<fossy>they mean the 3 tired bootstrap <fossy>Which I dont care about , removing that and doing it manually dosent change the dependencies and the like <nimaje>"make failure cheaper and faster" sounds a little bit like erlangs failfast/"Let it crash" style <xentrac>not sure how that came out as "ketas: w" ***stikonas_ is now known as stikonas
<OriansJ>pabs3: I am pretty sure all of the "easy wins" in bootstrapping have already been mined out and what remains is long term and slow development (not often the place of successful grants) <OriansJ>vagrantc: I forget if there were issues with mescc-tools (as I got completely take off track) <vagrantc>OriansJ: overall looks fine to me ... there was one seemingly random issue i should dig up <OriansJ>nimaje: well it is a pretty old idea but doesn't work to well in environments where the cost of failure by default must be extremely high. <vagrantc>mysteriously armhf and arm64 of the same version built fine ... <vagrantc>test/test3/hello.sh: 23: GET_MACHINE_FLAGS: parameter not set <vagrantc>and previously, the same version built fine on amd64 too ... <vagrantc>something in the toolchain changed that only broke amd64 ... weird. <vagrantc>i wasn't able to reproduce the failure locally ... <OriansJ>because get_machine wasn't built before the test ran <OriansJ>and I just pushed a makefile fix that should prevent that from showing up in the future