Greetings to all,

     This project has great interest to me for many reasons, but mostly
because I always find myself coming back to Debian, and, based on my 30+
years of sysprog/sysadmin/development experience, systemd simply represents
bad design in it's purest form, or just maybe, what SCO couldn't do,
systemd will.

      Having started my career on a mainframe in the early 80s, I quickly
noted the precise delineation between functional stacks, and within those
stacks, the intentional lack of dependency upon the presence of another
stack. This I call designing for flexibility and freedom of configuration.
systemd is the opposite; the mainframe equivalent of cramming RACF(security
stack), TSO/ISPF(ui stack/not quite a desktop), VTAM(sna comm stack),
TCPIP(ip comm stack), JES/JES2(job entry stack/init equiv), the MVS logger,
and other sundry items into one crude unintelligible, codependent wad.

     Not hate speech, science. Everyone understands the value of modular
and objectified approaches in system design, breaking down larger processes
into generic, reusable components with well-defined APIs, so how is it that
systemd has permeated the community at large?

     I'll stop there before I get a serious rant on.

     Regarding the Perl thang for Matteo, looks to me there's some pretty
strict REGEXs in that perl script which parses the output from eu-readelf.
Not sure if you can obtain a diff between output for a file,un-"optimized"
vs "optimized", but that would prolly reveal where the fix is needed.

Devuan? I like it, though I keep catching myself saying "Dev-Juan" :/

SWS