On Saturday 19 October 2002 16:06, ian wrote: Sorry about the preceding empty message.
> > We should do the release only after the code is stable. The only way to > > know the code is stable is to test it. > > This is exactly what we have been doing for the last 15 days. Frozen means you don't change the code. The last change I saw was 3 hours ago. We have not been testing the same release image for 15 days. > > > I would suggest at a minimum, *after* the Windows installer is working > > perfectly we freeze CVS, bump the minimum build number to force all > > non-release candidate nodes off the network and test for at least one > > week. Two would be better. > > The Windows installer is irrelevant, it has its own release schedule. I disagree. Most users of a stable release will use windows. There are Windows (or at least windows JVM) specific bugs. We need to get the Windows installer working so we can get as many windows nodes as possible up to test out the network *before* the release. > > > There two things that I want to get in before a freeze: > > 0) Cleaner handling of undecodable splitfiles in the fproxy SplitFile > > downloading ui. > > 1) Update the FEC code in freenet_ext.jar to use the onionnetworks 1.0.3 > > release which has full source for all support libraries. The changes > > between 1.0.2 and 1.0.3 are small. The output should be bit-for-bit > > compatible so this shouldn't be a disruptive change. > > > > I should be done with these by the end of the weekend. > > If they can be suitably tested by wednesday, then do it - if not, wait > until 0.5.1. ok. > > > I wouldn't even consider making announcements to press until after we > > have tested for at least a week. > > We have been testing for the last 15 days. We haven't been testing the final release candidate for 15 days. > > > > Getting good PR coverage is essential, > > > > Hyping freenet is a bad idea. When it works well people with tell other > > people about it and the userbase will grow organically until journalists > > start to call you. If it works well you will have more press than you know what to do with. The risk is that if you hype things before we are sure that its working well a lot of unsophisticated users will try it give up in disgust and go away forever. > > Don't be dumb - more users means more testing and more donations. More > donations means that we can pay Matthew to work for longer, and that is > good for the project. > > > > as it will lead to wide > > > deployment and more donations > > > > !!! > > We should not be manipulating press coverage to make money. > > That is the dumbest thing I have ever heard you say. > > > Are we really short of money? If so just put a request on the > > freenetproject.org site asking for donations. I would be surprised if > > you didn't get whatever is nescessary. > > We will, it will be part of the 0.5 release. We have been working on this for years. What is the motivation for rushing the release? Is the underlying problem that we are running out of money to pay mt? If so how much money do we have and how much do we need? --gj > > Ian. _______________________________________________ devl mailing list devl at freenetproject.org http://hawk.freenetproject.org/cgi-bin/mailman/listinfo/devl
