Oh, no blame taken. I'm actually really excited about all of the changes. Unfortunately, I've been missing for a while due to increased work responsibilities.
The changes are all things way beyond what I was initially envisioning as a 2.0. What led me to contribute was that I was using Chain and getting really frustrated about how "old" it felt as a consumer of the library. I wanted generics and extensibility. Also, I suspected that the library had a lot of serious thread safety issues and I didn't have the time to address those. I noticed that those are starting to get dealt with! Anyways, in the process of working on the project, I saw that a lot else was possible, but I had to hold back because my time was limited. I think you are doing great work and I'm happy to see a 2.0 get closer toward release. Thanks, -Elijah On Thu, Jun 27, 2013 at 4:17 AM, Benedikt Ritter <brit...@apache.org> wrote: > Hi Elijah, > > nice to have you with us again :-) > > > 2013/6/26 Elijah Zupancic <eli...@zupancic.name> > > > Hi Benedikt, > > > > I made the decision to inherit from ConcurrentHashMap because the > original > > implementation was inheriting from HashMap. I was doing an incremental > > refactoring approach and there was never a good justification for that > > design rather I was trying to make evolutionary improvements. Being able > to > > plug in any map implementation would be a far better design. As for the > > original design decision to use a HashMap, I have no insight. There were > > quite a few design enigmas that I encountered, however if you look at the > > date of the original project I don't think it is unusual. > > > > Thanks for the insights! My mail was not intended to blame you. You did > great work on chain. > Would be great if you could manage to share your thoughts about the > upcoming design changes. > > Benedikt > > > > > > Best of luck, > > -Elijah > > > > > > On Tue, Jun 25, 2013 at 11:24 AM, Benedikt Ritter <benerit...@gmail.com > > >wrote: > > > > > I have created CHAIN-101 [1] > > > > > > Benedikt > > > > > > [1] https://issues.apache.org/jira/browse/CHAIN-101 > > > > > > Am 24.06.2013 um 20:57 schrieb Adrian Crum < > > > adrian.c...@sandglass-software.com>: > > > > > > > I have always preferred the "has-a" approach over the "is-a" > approach. > > > It makes things easier to refactor down the road. > > > > > > > > -Adrian > > > > > > > > On 6/24/2013 7:30 PM, Benedikt Ritter wrote: > > > >> Hi, > > > >> > > > >> I just wonder why ContextMap inherits from ConcurrentHashMap. This > > > seems like an unnecessary restriction. The context interface makes > > > explicit, that implementations do not have to be thread safe. Beside > that > > > we loose all thread safety a ConcurrentHashMap provides with our > > > not-so-thread-safe implementations in ContextMap and ContextBase. I'd > > > suggest to switch from an inheritance based approach to a delegation > > based > > > approach. That leaves users with the liberty to choose what ever > > underlying > > > Map implementation they like for their context. > > > >> > > > >> WDYT? > > > >> Benedikt > > > >> > --------------------------------------------------------------------- > > > >> To unsubscribe, e-mail: dev-unsubscr...@commons.apache.org > > > >> For additional commands, e-mail: dev-h...@commons.apache.org > > > >> > > > > > > > > > > > > --------------------------------------------------------------------- > > > > To unsubscribe, e-mail: dev-unsubscr...@commons.apache.org > > > > For additional commands, e-mail: dev-h...@commons.apache.org > > > > > > > > > > > > > --------------------------------------------------------------------- > > > To unsubscribe, e-mail: dev-unsubscr...@commons.apache.org > > > For additional commands, e-mail: dev-h...@commons.apache.org > > > > > > > > > > > > -- > > -Elijah > > > > > > -- > http://people.apache.org/~britter/ > http://www.systemoutprintln.de/ > http://twitter.com/BenediktRitter > http://github.com/britter > -- -Elijah