> On Tue, Jul 15, 2003 at 09:31:17AM -0700, Richard A. Hecker wrote: ...<snip>...
> > There is a Debian package of Freenet. It is currently build 629 and I > > will be the new maintainer since Robert Bihlmeyer has been MIA for a > > while. I have no previous JAVA experience and that is slowing me down > > but there are other issues that weigh more. A typical Debian user will > > expect another Debian package to supply the JRE. Relying on the Sun JRE > > for newer features hurts the Debian package. The stability of a package > > is a serious concern. I try to keep the big picture in mind and take a > > long term view. The typical demands on my time are easier to justify > > when I amortize them over a few years. > > Heh. > > We cannot make a non-sun-dependant Freenet. You cannot make a package > of Sun Java, maybe not even an installer package. We apparently care > more about the windows leechers who probably won't donate to the project > than the people who actually care about Freenet who probably run Linux > or BSD, for security reasons as much as anything else. We therefore > cannot contribute project time to making Freenet work on GCJ. I suggest > you go put your head in a bucket of sand. It will go away eventually! > >=20 I know it will _eventually_ go away. The ostrich approach does not solve any problem. The question is where do you want Freenet to be in a year? In 5 years? Maybe M$oft is your target audience and the rest of us are simple niches. I understand the need for NIO and the willingness to expand the envelope. I also know that it is a trap that immature developers frequently snag themselves on. It is easy to spend all your time putting out fires when a better design would have eliminated them. My initial attempt to configure a node four years ago was a total waste. The current package with build 629 is an improvement but it still requires hacks. As a person I can handle the manual restarts and the 4 daily cron restarts. I might be totally happy with it as a toy to spend some time on. As a package maintainer though, I will focus on ways to improve things. I recently subscribed to this list so my impression may be off base. I do get the feeling that the SUN JRE is a foregone conclusion. Is anyone keeping track of how long it takes to sync up after such a divergent choice? For 20 years I have watched projects chase the fresh new ideas. I was paid to reach the production stage. Implementing new technology is fine and good but we still must reach production. I would like to see Freenet running on multiple platforms with whatever JRE is available. I am not suggesting a least commmon denominator approach, I would like to see recognition of these divergent choices with a general plan for eliminating them. Hiding my head in the sand will not improve the package. I would rather describe the things I see and let everyone fill in the missing details. I currently see a willingness to accept the SUN JRE limitation. I would love to hear ideas about when and how it will be eliminated. It is not reasonable for the Freenet developers to fix kaffe or gcj problems. It is reasonable to hold a civil dialog about what the problems are and what might be useful solutions. I only read the last 2 years of the devel archives but I remember very little of this conversation. I avoid the Debian Policy flamewars but support the general intentions. I find the technical interchanges rewarding as ideas and implementations are hashed out. The archives frequently contain insights never spelled out in the FAQs and HOWTOs. It is nice when this type of transparency exists. Richard _______________________________________________ devl mailing list [EMAIL PROTECTED] http://hawk.freenetproject.org:8080/cgi-bin/mailman/listinfo/devl
