Re: Dependencies and RPM hell, revisited - time for a fork?
Robin Green <[email protected]> Sun, 19 Oct 2003 19:42:49 +0100
| Newsgroups | gmane.linux.usability.annoyances |
|---|---|
| Message-ID | <[email protected]> |
On Sun, Oct 19, 2003 at 12:55:56PM +0200, Daniel C. von Asmuth wrote: > Thus quoth Robin Green on Sun, Oct 19, 2003 at 01:29:02AM +0100: > > Those of you who participated in the long and repititive "dependency > > hell" / "packaging hell" / "RPM hell" discussions earlier on this > > list may be interested to the "ultimate solution" I've worked out: > > Afraid I didn't. What problem are you trying to solve? It says in the blog (although note that the first of the 3 problems has been 90% solved already). > Sounds like creating a huge problem to solve an extremely minor one. Believe me, packaging problems are no joke. I just tried to install straw, and it took me hours. The actual amount of code I needed to download was quite small, but getting it working was the hard bit. On my OS you will just be able to type one command and let it get on with it - even if you already have incompatible packages installed (as I found out I had when I tried to install straw) and need those incompatible packages to stay. So actually, installing straw on Fedora Core Test 2 is a near-perfect example of the kind of situation my system would be able to deal with. > If you want to write an OS, maybe you could consider making it more > reliable than some existing ones (including Linux), by chosing or even > better inventing a programming language that is less error-prone than C. If you look at my blog that's exactly what I am in the process of doing, inventing a new one. I don't actually say on there that it will be less error-prone than C but that goes without saying ;-) > You could consider a kernel that is able to use Linux kernel modules > as device drivers. Just remember the GPL. Yes, and HURD is a possibility for the future, if it ever becomes really as solid as the Linux stable kernel series. -- Robin