Compressing a huge paper for a double blind review process

Question

Continuing How to cite a technical report for a double blind review process?, how would you proceed if your paper is acutally 53 pages long, most of it occupied by mathematical proofs, but you have to compress it into 10 pages (including all the appendices, and allowing for two additional pages for the bibliography) for an A*-level conference with a double-blind review process, a rebuttal and an explicit ban on relying on supplemental material? The time between the submission and the 2-day-long rebuttal is 2.5 months, which is then followed by a month till notification.

Here is a part of the guidelines:

Do not rely on supplementary material (your web site, your github repository, a youTube channel, a companion technical report or thesis) in the paper or in the rebuttal submitted during the clarification period. Supplementary information might result in revealing author identities.

Or would you consider not submitting to that particular conference, since it involves too much work which does not improve any of your results?

Several technical options are available, but I'm unsure about their acceptance in the community:

1. Use the PDF file with the main paper text as a container for the anonymized proofs as an embedded file, or

2. Upload the anonymized proofs to a (megaupload-like) file-storage service and provide the reader with a URL, or

3. Any of 1. and 2., with password-protected supplemental material, where the password is exposed with/after the rebuttal.

Are such solutions accepted by the community?

Answer 1

The real problem with such a paper is that conferences have deadlines. Therefore the review process needs to be expeditious. Journals can in most cases handle long papers since they don't need to promise when the paper will appear and try to have a big enough backlog that they can fill an issue with interesting work. But if the call asks for 10 page papers that is what they really are depending on.

I don't think the "method" by which you make the longer work available is the issue. You can probably find a way to make it blind. But the length isn't copacetic for purposes of their very hard deadlines.

For the conference, write something else. Or write some small excerpt of the longer work that fits the expectations of the committee.

But if you prove a long standing and important theorem, a short, even one page, announcement is probably enough for a conference, ending with "Proof will follow." You will want to include some information about why you think your proof stands up - vetting by some other folks. If you want to get bold, mention the key element of the proof (the crux) that made it come together. Now you are up to around, say, 10 pages. All is good. The full proof can appear later in a more appropriate journal.

Answer 2

Several options:

1. Split the paper into several parts, submit only the first part. Not every paper can be split up like that, but often they can. If you're not sure where to make the split since you're too invested in the "wholeness" of the paper - ask a colleague who has taken a look at it.
2. It's the wrong venue for your paper - try someplace else. Perhaps a journal which takes really long works? Or maybe extend it into a monograph?
3. Ask for a special exception from the PC. Probably won't work, but who knows, right?
4. Guerilla publishing. Print a "draft" of it as a booklet, attend the conference, distribute it to people. Also, put a draft online somewhere and pitch it to people you know / sort-of-know with a link.

I wouldn't try any of the three complicated tricks you've listed.