On 10/07/2020 15:56, Andrew Sharff wrote:
On Fri, 10 Jul 2020, Paul Emsley wrote:
Thank you for the reply.
I hope that the recent commits you have made address the issues.
It seems to me that they are not going to be easy to diagnose,
especially our script visualise-geometry-coot, which you run and give
it an output directory from a BUSTER refinement. It invokes coot with
two scheme files and should load the pdb and mtz files in the
directory as well as opening dialog boxes to highlight poor geometry
etc. This one opens coot but neither the pdb or mtz file are loaded.
There are NO errors or warning messages, indeed no indication at all
of what is wrong.
The bug I fixed (hope that I've fixed (it works for me)) was for
handling the command line. I don't know why it had stopped working for
some people after all these years.
The second script, opens coot and loads pdb files and an mtz file, but
if you click on the "next" (Structure) box on the dialoguw box, then
coot crashes with a segmentation fault.
send me the core dump and we can make progress.
I would love to try today's revision - where can I download a MacOS
binary from?
I don't know. I don't make Mac OSX binaries - I don't have a computer to
do so.
Bill Scott used to make Mac OS X binaries, but it is a time-consuming
near thankless task. CCP-EM have a jhbuild-like script I think. I'd like
to give it a go at some stage.
You only have linux binaries on your website - we only see the
problems on MacOS.
I have never been able - well never even tried - to compile Coot from
source on a Mac!
I haven't tried for a while, but I would start with the build-it script.
Paul.
Andrew
>
On 07/07/2020 14:37, Andrew Sharff wrote:
Dear All,
Since moving to ccp4 7.1 (and Coot 0.9), we have been seeing several
issues on MacOS with some of our utility programs, which start Coot
with
both distributed and automatically generated Coot scheme
scripts.
I don't really follow what CCP4 do to Coot.
These issues range from Coot not acting on commands in the scheme
scripts
(and not reporting any errors or warning messages) to crashing with
segmentation faults.
I have recently made commits to address issues in both of these
areas, but without a core dump, I don't know if it's related.
These problems only affect MacOS - they all work perfectly on Linux
- and
only with the latest version of Coot (0.9). We have tested and see the
same behaviour on machines running MacOS 10.13, 10.14 and
10.15.
OK.
We have also tried the Coot 0.9 binary that is distributed in
ccp-em but
in this case, it crashes with a segmentation fault every time on
MacOS.
Colin has been in touch about a command line-related issue. I've
tried to fix it. Hideous archaic baroque C-code. Bleugh. No fun at all.
Has anyone else encountered any unexpected behaviour / probems running
Coot 0.9 with scheme input scripts and if so, do you have any
insight into
what the problems (and better still, solutions) are?
Maybe you could give today's reversion a bash
(efde451d59318f6f58c7f8295c93c86f764126e5)
Paul.
| Dr Andrew Sharff |
| Global Phasing Ltd., Cambridge UK |
| http://www.globalphasing.com
########################################################################
To unsubscribe from the COOT list, click the following link:
https://www.jiscmail.ac.uk/cgi-bin/WA-JISC.exe?SUBED1=COOT&A=1
This message was issued to members of www.jiscmail.ac.uk/COOT, a mailing list
hosted by www.jiscmail.ac.uk, terms & conditions are available at
https://www.jiscmail.ac.uk/policyandsecurity/
