I can see strong arguments for using dash as the default sh, but I find the Ubuntu's devs approach wrong and triggering unnecessary pain to users and Ubuntu's image.
The switch should be a slower process. IMHO it should be initiated as probono at 2007-02-18 14:59:18 explained - ie. start it with using dash explicitely only for Ubuntu's own sh scripts, which are not dependant on any upstream. Then a message should follow via all the Ubuntu broadcasting channels, and to all the upstreams providing any sh scripts, that bash->dash switch is planned. The message should contain the outline of benefits: 1. examples of speed gain in the first place, based on the experience from uisng dash for Ubuntu own scripts 2. followed by POSIX compliance rationale 3. then with "make the world better" argument as the last one If possible, spread the word to other ditros and encourage them to join you. After most upstreams are done with transition, or at least known to be in the process, make the switch. That would hurt much less. Mostly, only third party non mantained software could remain broken. However, thanks to wide promotion and community feedback, those could at least be less surprising for users. Although I sympathize with those who find 3 and 2 are actually most important, I believe that trying to achieve the goal by "dash or death" approach was absolutely not appropriate. It fixed something, but broke another thing. While if the transition went smoother and more transparently, it could be almost all-wins process and a success story. Ubuntu's image would not worsen, correctness would prevail, less unhappy users.. Evolution istead of revolution, so to speak. Add GRASS GIS to the list of dash incompatible software. We are working on it. Hopefully soon to be fixed. But did Ubuntu really had to do it the hard way? -- Script that are using bash could be broken with the new symlink https://launchpad.net/bugs/61463 -- ubuntu-bugs mailing list ubuntu-bugs@lists.ubuntu.com https://lists.ubuntu.com/mailman/listinfo/ubuntu-bugs
