On 02-02-2017 barjac reported on the #geda IRC channel:
<barjac> Hi. Building 4.0.0 "/bin/sh: ../depcomp: No such file or directory" Any ideas?
<barjac> http://mtf.no-ip.co.uk/pub/linux/barjac/distrib/cauldron/x86_64/log/pcb-4.0.0-0.3.mga6.src.rpm/build.0.20170202162038.log
<barjac> It is in the tarball, but checking chroot when build fails it's missing :\
<barjac> This is pcb-4.0.0 from release tarball
<barjac> Finally built with --disable-dependency-tracking
<bert> barjac: I am in the process of 1) mobilizing the pcb developer team to solve this issue, 2) to add a bug report, 3) to look into the problem that is causing this
<bert> bug
<barjac> bert: Thanks! :)
<bert> I have a meeting planned next weekend with one of the key-developers during FOSDEM in Bruxelles, I will put this on our agendas
<barjac> bert: That's great - Say hello to any Mageia people you meet there - no stand this year sadly.
<bert> barjac: ad 2) https://bugs.launchpad.net/pcb/+bug/1661348
<bert> barjac: ad 1) email is sent so I'm going on with 3)
<barjac> bert: I did try using autoreconf -vif but then hit 3 missing ..../gs/*.in files IIRC so that failed.
More from the #geda IRC channel:
barjac> bert: The last test fails: Test: MinMaskGap. (pcb:16968): Gtk-WARNING **: cannot open display:
<barjac> bert: ../../.
<barjac> bert 1. Is this expected in a build chroot? 2. If so how can one test be skipped?
<dan> we probably need to find a way to make that test conditional on there being a display available
<barjac> bert: For now I have just used: "%make check ||:" to stop any test failures breaking the package build.
<barjac> dan: Yes that would be good ;)
<dan> I wonder if it is enough to check for $DISPLAY or if there is a more complete check that $DISPLAY is set and functional
<barjac> dan: BTW what is the small dialog top right-ish of screen with _._|_._ etc in it?
<dan> coordinates or measurements probably. Not in front of pcb at the moment
<barjac> dan: I'm not the one to ask - but I see the same issue in some gnuradio tests.
<barjac> dan: Ah OK that makes sense it looked like a corrupted graphic with no .pcb loaded :)
<dan> try opening the examples/LED.pcb file
<dan> anyway, gotta run
<barjac> dan: OK - it's not dims - it's left of them - has r-.- phi-.- etc. - bye thanks
<dan> dimensions while drawing a trace
<barjac> dan: Thanks