summaryrefslogtreecommitdiff
path: root/regress
AgeCommit message (Collapse)Author
2010-01-05...and this one tooMarc Espie
2010-01-05regress checks tend to work better when you don't forget the output...Marc Espie
2010-01-04add a script that I use before ever release to check that ourTheo de Raadt
edge conditions for disklabel -A are going to be acceptable ok krw
2010-01-03ouch, bad loopMarc Espie
2010-01-03Use CHECKx macro and enable setsockopt tests.Federico G. Schwindt
2010-01-02Check that end of .It does not restore .Sm stateIngo Schwarze
to the value it had before entering .It; fixed by kristaps@ in release 1.9.15.
2010-01-01When a .Fl macro without an argument is followed by text,Ingo Schwarze
the space after the dash appears to be missing. Needs fixing. Test triggered in a discussion with kristaps@.
2010-01-01check for most update situationsMarc Espie
2010-01-01make sure there's a MAINTAINER in thereMarc Espie
2010-01-01don't use -cMarc Espie
2009-12-31and yet another library check errorMarc Espie
2009-12-31remove all noise, have myadd display a simplified version of what it's doingMarc Espie
2009-12-31add some lib report checks where the error messages are currently very bad.Marc Espie
2009-12-31simplify and rationalize namesMarc Espie
2009-12-31check for infinite loops in case of install problemsMarc Espie
2009-12-26Add tests to check timeout on sockets under different situations. kurt@ ok.Federico G. Schwindt
2009-12-25Regression test for .Pa below .Bl in the FILES section;Ingo Schwarze
still broken in mandoc 1.9.14, fix being sent to kristaps@. Path names (.Pa) are usually underlined, except below .Bl -inset, -hang, -ohang, -tag or -column in the FILES section (yuck). Note that .Bl -bullet is subtly broken in an urelated way: That will need fixing as well before this test can fully succeed.
2009-12-24Regression test for .Bl -column;Ingo Schwarze
mandoc 1.9.14 still fails this one, patch being sent to kristaps@. The groff .Bl -column spacing we are trying to follow is weird: with less than 5 columns, groff puts 4 blanks between columns; with exactly 5 columns, groff puts 3 blanks between columns; with more than 5 columns, and also with more than 78 characters per line, groff usually falls over and bites the dust - though the exact failures look rather different in various cases, and a few cases even work... No use trying to test anything for more than 5 columns, though.
2009-12-24activate mdoc/Fl tests (duh.)Ingo Schwarze
2009-12-24Regression test for .Fl with multiple arguments.Ingo Schwarze
Still broken in 1.9.14, fix being sent to kristaps@.
2009-12-23regression test for non-breaking space (\~);Ingo Schwarze
fails with 1.9.13, succeeds with 1.9.14
2009-12-14Add a test for sleep(3), to verify that it both (a) returns zero when itPhilip Guenthe
completes, and (b) returns the unslept time in seconds when interrupted
2009-12-12test case for the gamin/fam scenarioMarc Espie
2009-12-12more complex testcase: let q1-q4 have dependencies with outside q5,q6.Marc Espie
2009-12-11sample case we fail (kili/edd): if updatesets contain "self" dependencies,Marc Espie
they should install/uninstall in the correct order to be able to handle any @exec activity...
2009-12-11simplify building of libsMarc Espie
2009-12-03split-update test with shared libraries, twice the fun.Marc Espie
2009-12-02and check resultMarc Espie
2009-12-02splitted update: a -> a + b with inter-dependenciesMarc Espie
2009-11-22and so, check the other case as well (this one does work by accident)Marc Espie
2009-11-22new test, modeled after some old kde update.Marc Espie
When a file moves between packages, one new package will conflict with two old packages, in which case pkg_add should update both (and it does not, not yet). We distinguish from the case where a new package is an update for two packages by the absence of a @pkgpath annotation.
2009-11-22partial updates ought to update only their dep treeMarc Espie
2009-11-21Add a (hopefully) comprehensive regression test of how stdio functionsPhilip Guenthe
affect stream orientation
2009-11-19Add regression tests for stdio threading. Originally by blambert withPhilip Guenthe
some further hacking by me
2009-11-16verify that fuzzy list stuff worksMarc Espie
2009-11-15tweakMarc Espie
2009-11-15add check for always-updateMarc Espie
2009-11-15repair things if dependencies existMarc Espie
2009-11-14more testsMarc Espie
2009-11-14more half-baked regress-targetsMarc Espie
2009-11-14forgotMarc Espie
2009-11-14add a collision-check, don't activate yet, need to understand whatMarc Espie
pkg_add should output, and fix it.
2009-11-14variables for concisionMarc Espie
2009-11-14add a somewhat icky test that will check that packages can represent fairlyMarc Espie
long file (and link) names correctly.
2009-11-14fix regress test (needs -F downgrade to be able to go back)Marc Espie
Introduce two scripts to make it easier to write a "full" scaffolding for further tests, to be used soonish...
2009-11-12respect the block size returned by the driverJacob Meuser
2009-11-11Regress for skipping evaluation of else-type conditional if previousFederico G. Schwindt
corresponding one evaluted to true. espie@ ok.
2009-11-09add regression test for ssh-keygen pubkey conversionsDarren Tucker
2009-11-03actual fixed tests.Marc Espie
2009-11-03Add some (currently failing) version number comparison tests.Stuart Henderson
ok espie@