[
https://issues.apache.org/jira/browse/YETUS-156?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel&focusedCommentId=15255638#comment-15255638
]
Allen Wittenauer commented on YETUS-156:
----------------------------------------
bq. Any advantage to using a symlink and detecting the invocation name compared
to making the convenience wrapper pass --empty-patch when invoking
test-patch.sh?
There really isn't much difference other than having to do a lot less work if
we want to represent the "binary's" name since $0 is going to come back as qbt
instead of test-patch. So things like usage, etc, are correct.
bq. the lack of any tests in the source tree seems like a big red flag, but I
suppose this will say "everything is fine" too often to be useful? In full
mode could it only vote when it would vote -1?
I didn't convert it to run in full mode for exactly that reason: for the vast
majority of projects, there's going to be unit tests and it's going to be extra
filler on an already long report. Plus I'd assume that devs would know a
project doesn't have unit tests.
> [Umbrella] build driver
> -----------------------
>
> Key: YETUS-156
> URL: https://issues.apache.org/jira/browse/YETUS-156
> Project: Yetus
> Issue Type: New Feature
> Components: Test Patch
> Reporter: Allen Wittenauer
> Assignee: Allen Wittenauer
> Fix For: 0.3.0
>
> Attachments: YETUS-156.02.patch, YETUS-156.03.patch
>
>
> it might be useful to create something that runs test-patch's plug-ins as
> part of a build process. In other words, perform a full source build,
> generating reports from checkstyle, findbugs, shellcheck, multi-jdk, etc.
--
This message was sent by Atlassian JIRA
(v6.3.4#6332)
