Open tpetricek opened 8 years ago
JonathanMEdwards
commented
8 years ago I love the idea of this workshop! It sounds like you are soliciting written papers. But writing can itself impose a conceptual value system, emphasizing post-facto abstractions and logical narratives. What about demos/presentations? That is how we often communicate raw ideas before we get to the writing phase.
jamii
commented
8 years ago I'd second the call for demos. It's often very hard to convey the experience of a new idea through text. Imagine inventing spreadsheets and trying to explain the idea in a paper.
gasche
commented
8 years ago One thing that is not completely clear to me is what sub-field of computer science you are interested in discussing in this different way. I'm personally mostly interested in programming language research, so I will assume in the rest of my answer that the proposed workshop would be about it or at least have a good share of that.
Your workshop idea sounds like fun. I'm not necessarily convinced by the underlying criticism of formalism in programming language research (or maybe I guess I would formulate criticism differently), but I would be interested in attending it.
We had related discussions about the issue that "there are some people doing work on programming language design that can't find academic air time right now, and it's a shame" previously on Lambda-the-Ultimate, in the specific context of programming language design. In A sketch of a "design papers/pearls" category in academic conferences, I proposed an evaluation mechanism based on reviewers trying to use the proposed system after reading its description -- trying to create space for the less formalism-minded people.
Off the Beaten Track (OBT) is an obvious competitor to the proposed workshop. Another is NOOL that seems to be somewhat equivalent for object-oriented folks.
Finally, as far as people are concerned, two people that I know of and could be described as currently falling through the tracks (but certainly there are more) are Sean McDirmid and Jonathan Edwards. You already have the latter here, so, hum, I guess you're doing a good job.
(I'm not convinced by the idea that the mathematical formalism that is common in programming language as a research field is restricting our creativity as designers. I think that it is very expressive and can actually account for most ideas in the wild, in beneficial ways (it may clarify and improve them, and weed out some fragile ideas early on). On the other hand, the process of formalizing requires a lot of work, too much work in some cases to be done atomically, and I agree that we need spaces to discuss ideas before they are formalized. Finally, and I think that is the more important point, not sketched in the writings of @tpetricek I've read so far, requiring formalisation may have a negative social effect, in that it excludes people with less mathematical training that may struggle, or never manage to, or simply not take any pleasure in and thus avoid, using formalization as a methodology. I don't think formalization is a bad way to do research, but I think imposing it excludes some people at our loss.)
mcdirmid
commented
8 years ago There is nothing really wrong with PL research as it is occurring today, it is just that many of us want to work on different problems that just don't appear in POPL/PLDI/OOPSLA/ICFP. Fields are really just communities that are formed around a number of conversations: if you find yourself in a room where people are talking about things that you aren't particularly interested in (nor do they seem very interested in the topics you want to talk about), it might just be time to find a new room (just tell them "its not you, its me!"). These days, I find myself having as much in common with the standard PL conf attendee as I do with maybe someone in say databases or pure theory: that is just not very much at all. On the other hand, put me in a room with people who feel the same way as I do, it is not necessary that we would have much in common either :).
I don't think the solution is to start a new venue based on the "edgy part of PL", PL already has many edges. Onward!, SNAPL, NOOL, OBT...those approaches are not working! Work will always just drift back to the core interests of the community, while PL lacks dynamicity ATM.
I disagree that evaluation is much of a problem. Evaluation is good, it is just that if you lack interest in what is being evaluated, the evaluation is completely meaningless (to you). Formalisms aren't bad, but math for the sake of math is simply vacuous. However, since evaluation techniques are so technical and (for both users and reviewers) expensive, they definitely bias what work can pass into publication or not. Accordingly, doing something "new" or off-of-field is simply career suicide if your goal is to maximize your publication and citation count.
cofinalsubnets
commented
8 years ago Hi! I don't have anything deep to add to what folks have already written here, but I'm very interested to see what comes out of this.
JonathanMEdwards
commented
8 years ago 
rpgpoet
commented
8 years ago Hi. I have a lot of experience with leading writers' workshops, in both the creative writing and software worlds. I would love to act as such for the discussion parts of your workshop. Jonathan and Sean can comment on my style and level of accomplishment in this arena. The above-mentioned Feyerabend Project was my creation a long time ago. I also am in the habit of creating conferences to support work like you describe. Two of them are Onward! and Programming.
pierogitus
commented
8 years ago I'm on the non-academic practitioner side of the audience and this appeals to me.
To @mcdirmid's point about rooms, where are these conversations about the edges happening? Are they organized or ad hoc?
amb26
commented
8 years ago Hi @gasche - Yes, I agree completely. "getting a new room where people are likely to talk about things we are interested in" is exactly the goal here. In terms of "what sub-field of computer science we are interested in", this is rather begging the question since one of our primary perceptions is that we are unhappy with the way that the field has chosen to subdivide its problem space into specialisations. We are primarily interested in work that refutes this classification, by whatever route.
Our conversations are pretty ad hoc presently. One set of of interesting conversations recently occurred at PPIG: https://drive.google.com/open?id=0B-7G3GOHucdiMTNjdEtyMC1qWjQ (in this sort of space, papers from Alan Blackwell, Tomas Petricek, Antranig Basman, Luke Church, Elizabeth Hudnott and others).
Re demos - yes, we should welcome them but we shouldn't be exclusive. Many interesting and productive ideas can still be pushed forward before the stage that anything worthwhile can be done about them practically.
Re formalism, I'm not against it either (and enjoy using it where it is appropriate), but as we air in the other issue #1 it is usually deployed prematurely. As a whole, our discipline is unbalanced as a result of bias in favour of it and we need to right the balance by creating new venues where other methodologies and formats are welcome. In fact, we need to do this constantly and energetically, because as fast as these venues are created, they decay into becoming yet more venues where only formalism is acceptable.
tpetricek
commented
8 years ago I didn't expect so many great comments here, thanks everyone who joined the discussion!
As @amb26 said, that the initial motivation for this was exactly to find a room where people are likely to talk about things we (aka our small group of current organisers) are interested in - and I believe there are some places like that already (Future Programming, PPIG, Onward!) - but often when I try to explain to people in the PL community why I think we need to do very different things, they say that the work cannot be evaluated and so it is not "scientific". We spent a lot of time thinking about doing just unscientific/anarchy/witchcraft workshop (I did read "Against Method" and I'm a fan!), but I think the idea of "critical commentary" moves it a bit further to something with a format that can serve as a baseline for "progressing research programme" of some sort (and perhaps prevent this from drifting back to "core interests of the core community" - I think there are enough people who think differently!)
@JonathanMEdwards I really enjoyed the "screencast" format of the Future Programming workshop (I think it's not only great format for early work, but it makes it possible to do different kind of work.) The main reason why the workshop is more text-oriented so far is that I didn't want to change too many things at once (having an unorthodox critical review format might be "crazy" enough). But @jamii's point about spreadsheets is very good & I'd love screencast/interactive/etc. submissions - I'll try to include it as an option in the proposal.
@rpgpoet Thanks for the help offer with the writers workshop. I attended one at Future Programming (Pittsburgh 2015) and it was one of the best academic event experiences! If the workshop gets accepted, I'll definitely get in touch. I did discover your Feyerabend project some time ago (after reading Against Method myself), but only now noticed the ECOOP 02 workshop. In many ways, we definitely follow your earlier steps with this workshop proposal!
@gasche Thanks for commenting! Do you think you'd be able to attend if this was happening at the ‹Programming› conference? Perhaps I could tempt you to join the PC? It would be great to have someone who disagrees with the underlying criticism of formalism, but can also interestingly write about that!
@pierogitus @gxexj Thanks for the support, we plan to submit this to ‹Programming› call for workshops sometime later this week. If it gets accepted, I'll post an update here!
tpetricek
commented
8 years ago I'm very happy to report that the workshop proposal has been accepted! The important details are:
Thanks everyone for the lively discussion here and we're looking forward to submissions and/or meeting many of you in April!
If you come here, it's probably because someone (likely me) sent you a link to our work-in-progress workshop proposal or you followed a link from my blog. Welcome :-)
Just comment below or leave your name & contact details (or just watch this thread) and I'll let you know if the workshop gets accepted or when we plan something along its line of thoughs!