Earlier this week Reactor reached a milestone! I consider the current code in the subversion repository to be a Beta Candidate. I expect Reactor to be a Beta Candidate for about three weeks. After about three weeks of stability Ill probably declare Reactor to be beta and will lock down the API.
Once Reactor is in beta Im going to be formalizing a lot of the project. I plan to create a project section on DougHughes.net. (At least more than a blog category!) I plan to move the repository to Simeon Batemans servers and to start using Traq for bug tracking and the project roadmap.
For Reactor to hit 1.0 reactor must:
- Have all know bugs fixed or addressed (Traq will help with this)
- Have complete documentation
- Have a project section on doughughes.net
Unfortunately the documentation is a bit out of date with the current framework. Ill try to get this updated over the weekend or as time permits.
The zip available on DougHughes.net is quite dated. Sorry! If you want the latest code get it from Subversion at svn://www.alagad.com/reactor/trunk. Ill post a new zip a few days after I havent found any new bugs.
DougHughes.net 
Comments on: "Reactor Reaches Beta Candidate" (15)
Let me be the first to congratulate you. We’re thinking of Reactor for our next big CF project. Can you give a rough estimate of when you expect 1.0; the Fall maybe? Late Summer?
LikeLike
I’m pushing to reach 1.0 by CFUnited. No promises, but that’s the goal.
LikeLike
Doug,
But releasing 1.0, please put Reactor through a stress test. Some of the methods (selectall) take forever when the table has more than 50 records.
LikeLike
Tony – I’m not sure which version of Reactor you’ve been using. The latest is pretty quick in production mode.
Joe Rinehart has been using jmeter to test Reactor under load and it’s been performing quite well. Sean Corfield spent quite of time optimizing the oo queries.
All in all, as Reactor matures it will get faster.
Lastly, when using Reactor be sure to put it in production mode and turn Trusted Cache turned on.
LikeLike
Just to add to this, I’ve also been working to remove redundant or time-consuming calls to create object or create methods on the reactorFactory. This seems to be speeding things up too.
LikeLike
So was Oracle support ever added to Reactor?
LikeLike
Paul – Yes!!!
LikeLike
Doug,
is there a bug list for reactor? I have happened to find a strange behavior if i try to take advantage of some of the queryobject.getwhere() methods (i.e. isNotEqual()) in context with database-mapping: reactor then will not take the aliases but the original names. This affects alpha and beta as well.
Greating Joerg (I’m happy to work with reactor, though!)
LikeLike
I am getting an error here: svn://www.alagad.com/reactor/trunk
No connection could be made because the target machine actively refused it.
svn: Can’t connect to host ‘www.alagad.com’: No connection could be made because the target machine actively refused it.
Am I doing something wrong or is the server really “actively” refusing me?
LikeLike
Hi Doug,
I was wondering what would be involved with adding Ingres support to Reactor?
Is it simply a case of creating Convention.cfc and ObjectDao.cfc with the relevant Ingres-specific stuff in?
I know Ingres and PostgreSQL are related in some way – possibly it would just involve adapting the existing PostgreSQL components, if they’re close enough?
Thanks.
LikeLike
Peter – You’re pretty much right on. You’re welcome to take a shot at it if you’d like.
Doug
LikeLike
Cool. I had a go with adapting the PostgreSQL one, but there’s too many differences, so I might as well start with the blank ones.
(Hopefully I wont have to read the entire 662 page reference guide to figure out the bits I need.)
Once I’ve got it done I’ll upload it somewhere, on the off-chance that there are other Ingres users still roaming the earth. 😉
LikeLike
Peter – You should join the mailing list and keep us up todate on your progress.
Random quesiton – is Ingress freely avaliable? If not, it’ll be a hard one for me to test.
LikeLike
Hi Doug
I saw this article today: http://javalive.com/modules/articles/article.php?id=16 and now i have a question, in a clustered environment, will Reactor be affected by the disadvantages mentioned in that article??
I ask because maybe a future project of the company i work for (and in which we are considering the use of Reactor) will be built in a clustered environment.
LikeLike
Leomar – I finally had a chance to read that article. I don’t think you need to worry about this. This sentence from that article sums it up:
“A data cache is useful only when we have data that is accessed frequently but is rarely changed.”
Well, reactor caches data, but only stateless data. Records and Tos are not cached in reactor. Gateways, DAOs, Validators, etc are. In fact, only one instance of any of those cacheable objects ever exist in production for Reactor.
LikeLike