13 comments

The development of AspectJ integration is currently frozen.
It is still an open question whether the feature will make into the IDEA5.0 release.

--
Best regards,
Eugene Zhuravlev
Software Developer
JetBrains Inc.
http://www.jetbrains.com
"Develop with pleasure!"


0

Eugene Zhuravlev (JetBrains) wrote:

The development of AspectJ integration is currently frozen.
It is still an open question whether the feature will make into the IDEA5.0 release.


Not that I need AspectJ integration at this moment, but perhaps with the
new language openapi, it could be made a plugin.

Bas

0

Could you give an small note on what is currently working in
IDEA4.5/5.0? I sucessfully tried it a year ago but now it doesn't seem
to work (It correctly finds pointcuts but doesn't weaves into the
classfiles. Maybe it doesn't work with 1.5 language features?).

Thanks, Sven.

Eugene Zhuravlev (JetBrains) schrieb:

The development of AspectJ integration is currently frozen.
It is still an open question whether the feature will make into the IDEA5.0 release.

0

It does not correspond to the latest version of the AspectJ language and of course it does not honor any java 5 new features.

--
Best regards,
Eugene Zhuravlev
Software Developer
JetBrains Inc.
http://www.jetbrains.com
"Develop with pleasure!"

"Sven Steiniger" <sven.steiniger@gmx.de> wrote in message news:d0497k$cpm$1@is.intellij.net...

Could you give an small note on what is currently working in IDEA4.5/5.0? I sucessfully tried it a year ago but now it doesn't
seem to work (It correctly finds pointcuts but doesn't weaves into the classfiles. Maybe it doesn't work with 1.5 language
features?).

>

Thanks, Sven.

>

Eugene Zhuravlev (JetBrains) schrieb:

>> The development of AspectJ integration is currently frozen.
>> It is still an open question whether the feature will make into the IDEA5.0 release.
>>


0

My question was a bit misleading. I definied pointcuts to "normal" :)
Java 1.3 code and they are correctly shown in the "find pointcuts"
dialog. But they are not weaved into the classfiles.
The project itself is set to language level 5, so maybe this breaks the
weaver. Or maybe just the project's size.

Eugene Zhuravlev (JetBrains) schrieb:

It does not correspond to the latest version of the AspectJ language and of course it does not honor any java 5 new features.

0

The weaver is disabled currently.

--
Best regards,
Eugene Zhuravlev
Software Developer
JetBrains Inc.
http://www.jetbrains.com
"Develop with pleasure!"

"Sven Steiniger" <sven.steiniger@gmx.de> wrote in message news:d06io2$7i6$1@is.intellij.net...

My question was a bit misleading. I definied pointcuts to "normal" :) Java 1.3 code and they are correctly shown in the "find
pointcuts" dialog. But they are not weaved into the classfiles.
The project itself is set to language level 5, so maybe this breaks the weaver. Or maybe just the project's size.

>

Eugene Zhuravlev (JetBrains) schrieb:

>> It does not correspond to the latest version of the AspectJ language and of course it does not honor any java 5 new features.
>>


0

Not supporting aspectj is now costing you licenses. I don't know if anyone from Intellij went to the TSS convention, but the word I heard was that when a panellist asked if anyone in the audience was using aspects in production now or in the near future, 30% of the hands in the audience went up.

There is something of a trend in the java aspect community of consolidating around aspectj, and moving away from XML based and java src compatible configurations. The spring fellows are now recommending aspectj over their own aop framework; the Aspectwerkz project has merged into aspectj. And now with the imminent aspectj M2 release and support for java 1.5, I expect to see adoption to continue to steamroll.

Aspects are here to stay. Aspects require much more explicit tool support than plain old java technologies (jstl, struts, ejbs, etc), and will drive users away from tools that do not support it.

Idea's relationship with aspect code is not even oblivious, it is actively nasty - it (often silently) breaks almost all of the refactorings, as usages in aj files are not tracked. Don't give us aspectj users fancy features. Just keep the ide from breaking - figure out a hack to keep OO java refactorings working, and the overall project compiling. At least give us some partial support.

Please don't abandon us to Eclipse.

0

I'll second a vote for AspectJ support. The company I'm currently working for uses Eclipse and AspectJ. Class dependencies from a compilation perspective are managed as Eclipse plugin projects. With over 200+ projects, this makes my "rogue" use of IntelliJ difficult. However, lack of AspectJ support (which is provided by AJDT) eliminates the possibility completely. IntelliJ has always been an IDE on the leading edge. Aspect support is one place where this is not the case.

0

Yeah, except that the number of people currently using it was about 3 people. Funnily enough, LAST year when the question of 'do you plan on using it soon', more hands went up than this year, yet the small number of users remains the same.

You have a interesting definition of 'steamroll'. Reality would indicate its a lot closer to 'splutter and stumble about' that any kind of steamrolling.

You might well be a big aspect user. More power to you! You likely are part of a small active aspects community, more power to you all! Try to maintain a sense of perspective though, and face up to the fact that AOP is, at the time of writing, a small small niche with a tiny userbase (no matter how vocal they are).

0

On Sun, 20 Mar 2005 07:12:30 +0300, Hani Suleiman wrote:

Yeah, except that the number of people currently using it was about 3
people. Funnily enough, LAST year when the question of 'do you plan on
using it soon', more hands went up than this year, yet the small number of
users remains the same.


Did anyone at TSS ask the followup question of "how many are/will be using
AspectJ if/when they do use aspects?"

It's one thing to say "IDEA should support AOP" or that "AOP is
steamrolling and we don't want to be left in the dust", but backing ONE
vendor of AOP is probably worse than not.

I could see the use of Aspects and AOP are part of a larger framework,
such as Spring, may end up being more prominent.

Looking at AspectJ now, I see the latest versions are moving away from
using a custom syntax, to using Attributes - if IDEA had spent their
efforts support the older syntax, this work would have to be redone again,
and then redone whenever AspectJ change their minds again.

0

Hello andy,

Not supporting aspectj is now costing you licenses.


That comes across a rather agressive opening sentence. It's also probably
true. But consider the tradeoff between the effort involved in implementing
feature X, and the number of people that will buy licenses as a result. Substitute
AspectJ in the above sentence with some other proprietary technology of your
choice and see how it sounds.

Aspects require much more explicit tool support than plain old java technologies


Exactly. There's likely a lot of work involved in adding solid AspectJ support.
Is it worth their time and effort? I suspect they've decided it's not, and
as much as it probably pains you, I don't really blame them. Even though
you make it seem as though AspectJ is taking the Java community by storm,
I'm almost certain that's very far from being the case.

...the word I heard was that when a panellist asked if anyone in the audience

was using aspects in production now or in the near future, 30% of the hands
in the audience went up.

This was a panelist at an Aspect talk perhaps? If so, it's only to be expected
that a large percentage of the audience in such a talk would be using aspects.
In fact if it was a talk on Aspects, I'd say 30% was an embarrassingly low
figure.

My impression is that AspectJ has only a very very small niche, and I don't
see that changing much anytime soon. I'm basing this impression on having
talked to many developers from many different companies, both in passing
and through interviewing potential candidates, often for fairly senior positions.
Few have more than a passing knowledge of what AOP is all about and I am
yet to meet a single person who is even considering using AOP commercially.
I'm sure these people exist, but I'm also sure there's not very many of them!

I'm sure JetBrains would love to be able to provide the support you're after
in their product, but I also think they've made the right business decision
to put the work on hold, giving preference to improved webapp development
and the like. They'll pick up 10x more licenses than they lose that way.


0

I wonder if we could express better. Thanks Chris!
-


Maxim Shafirov
http://www.jetbrains.com
"Develop with pleasure!"

Hello andy,

>> Not supporting aspectj is now costing you licenses.
>>

That comes across a rather agressive opening sentence. It's also
probably true. But consider the tradeoff between the effort involved
in implementing feature X, and the number of people that will buy
licenses as a result. Substitute AspectJ in the above sentence with
some other proprietary technology of your choice and see how it
sounds.

>> Aspects require much more explicit tool support than plain old java
>> technologies
>>

Exactly. There's likely a lot of work involved in adding solid AspectJ
support. Is it worth their time and effort? I suspect they've decided
it's not, and as much as it probably pains you, I don't really blame
them. Even though you make it seem as though AspectJ is taking the
Java community by storm, I'm almost certain that's very far from being
the case.

>> ...the word I heard was that when a panellist asked if anyone in the
>> audience
>>

was using aspects in production now or in the near future, 30% of the
hands in the audience went up.

This was a panelist at an Aspect talk perhaps? If so, it's only to be
expected that a large percentage of the audience in such a talk would
be using aspects. In fact if it was a talk on Aspects, I'd say 30% was
an embarrassingly low figure.

My impression is that AspectJ has only a very very small niche, and I
don't see that changing much anytime soon. I'm basing this impression
on having talked to many developers from many different companies,
both in passing and through interviewing potential candidates, often
for fairly senior positions. Few have more than a passing knowledge of
what AOP is all about and I am yet to meet a single person who is even
considering using AOP commercially. I'm sure these people exist, but
I'm also sure there's not very many of them!

I'm sure JetBrains would love to be able to provide the support you're
after in their product, but I also think they've made the right
business decision to put the work on hold, giving preference to
improved webapp development and the like. They'll pick up 10x more
licenses than they lose that way.




0


In article <c8a8a1bf6c328c6fba6957778c1@news.jetbrains.com>,
Maxim Shafirov <max@jetbrains.com> wrote:

I wonder if we could express better. Thanks Chris!


Probably not. This was really well said.

-------------------
Maxim Shafirov
http://www.jetbrains.com
"Develop with pleasure!"

Hello andy,

>> Not supporting aspectj is now costing you licenses.
>>

That comes across a rather agressive opening sentence. It's also
probably true. But consider the tradeoff between the effort involved
in implementing feature X, and the number of people that will buy
licenses as a result. Substitute AspectJ in the above sentence with
some other proprietary technology of your choice and see how it
sounds.

>> Aspects require much more explicit tool support than plain old java
>> technologies
>>

Exactly. There's likely a lot of work involved in adding solid AspectJ
support. Is it worth their time and effort? I suspect they've decided
it's not, and as much as it probably pains you, I don't really blame
them. Even though you make it seem as though AspectJ is taking the
Java community by storm, I'm almost certain that's very far from being
the case.

>> ...the word I heard was that when a panellist asked if anyone in the
>> audience
>>

was using aspects in production now or in the near future, 30% of the
hands in the audience went up.

This was a panelist at an Aspect talk perhaps? If so, it's only to be
expected that a large percentage of the audience in such a talk would
be using aspects. In fact if it was a talk on Aspects, I'd say 30% was
an embarrassingly low figure.

My impression is that AspectJ has only a very very small niche, and I
don't see that changing much anytime soon. I'm basing this impression
on having talked to many developers from many different companies,
both in passing and through interviewing potential candidates, often
for fairly senior positions. Few have more than a passing knowledge of
what AOP is all about and I am yet to meet a single person who is even
considering using AOP commercially. I'm sure these people exist, but
I'm also sure there's not very many of them!

I'm sure JetBrains would love to be able to provide the support you're
after in their product, but I also think they've made the right
business decision to put the work on hold, giving preference to
improved webapp development and the like. They'll pick up 10x more
licenses than they lose that way.

>

0

Please sign in to leave a comment.