Front page | perl.perl5.porters |
Postings from July 2018
Re: [perl #133311] BBC GRANTM/Encoding-FixLatin-1.04.tar.gz
From: Karl Williamson
July 1, 2018 01:42
Re: [perl #133311] BBC GRANTM/Encoding-FixLatin-1.04.tar.gz
Message ID: 27826_1530409314_5B38315D_27826_11_1_d57b168cfirstname.lastname@example.org
On 06/30/2018 11:15 AM, Andreas Koenig wrote:
>>>>>> On Sat, 30 Jun 2018 08:56:01 -0600, Karl Williamson <email@example.com> said:
> > On 06/29/2018 12:12 AM, Andreas Koenig wrote:
> >>>>>>> On Thu, 28 Jun 2018 21:02:19 -0700, "James E Keenan via RT" <firstname.lastname@example.org> said:
> >> > I cannot reproduce this, at least in this typical case:
> >> Please note that the report already has a commit ID.
> > But that wasn't Jim's question.
> Jim's posting revealed that he tried with some 5.28.0. And since I was
> about to rush out the door I sent a quick hint that at least one of you
> two should understand. You know, bleadperl has since moved on to 5.29.0.
I did not pick up on this hint, nor that Jim hadn't used a version of
blead that is broken. I can understand how, in your haste, you didn't
say clearly what you intended. Just about everybody says things that
are unclear, even when unhurried. Perhaps you can understand how
someone, like me, might have missed your intent.
> > We can't fix it if we can't reproduce it.
> Are you saying you, Karl, have tried yourself and could not reproduce?
> Or are you saying, if Jim could not, then you won't even try yourself?
> If the latter, then I urge you to show some effort yourself. If the
> former, then please provide the details that may lead to interesting
Your above statement is troubling to me. You give two options,
apparently assuming that one or the other must be true, and if the
second is true, you tentatively are chiding me for being, for lack of a
better word, "lazy".
But in fact neither statement is true. I have learned to try to be
careful to realize that often there are other possibilities I haven't
considered in any given situation. That makes me a better programmer,
and happier in life.
When I used the word "We", I was referring to the project as a whole.
If your earlier email had been clearer, I would have realized that Jim
hadn't tested on a broken blead. But when I wrote the sentence you
quoted, I believed that Jim had tried and failed to reproduce it on a
broken blead. That made it seem quite possible that this problem might
surface only under some particular combination of Configure options. It
would be a waste of my time to try all combinations when someone else
already should have at their finger tips the correct combination. I
could spend that time on more productive uses.
Almost certainly the reason this ticket didn't go through the first time
is that you did fail to include the output of 'perl -V'. That's been
the experience of many posters.
Given that I thought this bug didn't always reproduce, I waited a couple
of days to see what further communication there was. When there was
none, I asked for the perl -V output.
Your statement is especially troubling since I wrote to you privately
when I first saw the ticket, thanking you for finding the bug, and
telling you I would work on it this weekend.
Did you not get this email?
Did you forget you got this email?
Did you not believe what I said?
Is there another possibility I haven't considered?
I am unaware of blowing off any ticket you have filed that concerned an
area of the core that I feel responsible for. If there are such then
I'm sorry; please let me know which so that they can be addressed in the
open, and I can try to regain your confidence in my willingness to fix
what I break.
If there aren't any, then I'm sorry for you, because that means your
life is unnecessarily negative. Hanlon's razor is a recipe for happier
Let me sum up. If you had used 'perl -V' initially, it would ultimately
have saved you time, Jim time, me time, and everyone reading this time.
Same for if you had been clearer in your response to Jim. These are
both easy mistakes to make, and are understandable and forgivable. We
all do things like this on a regular basis. But in this case, it wasn't
me who was failing to exert effort. And it is counter productive to the
project to tentatively chide someone, as you did me.
The bug, BTW, is fixed in blead. Thank you for finding and reporting
it. I'll close the ticket at some point after the systems to do so come
> > And the output of perl -V will tell us the particular combination of
> > Configure options (and less likely, platform) that led to the
> > problem.
> I have reproduced with 8 different perl configurations. Here is the
> 'perl -V' from one of them:
> Summary of my perl5 (revision 5 version 29 subversion 0) configuration:
> Commit id: a74bb78e4469c9f5ea806b57b155df6265d07975
> uname='linux k93msid 4.14.0-3-amd64 #1 smp debian 4.14.12-2 (2018-01-06) x86_64 gnulinux '
> config_args='-Dprefix=/home/sand/src/perl/repoperls/installed-perls/host/k93msid/v5.28.0-40-ga74bb78e44/da1c -Dmyhostname=k93msid -Dinstallusrbinperl=n -Uversiononly -Dusedevel -des -Ui_db -Dlibswanted=cl pthread socket inet nsl gdbm dbm malloc dl ld sun m crypt sec util c cposix posix ucb BSD gdbm_compat -Uuseithreads -Uuselongdouble -DEBUGGING=-g'
> ccflags ='-fwrapv -fno-strict-aliasing -pipe -fstack-protector-strong -I/usr/local/include -D_LARGEFILE_SOURCE -D_FILE_OFFSET_BITS=64 -D_FORTIFY_SOURCE=2'
> optimize='-O2 -g'
> cppflags='-fwrapv -fno-strict-aliasing -pipe -fstack-protector-strong -I/usr/local/include'
> Linker and Libraries:
> ldflags =' -fstack-protector-strong -L/usr/local/lib'
> libpth=/usr/local/lib /usr/lib/gcc/x86_64-linux-gnu/7/include-fixed /usr/include/x86_64-linux-gnu /usr/lib /lib/x86_64-linux-gnu /lib/../lib /usr/lib/x86_64-linux-gnu /usr/lib/../lib /lib
> libs=-lpthread -lnsl -ldl -lm -lcrypt -lutil -lc
> perllibs=-lpthread -lnsl -ldl -lm -lcrypt -lutil -lc
> Dynamic Linking:
> lddlflags='-shared -O2 -g -L/usr/local/lib -fstack-protector-strong'
> Characteristics of this binary (from libperl):
> Compile-time options:
> Built under linux
> Compiled at Jun 28 2018 18:44:58