Yes, we should include source of other tools only when there is no reasonable alternative and it's required to run ATS at all. We hope to remove YAMLCPP once a sufficiently recent version is in general distribution.
I'm not sure not having the test disable is going to work in real life. It means that no test can use an obscure features without causing pain for the community. Susan has also mentioned to me that we may want to have tests that require privilege, which will not be runnable in many cases. On Tue, Jan 8, 2019 at 1:19 PM Leif Hedstrom <zw...@apache.org> wrote: > > > > On Jan 7, 2019, at 8:43 PM, Pushkar Pradhan <pprad...@oath.com.INVALID> > wrote: > > > > Leif, > > I am fine with that. What do you think of my proposal of including a > recent > > curl version in the repo? > > I don’t think we should do that. :) > > Getting an appropriate version of curl is not that big of a deal, > certainly no worse than getting the C++17 compiler tools that we need, > that’s a bigger hurdle IMO. > > > > > The other option is to fix the bootstrap.py to install it. > > On RedHat the new curl is available in httpd24 but running that requires > > prepending the curl command with scl enable…. > > Right, I think documenting around this is sufficient, just as we do for > the scl dependencies for compilers etc. > > Most people would never run autest :). > > Cheers, > > — Leif > > -- *Beware the fisherman who's casting out his line in to a dried up riverbed.* *Oh don't try to tell him 'cause he won't believe. Throw some bread to the ducks instead.* *It's easier that way. *- Genesis : Duke : VI 25-28
