On Jan 8, 2015, at 10:11 AM, Valeri Galtsev galtsev@kicp.uchicago.edu wrote:
I question intelligence of an attitude that something (that works for some people) has to be destroyed to make room for something else one thinks to be more appropriate.
The amount of actively-maintained software has always matched the available brainpower given over to its maintenance. Therefore, if we are going to add more features, some old things have to be left to die.
I do not mean “destroyed” in the physical sense. CDE is still there, if you want it. There just hasn’t been any new development on it in many years now, because the people who were doing that have all moved on.
If you want new features *and* everything old to continue to be maintained, you have a couple of options:
1. Grow the number of active programmers.
Every open source project I monitor closely frequently sends out calls for patches and other contributions, which are usually answered with crickets. Shortly after the call goes out, the crickets are drowned out by people demanding more bug fixes, and more features, and behavior changes, and better docs, and… I don’t know *any* open source project that has more active developers than it knows what to do with.
2. Reduce the amount of effort it takes to maintain a given feature set.
A lot of work has gone into that. It’s one reason software is moving to higher- and higher-level languages. Much of the Red Hat specific code in RHEL is written in Python, for example, not C, the traditional language of Linux.
Then we get old farts complaining that the new software is less efficient, because it isn’t written in C. That’s the tradeoff: computer efficiency for programmer efficiency, because programmers are more expensive and harder to come by.