Lambda the Ultimate

I see you've already submitted Ur for PLDI'10...

ACM is still living in the 20th century and trying to keep papers behind pay-gates. If you have an ACM Portal account, it's all here:

http://portal.acm.org/toc.cfm?id=1640089

ACM is still living in the 20th century

"in 1999, 70% percent of e-journals were free [6], in 2003, only 25% were free." It seems like you don't know what is happening in the 21st century, or have some new statistics I don't. Which is it? You should be used to contradictions about how publishing should work in the 21st century (Google Books and Harvard's weird relationship).

and trying to keep papers behind pay-gates.

Yes, because in the 21st century, everything is free! and we all drink Brawndo, The Thirst Quencher and because Brawndo's got electrolytes we'll even use it to quench the thirst of plants! This Joe Bauer guy is trying to say water is better for us than Brawndo, but Brawndo is cheap and tastes good!

Things need to be financed somehow, e.g. since SAE stopped sponsoring Ardour, the developer has been basically going month-to-month in search of independent contributions. Of course, there is no such thing as a free lunch.

As is common here on LtU, when somebody says something off-the-cuff, it's best to counter-attack with hard research studies debunking it, rather than incite a flame war:

The Business of Academic Publishing: A Strategic Analysis of the Academic Journal Publishing Industry and its Impact on the Future of Scholarly Publishing by Russell and McGuinon

Also, ACM Membership comes with Portal access, as well as access to tons and tons of Safari books. There is really no reason to not be a member of the ACM if you read LtU. Especially if you own a Kindle DX.

AND... If you are a professor who lectures students, you really should emphasize ethics throughout, including pushing students to join ACM via Student Membership rather than merely getting it through side channels via Upsilon Pi Epsilon induction or participating in International Collegiate Programming Competition. (Of course, there is a separate issue here that ACM sucks at: If you happen to get free memberships simultaneously from these two sources, while already having a membership, ACM will not 'add' to your membership. They also will, from experience, not return your e-mails about this matter. They also have a public LISTSERV mailing list for lapsed memberships that is "additive" so that if you signed back up, you are not removed from the mailing list. Bottom line: ACM has severe coordination problems, but paying for access is not one of them.)

Z-Bo said:

As is common here on LtU, when somebody says something off-the-cuff, it's best to counter-attack with hard research studies debunking it, rather than incite a flame war:

Right after:

Yes, because in the 21st century, everything is free! and we all drink Brawndo

Shouldn't you have quoted the "hard research study" before resorting to ridicule?

The business model for academic publishing, especially for papers related to research paid for by public money, is a complex and contentious issue, not to be solved by one paper related to the impact of industry trends on academic libraries, or by invoking insulting caricatures.

Shouldn't you have quoted the "hard research study" before resorting to ridicule?

I was viewing it as satire, not just ridicule. Readers are entitled to make their own value judgments, just as they are entitled to make value judgments on people who describe ACM's system as "20th century." (The fact I talk like a redneck is a different matter entirely.)

The business model for academic publishing, especially for papers related to research paid for by public money, is a complex and contentious issue, not to be solved by one paper related to the impact of industry trends on academic libraries, or by invoking insulting caricatures.

What does that have to do with the fact that if you are an ACM Member, you get OOPSLA journals for no additional cost? What does that have to do with the fact that the cost for an ACM Membership is well below the cost of many academic journals, which often cost in the thousands of dollars?

Really, OOSPLA also gives you free videos of talks, too, if you are a member. OOPSLA is simply fantastic and the phrase "pay-gates" rubs me the wrong way.

The average career in high-tech for programmers is a mere 7 years. Studies show that programmers who go beyond this average often tend to expose themselves to many programming paradigms, thereby making themselves more valuable than a horde of recent graduates who see nails and hammers.

ACM's pay-gates are really pearly gates to a long, prosperous career. I am not going to continue this discussion any further than that.

I was viewing it as satire, not just ridicule.

For what it is worth, I respect Adam's work as an academic. Actually, it would be pretty cool if OOPSLA could have a gauntlet of workshops on more modern approaches to web programming, in general. Just line 'em up. Have somebody coordinate it. Have a BoaF afterwards where people can constructively criticize and map out weaknesses in each approach and systematically build each other up.

For what it is worth, Martin Rinard once told me "web programming is a solved problem" (when I was visiting MIT last year, although he didn't know who I was when somebody at MIT introduced us briefly), so it would be pretty funny to see, say, a panel on web programming at OOPSLA.

[Edit: Fun fact for those of you unaware. One of MIT's hardest undergraduate courses (and I am told most rewarding), is Software Engineering for Internet Applications. It has one of the heaviest courseloads of any undergrad CS class (this is based on stats reported to me by a grad student at MIT). When professors at Nassau Community College asked me what to base a two-year degree in web software engineering on, I told them that MIT's course could probably just be expanded into two or three junior college courses. I also pointed to WebDAV creator James Whitehead's proposed Masters in Web Engineering curriculum.]

Also, ACM Membership comes with Portal access, as well as access to tons and tons of Safari books.

Combining a wrong with a right doesn't make the wrong right.

Things need to be financed somehow

What needs to be financed? The hosting costs? I would gladly pay $1 per gigabyte of papers that I download. It is mind boggling that institutions pay serious money for the journals that they provide with content for free.

If you are a professor who lectures students, you really should emphasize ethics throughout,

By telling your students that they should never publish in this kind of journal. It only serves to fill the pockets of the journal at the expense of keeping knowledge from the poor.

By telling your students that they should never publish in this kind of journal. It only serves to fill the pockets of the journal at the expense of keeping knowledge from the poor.

But even printed copies of OOPSLA are under $100! What's more, is that after 10 years, most libraries that buy these journals are now unloading them for around $5-20. I actually recently bought the first 10 years of OOPSLA journals on Amazon.com for $300. You probably couldn't even purchase a catalog from Elsevier for that "little".

Even Chemists who have decided to create their own journal and publishing company cannot beat the price of OOPSLA journals from Springer-Verlag. Yet you can also get OOSPLA online as an e-journal and only pay ACM membership fees, giving you a bunch of additional access to grow your skills with other sources of knowledge.

Bottom line: ACM is doing a lot of things to help push costs down, not up. ACM's size gives it economies of scale, and by becoming a member, you add to that economy of scale.

the expense of keeping knowledge from the poor.

Now that is a separate idea. How to help the extremely poor in third world countries? ACM could donate to the One-Laptop-Per-Child campaign a free lifetime subscription to ACM. Every child that gets a laptop would have virtually unlimited scientific information. I'll e-mail Wendy Hall this suggestion right now.

You are missing my point I think...there is no doubt that an ACM membership is a good deal in many situations. The question is: what is the point of filling the pockets of the ACM? Why not provide electronic versions for free? The costs are negligible.

$100 is ridiculous. I can get magazines for $2. And unlike OOPSLA, these magazines do not get their content for free. What is it exactly that OOPSLA does to add $98 of value?

The ACM does not get paid for the printing or for the hosting costs; these costs are orders of magnitude lower than what they charge. They get paid because they lock down scientific knowledge that should be freely accessible.

This is why Marc Hamann said this is a complex issue, and rightly ripped me for being a bit of a hypocrite.

Magazines subsidize their costs through advertisements and the fact that they have focused readerships. They discount subscriptions to improve their statistics on what kind of readership their advertisers are getting. Advertisers love 'narrowcasting', and very few situations in advertising lend themselves well to 'broadcasting' (such as the Super Bowl).

You don't have to buy OOPSLA in print. If you own a Kindle DX or some other e-reader w/o DRM, then you can read OOPSLA on that device. Then your cost is simply ACM Membership, which also looks good on a resume (the majority of programmers do not belong to ACM).

I am not totally aware of ACM's cost structure, but I imagine it is a lot like any organization in that it often takes that extra income and devotes it to initiatives that otherwise wouldn't be funded (such as sponsoring scholars). Printed journals are a separate matter, since much of the inflated cost goes to the publisher's costs and net revenue.

Whether it's good to have an ACM membership is irrelevant. We should not compare member of the ACM vs not getting access to papers. Sure, in that case it's good to be a member of ACM. We should compare having to pay for locked knowledge vs free knowledge. Remember that the government and thus the general public pays a large amount of money to scientists to generate knowledge, so the knowledge should belong to the general public, not the ACM. In my opinion the government should only pay scientists for work that will be publicly available.

Can we please all post links to citeseeer. ;)

Academic knowledge should be freely available. That's the whole, and often only, point of a library. ACM is a supplier for libraries, academics are suppliers of ACM. The question is not how to build a bussiness model, but how to break the circle.

To be honest, what's the point of hard-copy ACM nowadays anyway? Does anyone even bother to ever walk the steps into a library anymore? The only point of the ACM is that you get a paper accepted in a well-seen journal. So, it's only accreditation people are after - and not even of the ACM, but the people in a committee the ACM employs.

I guess academics need another way of forming accreditated committees, and another free-as-in-speak hosting provider. Which, probably, will end up to be either Mickeysoft or Google. You can always license the ACM to print after that.

There was a response from Moshe Vardi on Openness and ACM last summer.
Open, Closed, or Clopen Access?

As for ACM's stand on the open-access issue, I'd describe it as "clopen," somewhere between open and closed. (In topology, a clopen set is one that is both open and closed.) ACM does charge a price for its publications, but this price is very reasonable. (If you do not believe me, ask your librarian.) ACM's modest publication revenues first go to cover ACM's publication costs that go beyond print costs to include the cost of online distribution and preservation, and then to support the rest of ACM activities. To me, this is a very important point. The "profits" do not go to some corporate owners; they are used to support the activities of the association, and the association is us, the readers, authors, reviewers, and editors of ACM publications. Furthermore, ACM operates as a democratic association. If you believe that ACM should change its publishing business model, then you should lobby for this position.

There's a new article on the subject in this month's CACM. The good, the bad, and the ugly.

Ridiculous.

But the proponents of Open Access quickly realized that online publishing is not free, nor cheap. [...] But how much are authors ready to pay to publish an article? A few hundred dollars? The most prominent Open Access publisher, the Public Library of Science (PLOS), is a nonprofit organization that has received several million dollars in donations. Yet it charges between $1,350 and $2,900 per paper, depending on the journal.d In fact, many in the profession estimate that to be sustainable, the author-pay model will need to charge up to $5,000–$8,000 per publication.

So they are saying that you need to buy 5 servers *to host ONE paper*!.

Lets say one paper is 1 megabyte (usually more like 100 kilobyte). You can easily get 200 megabyte storage and 5 gigabyte transfer for 1 euro/month. So you can host 200 papers for 1 euro/month. Even if you multiply this by 100 to be safe, it's still three orders of magnitude less than what that article (i.e. the ACM) wants you to believe.

I bet that many companies would be happy to donate free hosting for papers.

I think those costs include professional editing, getting the latex files into the format, making sure the images are sufficient quality, etc. that produce the document that will eventually be hosted. Of course, as long as the paper is released under Creative Commons, the cost of these other services can differ from journal to journal while papers are hosted elsewhere.

I remember reading this article, but I couldn't find it on Google. I wanted to link it earlier, but just couldn't find it!

First, we're talking about conference proceedings here, not a journal. To produce the final product that you see, this involves substantially less investment in critical services per article.

Obviously this debate is far from settled, but the most appealing suggestion I've seen is charging authors of submitted or accepted papers a small fee, which should be more than enough to cover perpetual, redundant hosting of papers online. To me, hardcopies are irrelevant, so online distribution and the personnel supporting it are the only serious expenses. We might want to consider explicit ways to compensate program committees for their time, but that isn't happening with ACM's current practices, so I don't think a credible proposal needs to address it.

Actually, I am able to download the PDFs without any mention of signing up for an account with them.

Probably because your institution is already an ACM subscriber. At work I can use the ACM, at home I can't, this is kind of annoying.

I understand the need for the ACM to make money, but as a resource its not very useful, since (a) its not very comprehensive (only hosts ACM papers, not Springer papers...) and (b) the paywall is inconvenient if not a huge barrier. Most modern papers are available somewhere else, so citeseer or even just Google are often more comprehensive and useful.

I find the whole Open Access discussion pretty ridiculous.

It can be dangerous to form an opinion without understanding the whole system within which papers are being published. This isn't just about access, but also about legitimacy. Publications are the basis upon which all academic hiring and promotion is done, and that to "really count", the publications need to be accepted by a peer-reviewed medium with some "standing".

This is an important issue, it is very complicated, many thoughtful and informed people are thinking about it, and there are no easy answers to it, nor any consensus yet.

This is something only people can dream up who are born a couple of decades before me.

It is hardly the case that no one over 30 has grokked cheap web-hosting. Is it possible that it's not that those "born a couple decades before you" don't understand the things that you do, but that they also understand something that you don't?

I have found that the more "many thoughtful and informed people are thinking about it" you have, the less results you get.

I do understand the whole system, by the way.

Again, the fee you cite is unrelated to a hypothetical author fee for one of the main SIGPLAN conferences (ICFP, OOPSLA, PLDI, POPL). These aren't journals, and, if we drop the paper proceedings, no one is getting paid to do anything but keep the author's PDF accessible online. I think something more in the range of "a few dollars" than "a few hundred dollars" would be more than adequate.

... that the conventional thinking about posting proceedings is correct, and we merely need to provide a free access mechanism.

I firmly believe ACM should operate more like archive.org. If a paper includes in its reference section a timestamped URL access reference, then the proceeding should include an immutable archive.org-like URL to that timestamped URL access. This includes things like binaries and source code.

There are some library science/information science researchers at UMich that think we need more than just a DOI. We need to use that DOI and provide much more richness to potential readers.

The extended scope of this year's OOPSLA bothers me a bit, from the CFP:

This is a great point, and it did bother me too. How can you get enough expert reviewers for papers if you spread yourself too thin? Ultimately, it just waters down the review process. There could be a lot of great papers that get rejected, simply because the depth of the paper is too much for the reviewers to take in given their particular background.

So the main issue with scope is not "What kind of conference is OOPSLA now?" That's the wrong question, Sean. The main question is, "How can you possibly have enough qualified reviewers to ensure you treat my submission fairly?"

It appears the answer is that the author should write their paper as if the target journal was Nature, Science, or Scientific American. Just an observation and not intended as a value judgment.

I'm sure the reviewers will be from the same crowd that usually review OOPSLA papers, which have always been a bit inter-disciplinary: you'll get some reviewers who do ICFP and OOPSLA, or PLDI and OOPSLA, or ICSE and OOPSLA.

My point is that...OOPSLA's grown scope expands the tent but at the necessary expense of quality. I would prefer an ICFP-sized high-quality conference on object technology, but maybe that's the problem: what is the scope of "object technology"?