Bug #624
Test FK5 to Galactic coordinate conversion in test_GSky.py fails for Python 3.
| Status: | Closed | Start date: | 12/06/2012 | |
|---|---|---|---|---|
| Priority: | Normal | Due date: | ||
| Assigned To: |  Deil Christoph | % Done: | 0% | |
| Category: | - | |||
| Target version: | - | |||
| Duration: |
Description
The GammaLib unit test fail on Python 3 due to a problem in the FK5 to Galactic coordinate conversion code in test_GSky.py.
See https://cta-jenkins.irap.omp.eu/view/CTA%20Python/job/gammalib-check-python/129/
The error message is
basic_string::_S_construct null not valid
Recurrence
No recurrence.
History
#1
Updated by Deil Christoph over 11 years ago
Jürgen Knödlseder wrote:
The GammaLib unit test fail on Python 3 due to a problem in the FK5 to Galactic coordinate conversion code in test_GSky.py.
I’ll try to reproduce and fix.
See https://cta-jenkins.irap.omp.eu/view/CTA%20Python/job/gammalib-check-python/129/
The error message is
[...]
Perfect example for what we are discussing in other tickets: we need a way to run Jenkins on branches before merging, and we need Python tracebacks.
:-)
#2
Updated by Deil Christoph over 11 years ago
- Status changed from New to Closed
Pushed a fix directly to devel here:
https://cta-redmine.irap.omp.eu/projects/gammalib/repository/revisions/c51883ef397c99ea22ee47232987f228a1b3974b/diff/test/test_GSky.py
Let’s see if I get an email from Jenkins ...
#3
Updated by Knödlseder Jürgen over 11 years ago
Christoph Deil wrote:
Jürgen Knödlseder wrote:
The GammaLib unit test fail on Python 3 due to a problem in the FK5 to Galactic coordinate conversion code in test_GSky.py.
I’ll try to reproduce and fix.
See https://cta-jenkins.irap.omp.eu/view/CTA%20Python/job/gammalib-check-python/129/
The error message is
[...]Perfect example for what we are discussing in other tickets: we need a way to run Jenkins on branches before merging, and we need Python tracebacks.
:-)
I agree about the tracebacks.
I’m not sure that we need to run Jenkins on all branches, this would anyways be too time consuming. That’s the main purpose of having an extended system setup on the devel branch. These problems should exactly be discovered there.
#4
Updated by Knödlseder Jürgen over 11 years ago
Christoph Deil wrote:
Pushed a fix directly to
develhere:
https://cta-redmine.irap.omp.eu/projects/gammalib/repository/revisions/c51883ef397c99ea22ee47232987f228a1b3974b/diff/test/test_GSky.pyLet’s see if I get an email from Jenkins ...
It’s a nighty pipeline ... but for the sake of testing I’ll kick it off.
#5
Updated by Deil Christoph over 11 years ago
Jürgen Knödlseder wrote:
I’m not sure that we need to run Jenkins on all branches, this would anyways be too time consuming. That’s the main purpose of having an extended system setup on the
develbranch. These problems should exactly be discovered there.
Ideally I would say that issues are fixed during code review before merging into devel and once the merge happens it’s done.
But of course it’s also possible to have a workflow where test failures revealed by Jenkins are fixed in a second pull request or directly in devel.
#6
Updated by Knödlseder Jürgen over 11 years ago
So in fact you propose to add a 4th branch for core review?
For the moment we have:- master (last release of the package)
- release (branch to fix issues before release)
- devel (main development branch, under nightly CI control)
I again propose to more this discussion to the forum (otherwise we risk to have the Git workflow discussion spread over various issues).