Re: Firefox stability?

["John D. Baker" <jdbaker%mylinuxisp.com@localhost>]

On Tue, 10 Jan 2017, John D. Baker wrote:

> I've done just this.  On my (relatively) fast built/test system, I
> removed all packages (except sudo) and cleaned up stragglers manually so
> there was no pkgsrc-2016Q4 residue.
> 
> I then checked out a parallel pkgsrc tree for pkgsrc-2016Q3, but rolled
> "devel/nspr", "devel/nss", and "www/firefox" forward to HEAD.
> 
> I then built "www/firefox" (firefox-50.1.0) with all other dependencies
> (except nspr and nss) built from their pkgsrc-2016Q3 versions.
> 
> In this configuration, engaging the print function does not hang firefox.
> 
> Now, to update the long list of dependencies and find out which one is
> at fault.

Since I was resigned to nuke everything and start over anyway, I decided
to double-check by rebuilding everything from scratch using pkgsrc-2016Q4.

The resulting firefox-50.1.0 does not hang when engaging the print function.

So the problem was apparently some local glitch in a dependency that
caused firefox to misbehave.  I'll blame the squirrels.  I don't recall
a power outage during the run of 'pkg_rolling-replace' that eventually
produced my faulty firefox, but I've slept since then.

The build/test machine is not on a UPS, so it's entirely possible that
a power-outage at just the wrong time left a glitch in a package.  I
had thought that difficult since my WRKOBJDIR is on tmpfs which should
be completely nuked by power loss.  My PACKAGES directory is on spinning
rust (RAID-5 via amr(4) SCSI RAID controller) using FFSv2 and mounted with
"-olog,nodev,nosuid".

Now to rebuild the rest of my usual complement.

-- 
|/"\ John D. Baker, KN5UKS               NetBSD     Darwin/MacOS X
|\ / jdbaker[snail]mylinuxisp[flyspeck]com    OpenBSD            FreeBSD
| X  No HTML/proprietary data in email.   BSD just sits there and works!
|/ \ GPGkeyID:  D703 4A7E 479F 63F8 D3F4  BD99 9572 8F23 E4AD 1645