Dreamsongs Wiki

TypesOfPapers

We have a dilemma here. OOPSLA is a major ACM conference and therefore we need to accept technical papers according to rules that enable people to get academic credit and then credentials based on those papers. OOPSLA used to be a place where people changing the world showed up, and partly because the technical papers are so incredibly boring, they don't show up-they go to their own niche conferences.

Somehow we have to attract the world-changers back or else settle OOPSLA into one of the ACM niche conferences where just the academics show up. All this while retaining the academic bent.

One idea is OOPSLAEssays. An !OOPSLAEssay is a paper that is more reflective or speculative than is usually accepted in academic conferences.

Another idea is a CoolProgramCommittee. Such a committee is composed of the usual sort of member, but somehow they accept more reflective or speculative papers.

Another idea is a FederatedConference. This means that the program is broken into interest groups, each group designing its own events, but tying them together with common keynotes and the interests of the historic OOPSLA crowd.

What changes can we make to the TechnicalProgram?

What changes can we make to the ConferenceProgram?

-rpg

There seems to be a dynamic tension between the academics and "industry professionals" where extremes do not coexist well.

For example a paper that is "too academic" will not attract an audience of industry professionals (and possibly other academics :-) ... while an industry-based experience report may be perceived as "diluting" the academic integrity of the conference.

While activity sessions, panels, special events offer a diversity of style options for sharing experience � one program element view I feel is missing is that of �management of technology� - asking questions on ROI, time-to-market and quantifying learning curve and technology transfer issues.

-StevenFraser

OOPSLAEssays solves some of this. The primary issue for 2005 will be space for parallel events - which would solve more of the problems you mention. -rpg

The basic nature of tech papers at OOPSLA

hasn't changed too much over the years. (To convince myself

of this, I spot-checked 1993 and 1997, and saw basically the

same kinds of papers as 2004.) What makes anyone think

that tech papers topics or content has anything to do with numbers of people who attend OOPSLA? -DougLea

The general nature of the papers has not changed, but the quality was lower back then. I recall a typical GC paper was usually a sketch of how it might work rather than a working implementation or a proven algorithm. I was laughed out of the program committee room for complaining about this in 1992.

Back then we had exhibits - which were exciting because companies were doing things with object-related technology then. The interesting ideas and developments came from industry because academia was AWOL at that time (from the mid-1980s through the 1990s). All the new ideas and thinking were why people came to OOPSLA. We don't have exhibits now because companies are nowhere at the moment - we're in one of the cyclical downturns of commercial innovation (and a small upturn of academic). My idea is that the new ideas are going to have to come from things other than the technical papers (as they always have), and that having more types of presentations and other stuff is how we attract the interesting crowds now that there are no new ideas in industry to do it with, and now that the technical papers are on the deadly side for the sort of audience we used to have.

Or...we could let OOPSLA become a symposium. -rpg

As much as I'd like to agree with this statement, it is still the case that a lot of

papers at OOPSLA are not implemented. After I switched my team to "build

then publish" in the 90s, we often ask authors for implementations and experience

with implementations. For a conference that is known for "practical" values, OOPSLA's authors are the most likely to say "we have a prototype and no

we don't really program in it" meaning they don't really have anything.

Let's devalue "publish at OOPSLA" and just bring lots of ideas there. Everyone

knows then that the speaker is talking of dreamware but someone might be

inspired to build, experiment, prove, validate, and experience with the idea

seriously and report on it elsewhere.

See radical suggestions for same topic. 'Nuff said. -- Matthias

I have been involved with a conference called OT in the UK since it began in the early 90s. Some of you will know the conference, many of you will know Bruce Anderson who was one of the founders. This conference has a variety of session types, but no real 'paper sessions' where attendees sit and listen to one person speaking. The conference is about interaction betweem presenters and audience. This makes it less attractive to academia, but I wonder whether we could look at having more of these kinds of sessions (the goldfish bowl is one example session type that has made it to OOPSLA). Indeed, the conference itself is being re-branded as SPA (software practice advancement) in order to reflect its moving and increasingly multi-disciplinary focus. I would vote for having a more multi-disciplinary focus for OOPSLA. --HelenSharp

BTW, these sessions all have outputs that are put up at the conference and on the website after the event... HelenSharp

It appears that the ASPLOS community has faced a similar problem to one we face: How to counter stagnation in the field. Here is what they wrote in 2000:

The quantitative approach to computer architecture has been the backbone of our field for a long time. Unfortunately, it has also served to discourage new ideas. Papers usually do not get admitted to ISCA or ASPLOS unless the systems that they describe are mature enough to run the Spec95 or SPLASH benchmark suites. Many have noticed that this has a chilling effect on the ideas generation process

encouraging incremental research.

We countered ours with Onward!, but the brief genre of Wild seems appealing. At the first WACI session, they collected all the submissions on a.

Looks interesting. -rpg

I like the idea of the ASPLOS WACI session. At first, I thought

that we could accomodate it within a new PC procedure where PC members voted with minutes. The WACI session would then be one where papers were chosen based on novelty rather than proof.

But the main advantage of advertising a WACI session and having

a chairman for it is that it would attract different papers than

would otherwise be attracted. The CP can still donate papers to the WACI session, just like it does now with practitioners reports.

I have been thinking about another kind of paper/presentation

that I will call (for now) an ApplicationExperience.

-RalphJohnson