| RE: [stellation-res] Using AspectJ? |
On Thu, 2002-09-05 at 21:43, Jonathan Gossage wrote: > > > > >-----Original Message----- > > >From: stellation-res-admin@xxxxxxxxxxxxxxx > > >[mailto:stellation-res-admin@xxxxxxxxxxxxxxx]On Behalf Of Mark C. > > >Chu-Carroll > > >Sent: September 5, 2002 11:03 AM > > >To: stellation-res@xxxxxxxxxxxxxxx > > >Subject: [stellation-res] Using AspectJ? > > > > > > > > >A question to the group: > > > > > >I'm going to be sending out a preliminary description of the Stellation > > >ACL mechanism later today. I'll be starting to work on implementing > > >that, and completing the trigger mechanism as soon as I get the latest > > >bugs taken care of. > > > > > >The implementation of those things is really complicating the code > > >in the repository, because so many things are getting tangled > > >together. (Look at the LocalHandle implementation, with the X/realX > > >methods to see what I mean...) Using a tool like AspectJ > > >would make things like this much easier to write without creating > > >a mess of the code. (Using AspectJ, the X/realX would disappear, > > >and the trigger code would became a separate file contianing an aspect > > >implementing an "around" wrapper for the relevant methods.) > > > > > >So I'd like to know what people think about the idea of adding > > >AspectJ to our suite of tools. > > > > > >Some things to know about it: > > >- The AspectJ compiler is open-source, under the Mozilla license. > > >- There is an AspectJ task for ant, so command line builds > > > will continue to work. > > >- There is an AspectJ plugin for Eclipse, so it will not interfere > > > with our Eclipse use. > > >- There should be no execution-time performance penalty for using > > > AspectJ. > > > > > >The main arguments against it: > > >- People will need to learn the basics of AspectJ to understand parts > > > of our code. > > >- The setup process for working on Stellation becomes even more > > > complicated, because you'll need to install AspectJ. > > >- AspectJ *can* make things significantly less complex; but it also > > > has the effect of introducing additional "scatter" to the codebase. > > > If it's not used with great care, it can create as many problems > > > as it solves. > > > > > >What do you all think? > > > > > > -Mark > > I took a look at AspectJ a few weeks ago and was impressed by its' > potential. I agree that there will be a learning curve, both for us and for > anyone interested in the code base. I also agree that we need to evaluate > carefully where it can help and not just use it because it is a neat new > tool. In the interest of full disclosure: IBM is one of the major pioneers in aspect oriented software One of the other groups in our department at Watson has a research tool called HyperJ which is a direct competitor to AspectJ. Jim Wright and I regularly attend the HyperJ design meetings, and I've published a number of papers in various aspect-oriented forums - in fact, I'm the co-chair of one at this years OOPSLA. Nn my opinion, HyperJ is vastly superior to AspectJ - AspectJ has serious scalabilit problems. Unfortunately, HyperJ is not open-source, and is distributed under the wretched 3-month alphaworks license. So, on the one hand, I've got a bit of baggage when it comes to AspectJ. On the other hand, I consider it better than nothing, and I've got a lot of background in understanding aspect oriented techniques, and when it's appropriate to use them. > I am certainly in favour of trying it and seing how well it works out. That's what I'm hoping the consensus is. I think aspect-oriented techniques *could* help us, and if the gang agrees, I'd like to give them a try. If they work, great. If not, we'll strip it out and go back to straight Java. But I want to be sure that people are willing to deal with the burden of installing and understanding yet another tool before I go ahead. -Mark -- Mark Craig Chu-Carroll, IBM T.J. Watson Research Center *** The Stellation project: Advanced SCM for Collaboration *** http://www.eclipse.org/stellation *** Work Email: mcc@xxxxxxxxxxxxxx ------- Personal Email: markcc@xxxxxxxxxxx
Attachment:
signature.asc
Description: This is a digitally signed message part