Benjamin Piwowarski
2014-04-30 12:39:29 UTC
Sorry to ask again, but there was no answer to my previous email.
In the master, with the provided patches (and provided I did not break anything for windows, but I tried to triple check everything), OS X compilation is (unless I am missing things) nearly feature complete. On OS X, it generates a full LyX and working bundle, with all the necessary dependencies - I checked with diffs on the autotools version.
So from what I understand, there is not much missing in order to have cmake working on all platforms; would it be wise to pursue a bit the effort in order to leave autotools behind and focus only on cmake? This would avoid, e.g. for the transition to QT5, to have to check that everything works with the two build systems. If going in this direction, maybe creating a meta bug on trac would make it easier to track down what is missing.
Tell me if this is worth pursuing a bit the effort, and what is missing in the current cmake system. I can work on linux/os x versions but cannot test with windows easily.
Â
Benjamin
In the master, with the provided patches (and provided I did not break anything for windows, but I tried to triple check everything), OS X compilation is (unless I am missing things) nearly feature complete. On OS X, it generates a full LyX and working bundle, with all the necessary dependencies - I checked with diffs on the autotools version.
So from what I understand, there is not much missing in order to have cmake working on all platforms; would it be wise to pursue a bit the effort in order to leave autotools behind and focus only on cmake? This would avoid, e.g. for the transition to QT5, to have to check that everything works with the two build systems. If going in this direction, maybe creating a meta bug on trac would make it easier to track down what is missing.
Tell me if this is worth pursuing a bit the effort, and what is missing in the current cmake system. I can work on linux/os x versions but cannot test with windows easily.
Â
Benjamin