Re: Finishing work
"La Monte H.P. Yarroll" <[email protected]> Mon, 21 Mar 2005 10:08:32 -0500
| Newsgroups | gmane.comp.misc.free-software-business |
|---|---|
| Message-ID | <[email protected]> |
Jamie Lokier wrote: >Michael R. Bernstein wrote: > > >>>From: Tom Lord <[email protected]> >>> >>>as a *hacker*, i've often been >>>disappointed when my big, richly complicated but totally nifty >>>ideas don't make me much money /while at the same time/, some >>>dork will take a very simple hacking idea, spend all of their >>>time on packaging and marketing, and Win Big. >>> >>>Those lucky dorks know something about what's important in markets >>>that my hacker-self doesn't. What, though? >>> >>> >>Hackers generally consider finishing work and polish to be part of 'packaging' but end users don't. >> >>The finishing work and polish that goes into creating a smooth >>experience (especially a smooth *installation and configuration* >>experience) is relatively boring to developers, and so frequently >>doesn't get done at all. >> >> > >These days we have package maintainers as well as developers. > >Package maintainers do precisely that work. > >Why, then, doesn't the combination of developers + package maintainers >result in "Winning Big"? > > I speak as a heavy consumer of the work of package maintainers. I work for a company which produces many custom Linux distributions. I think the social dynamics for package maintainers are very different from package developers. A very large proportion of package maintainers do it professionally. But to make a living maintaining packages, you have to do it wholesale--it is rarely profitable to put a lot of energy into any given package. There are not strong incentives for package maintainers to push changes upstream. It is vastly easier to throw in a patch which works around some random impediment than to craft something acceptable to the package maintainer, get it accepted, and then incorporate the resulting revised package into a distribution. In short, package maintainers are not really providing the kind of polish brought up by Tom Lord. -- "Beware of bugs in the above code; I have only proven it correct, not tried it." -Donald Knuth