Herrick, Mike wrote: > Hi Dan, > > I'm in the position of evaluating introducing new (to us) technology. In > order to introduce it, I want to see signs of health. I can only assume > that you are in this position as well in your new job. The roadmap is
So we should clarify that one. No, that's not my position as: (1) Jini is not a core feature of Betfair, distributed systems are. i.e. Technology is not the issue, design and architecture are and we don't necessarily use Jini for implementation. (2) I'm seen by many as "Mr Jini" and have previously maintained patched starter kits for various people. A number of people treat me as their first stop for Jini support and some are even expecting me to go fork JSTK and do something independent of River. In fact some have even encouraged me to fork. I suspect my position then is best termed "unique". I sometimes play "customer" and sometimes "developer" and sometimes "committer" whichever I think is useful for enlarging the community of users of Jini (enlarging the developer community for me is a response to a larger more vibrant user community). > just a logical starting point. It seems to be if not thee - one of the > first things people check when looking at an open source up start. And what are the other heartbeats as we're on the subject? > Perhaps I am harping on that too much so I'll stop. But look at the > state of it right now: > http://incubator.apache.org/river/roadmap.html This sends a message to > people (I think). > > And it isn't necessarily fair, but I think that something like > Jini/JavaSpaces has an even tougher road ahead of it because it has been > around for 10 years. It isn't some brand new project. People expect more I'm less worried about that. There are a number of similar opensource efforts in the universe that had similar issues of adoption and making it to mainstream. > perhaps because they are used to (I presume) Sun Marketing etc. And > sadly, Jini/JavaSpaces didn't ever "make it" into the mainstream > (compared to J2EE) so it has a certain perception amongst people outside > of the community. You know this better than anyone. When the project has > been around for 5 months and these things still are not in place I > wonder and grow concerned. Perhaps others won't. My point is don't make > it easy for them to ignore/write off Jini/JavaSpaces. > Well that's a discussion for beer time - it's way too large and messy to do here. > I think that I am the minority here in that I represent potential growth > of the Jini/JavaSpaces community. I am just trying to give you my > perspective. I assume that I represent part of the whole point of > bringing Jini/JavaSpaces to Apache - to grow the community? > > I harp on the Roadmap, but I'm really talking about everything. The > quickest way to show life IMHO is to get a couple things on the roadmap > and get those .1, .2 releases done (maybe that is the only thing on the > roadmap for now). > > Outside of getting something called River released and getting the > standard processes going, I'm interested in seeing a simplification of > Jini/JavaSpaces. 1. Someone else mentioned this, but forget the > installer, just unzip / untar (take a look at Tangosol or GridGain - it > is dirt simple) Yes on Maven, Ivy, etc. > 2. Configuration - simplify it into a property file or at least provide > that as the default (again Tangolos/GridGain are different, but they are > simple to set up) Properties? Why not some XML? More importantly, why do you consider this simpler? And do you mean you want everything configured in that single file or just some set of core things? And what are the core things? In addition I have another sampler question for you: What about convention over configuration? > 3. Logging - get rid of sys.out, use log4j (I vaguely remember that > somewhere) > Hmmm, where have you seen sys.out? Or do you mean don't use JDK logging use log4j? > Anyway, I want this to be successful and will help where I can. > > Cheers, > > Mike > Many thanks for taking time, Dan. > -----Original Message----- > From: Dan Creswell [mailto:[EMAIL PROTECTED] > Sent: Wednesday, May 23, 2007 3:14 AM > To: [email protected] > Subject: Re: Short term plan forward... (proposal) > > Hi Mike, > > I'm happy for you to wade in - more external feedback the better. > > My current concern is that at the moment you seem to be saying not much > more than "Bad River, do stuff, be there". > > Is that actually all you want or are there specific things you expect > from River? Those things mightn't be roadmap stuff, they might be > something else. In fact, there's a question - why do you want a > roadmap? What does that mean to you? Is it some kind of symbol of > longevity or maybe a hint of some planning/forethought that makes you > feel comfortable or something else? > > Dan. > > Herrick, Mike wrote: >> Mark, >> >> I will try to be more constructive in my criticism in the future. >> River so far has just been confusing to watch. As someone new to the >> community, this is discouraging. >> >> I did mention the same criticism you can find on my blog months ago on > >> this list (was more positive then). We just hit a point in our project > >> where the future looked too bleak that we didn't feel like we could >> continue to invest in this technology. I now have a little hope, but >> we will only pick things back up once it is clear that River is on a >> good track. >> >> I don't have a vote, but I like Jim Hurley's suggestion for a plan. >> Those are nice baby steps that will get things going. Any of the >> follow on stuff can just be .4, .5. .6 or whatever. >> >> In terms of what I can do to help - I can do some grunt work of some >> sort to start. If you have an idea let me know, otherwise I'll keep an > >> eye out. >> >> Mike >> >> -----Original Message----- >> From: Mark Brouwer [mailto:[EMAIL PROTECTED] >> Sent: Tuesday, May 22, 2007 12:31 AM >> To: [email protected] >> Subject: Re: Short term plan forward... (proposal) >> >> Herrick, Mike wrote: >>> Whatever is done, get it on the roadmap. Get a 12 month roadmap. >>> Start >>> acting like a project that has a chance at making it as a poddling. >> Hello Mike, >> >> It is hard not to miss the criticism here and in the blogosphere with >> regard to Jini and the Apache River project. Therefore I do have a >> remark (I just picked your posting so don't take it too personally). >> >> "And so, my fellow Apache River friends, ask not what the project can >> do for you; ask what you can do for the project." >> >> I say this as although it is indeed unfortunate to realize how long it > >> took for the code to arrive, sometimes I also have the feeling some >> people think we are here to fulfill their wishes for which they even >> have a hard time to make them public. >> >> I have no problem with criticism as it is often a healthy way to get >> you out of your comfort zone, at the same time I also think there is a > >> limit to it. This project will only become a success and graduate if >> we are able to collect new committers who want to backup their wishes >> with effort. I realize for a large part that will depend on our >> guidance/openness and atmosphere we create here, but equally important > >> are the ideas and will to contribute by those prospects. >> >> Asking for a roadmap is logical and fine, but so is suggesting ideas >> for that roadmap as it might well turn out that if nobody comes up >> with ideas and did some thinking how that fits in with the current >> investments/installed base it might well turn out that it ain't the >> Big-Bang some of you might be hoping for, while at the same time >> significant work is performed that is only perceived by others as >> 'maintenance'. >> >> This posting is not intended to shut people up, the contrary, but I >> hope we can get a collaborative atmosphere here even while at some >> moment you want to put the other in the drying machine because of >> his/her opinion >> ;-) >> -- >> Mark >> >
