RE: [abc] AAA keynote draft

On Sat, 21 Oct 2006, Eric Bodden wrote:

> Hi, Oege.
>
> Just a few comments:
>
> Slide 36: here you could probably mention that recently they added
> hasMethod/hasField (which bad enough works only in declare annotation)
> and also thought about synchronized pointcuts
> (https://bugs.eclipse.org/bugs/show_bug.cgi?id=123759). So even the ajc
> gang sometimes is not happy with the restrictions imposed by the JP
> model.

No, that is just extending the type pattern language, not the joinpoint
model.

> I don't understand slide 39. What has Jimple to do with encapsulation?
> The user never sees Jimple anyway!? So what are you trying to say?

If you can advise every Jimple instruction, you can see all of the
implementation detail; which is the main reason you currently are not
allowed to advise, say, an assignment to a local.

> In general I find that the OM stuff comes up very abrupt and
> unmotivated. It really looks like a second, unrelated presentation to
> me. But maybe you can overcome this impression by speech... Also, this
> part is really a little short. I am unsure if it's gonna be really
> beneficial to the audience.

Thanks for the tip; I believe it's crucial to consider the issue of
encapsulation hand-in-hand with making the joinpoint model more liberal.

Cheers,

-Oege

>
>
>> -----Original Message-----
>> From: Majordomo list server [mailto:majordomo@comlab.ox.ac.uk] On
>> Behalf Of Oege de Moor
>> Sent: Saturday, October 21, 2006 10:54 AM
>> To: abc@comlab.ox.ac.uk
>> Subject: [abc] AAA keynote draft
>>
>> Friends,
>>
>> I'm giving a keynote at the "Astrenet Aspect Analysis" workshop in
>> Benevento
>> this coming Tuesday [urk, missing OOPSLA on this account]. The
> audience
>> knows AspectJ; it's associated with a conference on re-engineering, so
>> their
>> notion of `analysis' is more like pointcut matching than like dataflow
>> analysis.
>>
>> The web page of the workshop can be found here:
>>
>> http://www.dcs.kcl.ac.uk/staff/zheng/astrenet/html/astrenet8.html
>>
>> I have a whopping 90 minutes to pontificate.
>>
>> A draft of the slides can be found here:
>>
>> http://musketeer.comlab.ox.ac.uk/~oege/AAA.pdf
>>
>> Thanks to Neil and Pavel for already giving me useful feedback on the
>> slides. Of course the remaining errors are all mine!
>>
>> Further suggestions for improvement would be very welcome, especially
>> before
>> I travel to Italy on Monday. As you can see, I go quite deeply into
> the
>> POPL
>> paper, but I intend the last part of the talk to be lighter, hopefully
>> stimulating some discussion on the AspectJ choice of joinpoints [throw
>> it
>> out! power to the people!], and open modules.
>>
>> Should I expand the last part a bit [45 slides is already a lot for 90
>> minutes]? Is there too much datalog in the talk?
>>
>> Cheers,
>>
>> -Oege
>>
>>
>
>
>
Received on Sat Oct 21 17:42:16 2006