History log of /seL4-l4v-master/HOL4/.travis.yml
Revision Date Author Comments
# c4ffbd66 17-Jul-2020 Thibault Gauthier <email@thibaultgauthier.fr>

travis: removing 'opam config env'


# 1d9fa47d 23-Sep-2019 Michael Norrish <Michael.Norrish@nicta.com.au>

Make early Moscow ML regression tests run on Travis


# 818ca782 07-May-2019 Michael Norrish <Michael.Norrish@nicta.com.au>

Give up on sending Travis notifications to hol-builds

There doesn't seem to be any easy way of getting this to work. I
believe it will now e-mail authors of commits instead, which isn't so
bad.


# 9bb1d3c9 10-Mar-2019 Michael Norrish <Michael.Norrish@nicta.com.au>

Turn off Travis CI notifications to the IRC #hol channel


# fcb9856c 19-Feb-2019 Michael Norrish <Michael.Norrish@nicta.com.au>

Make even shorter build seq. file for OpenTheory builds on Travis


# f5f56397 17-Feb-2019 Michael Norrish <Michael.Norrish@nicta.com.au>

Get Travis to test (once) with Moscow ML as well as Poly/ML


# e8f805e9 17-Feb-2019 Michael Norrish <Michael.Norrish@nicta.com.au>

Give Travis' opentheory build less to do

I hope this will stop it from exceeding Travis's cut-off limits for
long builds.


# f4477a5f 17-Feb-2019 Michael Norrish <Michael.Norrish@nicta.com.au>

Try to get Travis to build opentheory tool and test --otknl


# 3f4ad1c1 21-Sep-2018 Lorenz Leutgeb <lorenz@leutgeb.xyz>

Rewrite before_install.sh as install-poly.sh

Promote the contents of before_install.sh to be the de facto way of
setting up PolyML.

To make arguments for the script less confusing, use "T" and "F"
instead of empty/non-empty environment variables.
Also rename the environment arguments to be prefixed by POLY instead
of suffixed as I consider this more clear.

The renaming is in conformance with other {verb}-{noun} constructs in
/developers.


# 8cc9b9ac 11-Jan-2018 Michael Norrish <Michael.Norrish@nicta.com.au>

Don't install default apt (old/stale) OCaml (use opam's instead)


# f889b2e3 11-Jan-2018 Michael Norrish <Michael.Norrish@nicta.com.au>

Still not seeing holyhammer building on Travis

This attempt runs the opam command to update the PATH in the current
shell.


# 8ddbb9c1 11-Jan-2018 Michael Norrish <Michael.Norrish@nicta.com.au>

Remove c++ package from Travis tests (given 423c04e)


# ea70d489 11-Jan-2018 Michael Norrish <Michael.Norrish@nicta.com.au>

Another attempt at getting opam on Travis


# 86828584 11-Jan-2018 Michael Norrish <Michael.Norrish@nicta.com.au>

Seems like package opam doesn't automatically bring in package ocaml


# f12e4923 11-Jan-2018 Michael Norrish <Michael.Norrish@nicta.com.au>

Try to set up Travis environment to test holyhammer

Need both ocamlfind and ocaml's num package, and the straightforward
way to do this seems to be to use the opam machinery.


# 4629274d 30-Jan-2017 Michael Norrish <Michael.Norrish@nicta.com.au>

Remove a redundant Travis CI test

Was testing the same sequence with selftest on and off; given that the
former tests a superset of the latter, there doesn't seem much point
in having the latter chewing up resources.


# 1a60306c 30-Jan-2017 Michael Norrish <Michael.Norrish@nicta.com.au>

Fix regression whereby git-poly wasn't allowed to fail in CI test


# 4b4dbe20 29-Jan-2017 Michael Norrish <Michael.Norrish@nicta.com.au>

Fix another .travis.yml error


# 5e43b435 29-Jan-2017 Michael Norrish <Michael.Norrish@nicta.com.au>

Remove some Unicode that crept into .travis.yml file

This broke most of the CI tests


# b8f40d74 28-Jan-2017 Michael Norrish <Michael.Norrish@nicta.com.au>

Incorporate new c/line option-handling into build

The build in relocation option doesn't do anything at the moment.


# 5cf1fbc7 10-Nov-2016 Michael Norrish <Michael.Norrish@nicta.com.au>

Include a new Travis build

Mostly so that comparison (of timing at least) is possible between the
git poly option and the default 5.6.


# 37197c97 17-Sep-2016 Michael Norrish <Michael.Norrish@nicta.com.au>

Attempt to allow git-based poly CI tests to fail


# 9a2db1fe 26-Apr-2016 Michael Norrish <Michael.Norrish@nicta.com.au>

Allow OSX Travis builds to fail

The virtual OSX environment seems so flakey at the moment (possibly in
conjunction with PolyML 5.6 bugs) that it doesn't seem worth having
failures there affect final verdict.


# 8b415e57 08-Jan-2016 Michael Norrish <Michael.Norrish@nicta.com.au>

Delete OSX git Poly test on Travis

As per http://permalink.gmane.org/gmane.comp.lang.ml.polyml.general/1313
the errors seem to be function of the Travis setup rather than our code.


# 338aa706 06-Jan-2016 Michael Norrish <Michael.Norrish@nicta.com.au>

Travis to check with latest git version of poly


# 25b661b2 17-Nov-2015 Michael Norrish <Michael.Norrish@nicta.com.au>

Turn off sudo: none in Travis builds

Though recommended this seems to be causing our builds to get killed by
the CI infrastructure.


# 7953d7b6 17-Nov-2015 Michael Norrish <Michael.Norrish@nicta.com.au>

Specify apt-based packages in .travis.yml file

Don't attempt to install them separately as this requires sudo commands.


# 2411d871 17-Nov-2015 Michael Norrish <Michael.Norrish@nicta.com.au>

Fix .travis.yml syntax error


# ca445e54 17-Nov-2015 Michael Norrish <Michael.Norrish@nicta.com.au>

Try to use Travis' new "container-based" tech

As per page at

http://docs.travis-ci.com/user/migrating-from-legacy/


# da38fc16 04-Nov-2015 Michael Norrish <michael.norrish@nicta.com.au>

More tweaks for Travis on OSX


# 6540e324 04-Nov-2015 Michael Norrish <michael.norrish@nicta.com.au>

Test Travis' support for tests on OSX


# 09b179ba 13-Jan-2014 Michael Norrish <michael.norrish@nicta.com.au>

s/chat/irc/ in Travis notification config


# 2f742a8a 13-Jan-2014 Michael Norrish <michael.norrish@nicta.com.au>

Try to get Travis's IRC notifications working (correct syntax helps)


# 3e9a2de5 12-Jan-2014 Michael Norrish <michael.norrish@nicta.com.au>

Tweak/experiment-with the way Travis sends build notifications.


# 4226645c 16-Dec-2013 Michael Norrish <michael.norrish@nicta.com.au>

Travis: try building with a fresh svn checkout of poly.


# 8f17909c 24-Nov-2013 Michael Norrish <michael.norrish@nicta.com.au>

Make Travis build core system in -expk as well as -stdknl.

Also shortened the "built with Poly in /usr/bin" test-case. If this
can get as far as building bossLib, then there's no reason to imagine
it failing later on. And why stress the free service at Travis unduly?


# df422661 24-Nov-2013 Michael Norrish <michael.norrish@nicta.com.au>

Travis: try to capture some more examples in our automatic builds


# 56111c61 19-Nov-2013 Michael Norrish <michael.norrish@nicta.com.au>

Travis: test expk as well as standard kernel.


# 372fa511 19-Nov-2013 Michael Norrish <michael.norrish@nicta.com.au>

Travis: try to get some examples tested

Do this by not building some theories under src, using a custom build
sequence.


# 267409fa 18-Nov-2013 Michael Norrish <michael.norrish@nicta.com.au>

Travis: selftest is taking too long; remove it for the moment


# 49a0cd95 17-Nov-2013 Michael Norrish <michael.norrish@nicta.com.au>

Travis: another attempt to build Poly in both user- and root-land.


# 031f0009 17-Nov-2013 Michael Norrish <michael.norrish@nicta.com.au>

Travis: test PolyML installed as root and non-root both


# d4b804bd 17-Nov-2013 Michael Norrish <michael.norrish@nicta.com.au>

Try varying environment variables to get multiple Travis test instances.


# 6c269132 17-Nov-2013 Michael Norrish <michael.norrish@nicta.com.au>

See if selftest 1 will go through before Travis times out the build


# a786e29a 17-Nov-2013 Michael Norrish <michael.norrish@nicta.com.au>

Travis: set LD_LIBRARY_PATH variable.


# a8d93ead 17-Nov-2013 Michael Norrish <michael.norrish@nicta.com.au>

Correct typo in .travis.yml


# 3ca03917 17-Nov-2013 Michael Norrish <michael.norrish@nicta.com.au>

Experimenting with Travis integration (see travis-ci.org)