Fri, 30 Dec 2016 22:44:34 +0000 Various fixes to build a working Smalltalk/X jv-branch distribution
Jan Vrany <jan.vrany@fit.cvut.cz> [Fri, 30 Dec 2016 22:44:34 +0000] rev 112
Various fixes to build a working Smalltalk/X jv-branch distribution * correctly install `librun.so`, `librun.dll` * install all `rc.d` files * fixes installation process to handle packages with no `.so` / `.dll` (source-only package such as test packages)
Tue, 27 Dec 2016 21:25:27 +0000 Win32: A nasty workaround to avoid Borland make to crash under Windows 10
Jan Vrany <jan.vrany@fit.cvut.cz> [Tue, 27 Dec 2016 21:25:27 +0000] rev 111
Win32: A nasty workaround to avoid Borland make to crash under Windows 10 Recently a Borland's ancient `make.exe` started to crash on one of the Windows 10 slaves. The reason is unknown but hell, this is so old and not supported anymore so no wonder. The crash occured on 64bit builds when one tries to remake previously built librun. To avoud that crash, always recompile. This spoils the idea of make, but at least, does not crash. Sigh...
Fri, 30 Dec 2016 00:47:33 +0000 CI: Updated prebuilt STC and stx:librun download code to catch up with new pipleline jobs
Jan Vrany <jan.vrany@fit.cvut.cz> [Fri, 30 Dec 2016 00:47:33 +0000] rev 110
CI: Updated prebuilt STC and stx:librun download code to catch up with new pipleline jobs ...used on SWING CI.
Sun, 25 Dec 2016 22:24:40 +0000 CI: Fixed old-style CI test jobs
Jan Vrany <jan.vrany@fit.cvut.cz> [Sun, 25 Dec 2016 22:24:40 +0000] rev 109
CI: Fixed old-style CI test jobs ...run directly via `Jenkinsfile.rb` from a matrix job. In that case have to to explicitly inhibit package test settings enabled in spec project `stx:jv-branch` (which are there to define "standard" test suite.
Thu, 22 Dec 2016 22:25:06 +0000 Various fixes
Jan Vrany <jan.vrany@fit.cvut.cz> [Thu, 22 Dec 2016 22:25:06 +0000] rev 108
Various fixes * make `update` task to work with binary STC and librun * don't push STC and librun to upstream - this always has to be done manually.
Tue, 20 Dec 2016 17:50:14 +0000 Oops, fixed `should_remove_stc_and_librun_sources()`
Jan Vrany <jan.vrany@fit.cvut.cz> [Tue, 20 Dec 2016 17:50:14 +0000] rev 107
Oops, fixed `should_remove_stc_and_librun_sources()` Sigh, due to use of single `=` instead of `==` in comparison the condition was always trivialy satisfied. Sigh, switching from Smalltalk to Ruby way to often...
Sat, 17 Dec 2016 22:22:43 +0000 CI: Allow to use different credentials to push to upstream
Jan Vrany <jan.vrany@fit.cvut.cz> [Sat, 17 Dec 2016 22:22:43 +0000] rev 106
CI: Allow to use different credentials to push to upstream
Tue, 13 Dec 2016 22:48:25 +0000 CI: Revise pipelines, use "integration" pipeline by default
Jan Vrany <jan.vrany@fit.cvut.cz> [Tue, 13 Dec 2016 22:48:25 +0000] rev 105
CI: Revise pipelines, use "integration" pipeline by default ...and "build" pipeline for feature branches.
Tue, 13 Dec 2016 22:06:34 +0000 CI: Remove hack to disable standard test suite on SWING CI
Jan Vrany <jan.vrany@fit.cvut.cz> [Tue, 13 Dec 2016 22:06:34 +0000] rev 104
CI: Remove hack to disable standard test suite on SWING CI ...as a preparation to move SWING CI from matrix jobs to pipeline jobs.
Tue, 13 Dec 2016 13:50:58 +0000 CI: Pass `--ssh ssh` option to hg when checking out / updating rakefiles
Jan Vrany <jan.vrany@fit.cvut.cz> [Tue, 13 Dec 2016 13:50:58 +0000] rev 103
CI: Pass `--ssh ssh` option to hg when checking out / updating rakefiles ...in order to make it work with `ssh-agent-plugin` on Windows. The problem is that on Windows Mercurial is configured to use PuTTY's `plink.exe` as SSH client. `plink.exe` cannot don't know how to talk to OpenSSH's key agent. On CI setups we use OpenSSH's agent run via Jenkins's `ssh-agent-plugin` so we have to use OpenSSH client. Hence the `--ssh ssh` option. This affects only Jenkins setups, user may still use properly configured `plink.exe`!
(0) -100 -10 +10 +100 tip