Re: Conflict between R 4.5 and some (mostly r-bioc 3.20) R packages

["Rebecca N. Palmer" <rebecca_palmer@zoho.com>]

r-bioc-graph and r-bioc-iranges are both segfaults attempting to access 
an empty object, and are known to upstream.  (I'm not sure why they were 
only exposed by R 4.5; they look like they're bugs independent of R 
version.)

I have fixed versions of both in Salsa (r4p5test branches), using the 
upstream fix for r-bioc-graph, but my own for r-bioc-iranges because the 
upstream fix was a large rewrite that might not be allowed at this point 
in the freeze.

As previously noted, it is likely (but untested) that the r-bioc-iranges 
fix also fixes at least r-bioc-delayedmatrixstats, and maybe also 
r-bioc-pwalign and r-bioc-shortread.