Flod - Arch 2016

[Thomas Lauf]

Hello!

First: Nice work on flod! :)

I see the tests for totals.py failing on Arch 2016. What version of
python is installed there? Is there any more output for these test there?

How are the numbers caclulated? I see "798/805" but there are only 3
failed tests and 1 compiler warning listed and when I run the tests
locally I get

Passed: 806
Failed: 3
Unexpected successes: 0
Skipped: 0
Expected failures: 0

which would let me expect something like "806/809".

Regards,
Thomas

--
Dr. Thomas Lauf * thoma...@tngtech.com * +49 174 3180 086
TNG Technology Consulting GmbH, Betastr. 13a, 85774 Unterföhring
Geschäftsführer: Henrik Klagges, Christoph Stock, Dr. Robert Dahlke
Sitz: Unterföhring * Amtsgericht München, HRB 135082

[Paul Beckingham]

Hi Thomas,

[1] The Arch 2016 satellite (sputnik4.tasktools.org) is running:

[paul@sputnik4 ~]$ python --version

Python 2.7.5

[2] There is not output remaining on that arch machine.  A transcript of the whole process gets sent back to flod (on sitka.tasktools.org), so we have it, but do not currently expose it.  The plan is to provide access in two ways:

a) To put up a web server where the docroot points to the flod results directory. This will let us all navigate the full transcripts or all builds.

b) To put our keys on all the build machines, so we can ssh there and debug/build/test on the actual build machine.

Fredde, can we get everyone’s keys on all machines?

[3] Let’s investigate.  Walking through it, the latest timew commit is 872499a5, so let’s check the test numbers for Fedora 25 for that commit, and see where the problem is.  Looking the build transcript for that commit, for fedora, we see:

$ cd /opt/flod2/central

$ tail -14 $(grep -l fedora $(grep -l 872499a5 result/archive/*)) | head -6

Passed:                           798

Failed:                             3

Unexpected successes:               0

Skipped:                            0

Expected failures:                  0

Runtime:                         8.61 seconds

So there we see right away that you are right, and flod is wrong.  But is it inherently wrong, or is it being deceived?  There are only three failures, zero skips. The next step is to see what the digest event looks like, which is a summary of the transcript.  We expect it to list warnings, error, and test failures.

$ cat $(grep -l fedora $(grep -l 872499a5 digest/archive/*))

change.repository: /opt/gitea/repositories/tm/timew.git

change.timestamp: 1492033560

digest.testsExpected: 805

digest.testsPassed: 798

digest.testsFailed: 3

project.ignore: README*,*.md

build.duration: 37.984548

change.project: timew

digest.testsSkipped: 0

project.job: {bash}{nl}{nl}set -x{nl}git clone --recursive -b 1.1.0 https://git.tasktools.org/TM/timew.git timew.git{nl}cd timew.git{nl}git reset --hard 872499a54e234174ec7dea7be20cda57c23491b8{nl}cmake .{nl}make{nl}make test{nl}

platform.description: Fedora 25

change.commit: 872499a54e234174ec7dea7be20cda57c23491b8

change.author: Paul Beckingham

platform.dependencies: cmake,git,build-essential,gnutls-dev,uuid-dev,python

digest.buildErrors: 1

digest.funky: 1

platform.host: sputnik2.tasktools.org

platform.scratch: /tmp

project.repository: https://git.tasktools.org/TM/timew.git

platform.details: 64-bit;GCC 6.3.1;CMake 3.6.2

digest.buildWarnings: 1

change.branch: 1.1.0

platform.name: fedora25

digest.clean: 0

/tmp/timew.754b08fc64836ed0c8eaaff65155fd127136161e/timew.git/src/validate.cpp:162:13: warning: ‘void warnOnNewTag(const Rules&, Database&, const Interval&)’ defined but not used [-Wunused-function]

not ok 3 - move.t: Move an interval over another with :adjust

not ok 1 - tags.t: Test that tags command filtering excludes tags that are outside the filter range

not ok 6 - start.t: Test timed start

# ERROR: CommandError on file /tmp/timew.872499a54e234174ec7dea7be20cda57c23491b8/timew.git/test/start.t line 69 in test_timed_start: 'self.t("start 10:30am")':

The problem appears to be that it thinks there should be 805 tests.  That’s not right.  Where did that number come from?

$ grep '^1..' $(grep -l fedora $(grep -l 872499a5 result/archive/*)) | cut -c4-

129

96

101

261

39

7

6

4

3

1

4

3

3

7

7

1

1

3

5

6

20

4

3

5

14

36

2

2

5

2

1

3

2

7

12

Those add up to 805.  That means some timew test script is emitting “1..N” as the first line of output, and it should be “1..N-2”.  That should be easy to find.  It was, and here’s the problem:

# rules.t

1..6

ok 1 - Rules set string, get string

ok 2 - Rules set string, get string

terminate called after throwing an instance of 'std::__cxx11::basic_string<char, std::char_traits<char>, std::allocator<char> >'

So not a flod problem this time but a test crash.  I’ll fix that.  Well spotted Thomas!

P

--
You received this message because you are subscribed to the Google Groups "taskwarrior-dev" group.
To unsubscribe from this group and stop receiving emails from it, send an email to taskwarrior-d...@googlegroups.com.
For more options, visit https://groups.google.com/d/optout.

[Federico Hernandez]

1) That was CentOs. Arch on sputnik3 is 3.5.2.

2) I will prepare something.

To unsubscribe from this group and stop receiving emails from it, send an email to taskwarrior-dev+unsubscribe@googlegroups.com.

For more options, visit https://groups.google.com/d/optout.

--
You received this message because you are subscribed to the Google Groups "taskwarrior-dev" group.

To unsubscribe from this group and stop receiving emails from it, send an email to taskwarrior-dev+unsubscribe@googlegroups.com.