03 -0600, Dirk Eddelbuettel <e...@debian.org> :
| > One of my packages which still didn't build reproduciby is littler -- for
| > which I am upstream. I rewrote the build process, and even though it 
| > a small binary (which embeds R for use in #! scripts etc) it now ships as an
| > R package on CRAN.  Which all build reproducibly.
| >
| > Yet I have this:
| >    https://reproducible.debian.net/rb-pkg/unstable/amd64/littler.html
| >
| > And I don't understand the 'dbgsym' part.  What turns that on? How can I 
| > it off?
| See man dh_strip. However, the correct action would be to find why there
| is a difference in the debug symbols. The difference seems to stem from

I was thinking that too but didn't test.  As an R package, it installs into a
'tree' below /usr/lib/R/site-library/$PACKAGE but I also copy the resulting
binary to /usr/bin/r. I think I need to strip that.

Will the weird dbgsym then be gone?

| the way some strings are sorted (see the end of the differences). They
| come from the r-base package. I don't know why they are embedded into
| debug symbols.



http://dirk.eddelbuettel.com | @eddelbuettel | e...@debian.org

Reproducible-builds mailing list

Reply via email to