There's been some churn in the community of folks wanting some support for
continuous integration with the Guacamole repos.  Travis CI has come up a
couple of times - I stumbled across it and thought it might be useful, and
another contributor opened a PR for adding the travis.yml files for the
client and server repos.  However, Mike also pointed out that ASF also
operates a Jenkins instance that we could leverage.

So, the two questions up for discussion are:

1) Do we want to leverage a CI system, such as Jenkins or Travis, for the
Guacamole code?

2) If (1 == Yes), is there any reason we would not just use ASF's Jenkins
system?  Is there any compelling reason to go to Travis CI (or any other
system) over ASF's Jenkins?

My personal votes are:

1) Yes, we should leverage some sort of CI system, like Jenkins, for making
up-to-date builds available.  While the Guacamole code is relatively small
and easy to build yourself, there are certainly instances, both while
developing and while testing and using the product, that it's useful to
have those pre-built binaries available.

2) Personally I don't have an affinity for one CI system or another - if
ASF provides Jenkins, seems like that's probably the best way to go, but
I'm not familiar enough with Jenkins or Travis or any of the others to know
what the merits of one are vs. another.

-Nick

Reply via email to