.travis: Convert TRAVIS_COMMIT_RANGE base...head to base..head by wking · Pull Request #216 · opencontainers/runtime-spec

wking · 2015-10-06T17:30:10Z

Work around travis-ci/travis-ci#4596 until that is fixed upstream.
This avoids pulling in commits from the base tip that aren't reachable
from the head tip (e.g. if master has advanced since the PR branched
off, and the PR is against master). We only want to check commits
that are in the head branch but not in the base branch (more details
on the range syntax here).

Once the Travis bug does get fixed, the shell replacement will be a
no-op. So we don't have to worry about checks breaking once the bug
gets fixed, and can periodically poll the bug and remove the
workaround at out leisure after the fix.

Spun off from #182 now that #215 has landed.

vbatts · 2016-01-04T18:39:11Z

is this really an issue to us?

wking · 2016-01-04T18:56:24Z

On Mon, Jan 04, 2016 at 10:39:14AM -0800, Vincent Batts wrote:

is this really an issue to us?

It's only an issue if there are broken commits in the master branch.
We seem to have avoided that so far, with the only merged failure
being the (Travis hiccup?) [1,2]. So not a critical fix, but still
something that I think is worth landing.

vbatts · 2016-01-13T23:43:54Z

as this has not impacted us, and may not for some foreseeable future, I'm inclined to close it. It could always be reopened if the need be.

cyphar · 2016-07-22T17:00:10Z

Time to reopen I guess.

vbatts · 2016-07-25T18:54:43Z

@cyphar what is needing the reopening?

wking · 2016-07-25T19:00:28Z

On Mon, Jul 25, 2016 at 11:54:45AM -0700, Vincent Batts wrote:

@cyphar what is needing the reopening?

#520 was reporting errors in old master commits 1, but its
replacement #521 did not. So you can probably re-close this and kick
the can down the road again ;).

vbatts · 2016-07-25T19:02:12Z

😕

wking · 2016-07-25T19:07:08Z

On Mon, Jul 25, 2016 at 12:02:13PM -0700, Vincent Batts wrote:

😕

I still think this is a useful change, and could land now. You
(presumably) still think we should wait until we are impacted by not
landing it 1. We may have been impacted by it in #520 2, but
that's no longer open. So we're back to “no currently open PRs are
impacted by this not being merged”, and the last time we were there
you closed the PR 1.

vbatts · 2017-01-18T22:09:00Z

lets give this a try. could you rebase?

wking · 2017-01-18T23:25:53Z

On Wed, Jan 18, 2017 at 02:09:01PM -0800, Vincent Batts wrote: lets give this a try. could you rebase?

Rebased with d815fa9 → 40035db.

Work around travis-ci/travis-ci#4596 until that is fixed upstream [1]. This avoids pulling in commits from the base tip that aren't reachable from the head tip (e.g. if master has advanced since the PR branched off, and the PR is against master). We only want to check commits that are in the head branch but not in the base branch (more details on the range syntax in [2]). Once the Travis bug does get fixed, the shell replacement will be a no-op. So we don't have to worry about checks breaking once the bug gets fixed, and can periodically poll the bug and remove the workaround at out leisure after the fix. [1]: travis-ci/travis-ci#4596 [2]: http://git-scm.com/docs/gitrevisions#_specifying_ranges Signed-off-by: W. Trevor King <wking@tremily.us>

wking · 2017-01-19T05:04:18Z

I fixed up passing the environment variable through to git-validation and shifted the base of the branch back a few commits for testing with 40035db → d9b7bc3. That leaves Travis testing a history like: $ git log --graph --decorate --oneline origin/master wking/travis-test-branch-commits * 995c556 (HEAD) Merge d9b7bc3 into c368be6 |\ | * d9b7bc3 (wking/travis-test-branch-commits, origin/pr/216) .travis: Convert TRAVIS_COMMIT_RANGE base...head to base..head * | c368be6 (origin/master, origin/HEAD) Merge pull request #650 from wking/readme-header-nesting |\ \ | * | 200e86f (origin/pr/650) README: Consistent header nesting * | | dc0fa75 Merge pull request #655 from Mashimiao/schema-fix-with-latest-spec |\ \ \ | |_|/ |/| | | * | 5076439 (origin/pr/655) schema: fix items based on latest spec * | | 937ea7b Merge pull request #490 from wking/json-schema-validate-http-schema … … … Travis is setting the buggy three-dot form (c368be6...d9b7bc3 [1]), which includes all the commits not reachable from both commits: $ git log --oneline c368be6...d9b7bc3d9b7bc3 .travis: Convert TRAVIS_COMMIT_RANGE base...head to base..head c368be6 Merge pull request #650 from wking/readme-header-nesting dc0fa75 Merge pull request #655 from Mashimiao/schema-fix-with-latest-spec 5076439 schema: fix items based on latest spec 200e86f README: Consistent header nesting This PR fixes that so git-validation is called with the two-dot form, which includes only the unique-to-this-PR commit: $ git log --oneline c368be6..d9b7bc3d9b7bc3 .travis: Convert TRAVIS_COMMIT_RANGE base...head to base..head [1]: https://travis-ci.org/opencontainers/runtime-spec/jobs/193276777#L292

Only use the auto-ranging when Travis tells us what the range is. Use our EPOCH_TEST_COMMIT-based range in all other cases. ifdef is described in [1]. [1]: https://www.gnu.org/software/make/manual/html_node/Conditional-Syntax.html Signed-off-by: W. Trevor King <wking@tremily.us>

To make make debugging Travis environment issues more straightforward. Signed-off-by: W. Trevor King <wking@tremily.us>

wking · 2017-01-24T00:01:57Z

I've also brought over the new commits from opencontainers/image-spec#521, which should help with failures from local-branch PRs (e.g. here for #661).

vbatts · 2017-01-24T15:49:10Z

LGTM

crosbymichael · 2017-01-24T18:04:46Z

LGTM

Through 3297cd5 (Merge pull request opencontainers#216 from wking/travis-test-branch-commits, 2017-01-24). Signed-off-by: W. Trevor King <wking@tremily.us>

wking mentioned this pull request Jan 8, 2016
Makefile: add a target to run tests #295
Merged

vbatts closed this Jan 13, 2016

wking mentioned this pull request Jul 22, 2016
Add governance and release guidelines #520
Closed

cyphar reopened this Jul 22, 2016

wking mentioned this pull request Aug 24, 2016
Fix a grammar mistake in serialization.md and don't imply _ is valid in dns hostnames opencontainers/image-spec#163
Closed

RobDolinMS added this to the 1.0.0 milestone Jan 18, 2017

RobDolinMS assigned wking Jan 18, 2017

wking force-pushed the travis-test-branch-commits branch from d815fa9 to 40035dbCompare January 18, 2017 23:20

wking force-pushed the travis-test-branch-commits branch 2 times, most recently from a7d8845 to 8952a6bCompare January 19, 2017 04:46

wking force-pushed the travis-test-branch-commits branch from 8952a6b to d9b7bc3Compare January 19, 2017 04:49

wking mentioned this pull request Jan 20, 2017
specs-go/v1: use digest type on descriptor opencontainers/image-spec#514
Closed

wking added 2 commits January 23, 2017 15:59

.travis.yml: Print TRAVIS_* environment variables
550a73a
To make make debugging Travis environment issues more straightforward. Signed-off-by: W. Trevor King <wking@tremily.us>

crosbymichael merged commit 3297cd5 into opencontainers:masterJan 24, 2017

vbatts mentioned this pull request Jan 27, 2017
Bump version for 1.0.0-rc4 #672
Merged

wking deleted the travis-test-branch-commits branch January 27, 2017 21:07

wking mentioned this pull request Mar 21, 2017
Test merges again and return to using TRAVIS_COMMIT_RANGE vbatts/git-validation#17
Merged

Uh oh!

Conversation

wking commented Oct 6, 2015

Uh oh!

vbatts commented Jan 4, 2016

Uh oh!

wking commented Jan 4, 2016

Uh oh!

vbatts commented Jan 13, 2016

Uh oh!

cyphar commented Jul 22, 2016

Uh oh!

vbatts commented Jul 25, 2016

Uh oh!

wking commented Jul 25, 2016

Uh oh!

vbatts commented Jul 25, 2016

Uh oh!

wking commented Jul 25, 2016

Uh oh!

vbatts commented Jan 18, 2017

Uh oh!

wking commented Jan 18, 2017 via email

Uh oh!

wking commented Jan 19, 2017 via email

Uh oh!

wking commented Jan 24, 2017

Uh oh!

vbatts commented Jan 24, 2017• edited by caniszczyk Loading Uh oh! There was an error while loading. Please reload this page.

Uh oh!

Uh oh!

crosbymichael commented Jan 24, 2017• edited by caniszczyk Loading Uh oh! There was an error while loading. Please reload this page.

Uh oh!

Uh oh!

Reviewers

Assignees

Labels

Projects

Milestone

Development

Uh oh!

5 participants

vbatts commented Jan 24, 2017•
edited by caniszczyk
Loading

crosbymichael commented Jan 24, 2017•
edited by caniszczyk
Loading