Re: [fossil-dev] Current status of make test?

2016-01-05 Thread Joe Mistachkin
Ross Berteig wrote: > > I figure the first step to building test cases for JSON support is to > prove I can build a fossil that passes its existing tests. Apparently > that isn't the case. > Please try setting the TEMP environment variable to a directory writable by the current user. -- Joe

Re: [fossil-dev] Current status of make test?

2016-01-05 Thread Ross Berteig
On 1/5/2016 5:29 PM, Joe Mistachkin wrote: Ross Berteig wrote: I figure the first step to building test cases for JSON support is to prove I can build a fossil that passes its existing tests. Apparently that isn't the case. Please try setting the TEMP environment variable to a directory

Re: [fossil-dev] Current status of make test?

2016-01-05 Thread Joe Mistachkin
Ross Berteig wrote: > > Would it make sense to add a 000_FIRST.test (named something that > guarantees it to run first) that verifies easy to miss assumptions made > by the test suite itself, such as $TEMP is writeable? > I suppose we could just complain in "tester.tcl" if it's not set. > >