3 ms·
> But for longer works where the proof is decomposed into multiple lemmas and sub-lemmas, these can be very useful. Wouldn't they be better off as an actual ta
by jgwil2 5y ago
> But for longer works where the proof is decomposed into multiple lemmas and sub-lemmas, these can be very useful.
Wouldn't they be better off as an actual table of contents though, rather than in paragraph form?
- tikhonj 5y agoAt least in CS venues, page limits disincentivize formatting that takes up more space—even when it's more readable.
- Ar-Curunir 5y agoanother reason why we should get rid of page limits
- jmmcd 5y agoI don't want to review for a conference with no page limits.
- spekcular 5y agoNo, because it's hard to accurately represent the tree structure and interdependence of lemmas with a linear table of contents. (Sometimes – not always!) But for really nasty proofs, I think a graphic description of the structure is best. Check out Figure 3 on page 29 of this paper: https://arxiv.org/abs/1108.2291 https://arxiv.org/abs/1108.2291.