A new version of the notes is available at the usual place (the February 21 2024 version). Everything is in potentially nearly final form for an “official version”, except for typesetting/formatting and the index.
(a) Significant change, but not substantive: The figures are now basically done (meaning: redone). Because I think people should be comfortable making their own sketches in real time as they figure things out, I’ve deliberately gone all in on a hand-drawn aesthetic. This is atypical, even amateurish, for serious mathematics books. So maybe I will reconsider.
(b) About composition of projective morphisms (the old Exercise 17.3.B), and more generally 17.3: I realize now what the complication was in the old 17.3.B. People were stuck at many steps, but the real issue was only the last one, to get to the quasicompact target case once you already had the target line bundle. There was indeed a gap there. Given what is done in the notes (and not even by this point in the version previously posted), we can show it only in the Noetherian case. So the new version has 17.3 seriously rearranged in a number of ways. In particular, now the old 17.3.B is a bit later, and the argument is for when the final target is either affine or Noetherian. (Even this requires as a black box something that will only be proved in the cohomology chapter, which is Grothendieck’s coherence theorem for projective morphisms.) I think this is now rigorous and complete. Please let me know if there are issues. (Some of David Speyer’s ideas also in retrospect guided me on how to improve it.) I’ve moved all the double-starred bits to the end (and I hope the reader ignores them all). There is a single-starred section that is where the trouble lies, and that’s going to be hard going for those readers working through it.
(c) In the chapter on the 27 lines (you know which chapter number it is), I was always unhappy about needing Castelnuovo’s criterion. János Kollár pointed an explicit workaround that I like a lot. (I know that, roughly, doing it in this hands-on way, is very classical; but it is hard to do it rigorously without hand-waving, and you’ll notice this hand-waving in some expositions you may have seen.) This is now Proof 2 of Proposition 27.4.1. His explanation to me was direct and to the point; I’ve muddied it a little to fit the narrative, so the “worsening” is due to me. (Most mathematicians have their own particular kind of thinking, which they are best at. Kollár has this rather amazing ability of understanding very abstract things and very concrete things, both as well as anyone else, and those two things seem to be connected directly in his head in a way that they are not for most algebraic geometers. When I was in grad school, we secretly called him “The Mighty Kollár”, and even now it doesn’t seem an inappropriate name, although I would never say it to his face.)
General philosophical point: I’ve noticed that there is a tension between the kinds of requests people have of the notes. Roughly, on one hand, there is my desire to try to make it possible to cover some central core of the material in a year (for at least some people), which requires rather severe compromises. I’m trying to make different compromises than most people have made in the past (in particular, I’d like to expect less from the reader in terms of background, but then I need to expect more from the reader in other ways). The things I need to push back against are things like “This topic really needs to be included”, “This topic needs to be fleshed out more completely”, “This topic is not done in sufficient generality”, “This topic needs to be done more rigorously”, “The presentation of this topic is well tuned to me as a reader”, “I coudn’t solve this exercise”. In all of these cases, the suggestions are good ones, but at some point the manuscript will sink under the weight of items loaded onto it. Usually those making suggestions are happy to suggest what other things should be cut, but you might not be surprised to hear that these suggestions contradict each other. I’ve tried to help by starring and double-starring some topics, but I can’t seem to stop some readers from not skipping them (and then getting boggeddown). I’ve had a number of suggestions of things that “really are needed in such a work” that I’ve had to repeatedly decline. Even the ones I have said yes to have let it creep up to 850 pages even after the I passed a secret “no new material” line in my mind. (Some recent additions: on top of the ones mentioned above: a brief mention of projective normality; definition of Cartier divisor in generality; and more. But even these are things I think the reader can quickly read on their own from their web having read these notes, and needn’t be here.)
One form of the compromises I’ve made is: “If we need it, I don’t want to black-box it, and I want you to understand it, but I only need you to understand it well enough to use it and move on”, and “if we don’t need it, no matter how wonderful it is, we should just skip it, and you can learn it on your own later” (so if included, those things are starred or double-starred). I’ve had a hard time maintaining this position consistently, and over time a lot of things have slipped by my defenses.
At this point I am still entertaining all sorts of suggestions, but am going to try to stick to things that particularly deal with mathematical errors (often leaving imperfections and imprecisions — and many of these were actually deliberate choices), or really affect the understanding of a significant portion of readers (which I have some broader sense of given comments over the years) and not just you personally.
Many of the recent comments (including some still for me to think about) are here on this website. Some excellent ones have come to me by email, from a group of students in Poland, by way of Joachim Jelisiejew. I want mention them here, and I also look forward to seeing what these students go on to do mathematically in a few years’ time, because they are clearly very talented.
February 26, 2024 at 7:46 pm
Is that going to be the actual cover?
February 27, 2024 at 5:30 am
It’s something I was playing with at one point, which makes me curious what people think. The picture certainly isn’t traditional for what a cover is supposed to look like, and may annoy some people. (The images aren’t completely randomly chosen of course.)
February 27, 2024 at 5:33 am
I personally adore it! My take is that since the book itself oozes personality so should the cover.
February 27, 2024 at 8:37 pm
+1 The cover looks beautiful
February 29, 2024 at 2:00 am
I also like the cover and at any rate I do not think that the book requires a neutral cover to claim authority. It would be great if you could manage to publish it in one volume: given the many cross references, it will facilitate reading and working with it. With a lighter grammage of paper (50-60gr*m^-2 ) this would be feasible. An example – also for an unusual cover – is the remarkable “Grimoire d’Algèbre Commutatuve” by Lorenzo Ramero.
March 9, 2024 at 7:12 am
I didn’t know about Ramero’s Grimoire, and it is really beautiful. I’m going to order a copy on Monday, and also I have also printed the cover to put on my office door for a bit.
Here is where you can buy a copy yourself. Here is link to the entire Grimoire project.
March 9, 2024 at 7:49 am
I forgot to also mention: yes, definitely in one volume, for the reasons you mention. It will be a bit hefty, but hopefully still something that can be comfortably carried about without being a burden.
March 4, 2024 at 7:36 pm
I would also like to express enthusiasm for this cover! Commonly said metaphors aside, having a math book with a nice cover really makes it stand out.
March 9, 2024 at 7:09 am
Thanks to you and others for the positive comments — it has made me now think about how to possibly make this work. (It would be likely something unofficial I would do, separate from the publisher, which will require some cleverness.)
February 27, 2024 at 8:43 pm
p. 121, Ex. 3.6.J(a), the Hint mentions the “residue field”, which won’t be introduced for another 19 pages (according to the index at least). Speaking personally, I would recognize the term as referring to the quotient of a ring by a maximal ideal, but what’s missing here is which ring and which maximal ideal.
February 27, 2024 at 9:57 pm
Seems like on P141 4.3.F. (b) the function should be “away from x-axis”? Or am I missing something. Also if anyone happens to have an official solution to 4.1.C. or previous counterpart please let me know, thanks!
April 30, 2024 at 9:51 pm
For 4.1.C, as with 4.1.B, I think the “right” way is to show that the restriction of the presheaf-on-a-base determined by A to any distinguished open subset D(f) is isomorphic to the presheaf-on-a-base determined by A_f. Once you have that, it’s easy to see that the statement of the base identity or base gluability axiom for any open cover of D(f) by distinguished open subsets of Spec A is exactly the statement of the same axiom for the corresponding open cover of Spec A_f by distinguished open subsets of Spec A_f. The preceding proof covers this statement, so we are done.
February 29, 2024 at 10:38 pm
This is a subtle but I think important nitpick. In 1.3.D, the two “translations” describe a universal property that is ostensibly satisfied by the identity map A -> A. In order to characterize S⁻¹A in these ways you need to require that every element of S is invertible in S⁻¹A. This is implicit in the defining universal property at the beginning of the exercise, because S⁻¹A is implicitly *among* the family of A-algebras B where every element of S is sent to an invertible element of B, but not really implicit in the translations. I expect it could be argued that the translations are meant to be interpreted in the context of the defining UMP, in which case you could also argue that invertibility of the elements of S remains implicit. But I think it probably would be better to call it out explicitly.
March 6, 2024 at 4:19 pm
that would be a stunning cover!
March 8, 2024 at 8:23 pm
Nitpicky nitpick: on p. 152, right before 4.5.H, “The points of Proj S• are the set of homogeneous prime ideals of S•…”, maybe just “are the homogeneous…”? Stripped of all the decoration this sentence has the form “The points are the set of ideals”.
March 8, 2024 at 11:17 pm
Oh sorry, I already pointed this out and you fixed it. I haven’t switched to the latest and greatest version yet.
March 9, 2024 at 6:02 am
Thanks Mark!
March 9, 2024 at 7:21 am
Hi, professor Vakil! Thank you very much again for the great book! Here are some more issues I’ve come up with after the previous batch:
13.3.2: In ‘We say p is node of C if … the map … is rank one’, should ‘rank one’ be describing the kernel of the map, rather than the map itself?
28.3.I: This exercise confused me. What does ‘Assuming it is possible’ mean? For C-curves, isn’t it always impossible to draw sketches on a paper? For an elliptic curve with suitable j-invariant, if we only consider its intersection with the real plane, it would be topologically the same as a straight line in the projective plane. In such cases how should we count holes?
28.4.1: The notation here seems ()^_q is evaluated first, before R^i pi_* . Perhaps it would be clearer to add parentheses here?
In the (Do you see why?) part, I think this is ‘Natuality of push-pull diagrams’. However, that seems unproven, and maybe takes some effort to prove.
28.5.B: Here 10.7.P has unnecessary assumption of k-schemes, making it insufficient for this exercise.
28.5.C: I suppose ‘contracting’ here means contracting to a point, but it could lead to confusion. For example, blowing up a line in 3-space could be considered contracting a 2-dimensional subscheme to a line. Also I think maybe connected isn’t needed as condition here, since the proof ensures it anyway.
28.6.F: In order to use Nagata’s Compactification Theorem, should pi be separated?
28.7.E: I think it might be helpful to move ‘remains exact upon taking global sections’ in 28.7.F here, according to my understanding of the usage of (28.7.1.2) in proving isomorphism as sets.
And here are some previously unreplied questions:
13.4.2: Bertini’s Theorem stated here seems not general enough for subsequent usages. I think a sufficiently strong version should cover the case of a smooth subvariety of P^n_k of pure dimension d, intersection m hypersurfaces whose degrees are fixed respectively, and still have an open locus in the parameter space with smooth intersection of pure dimension d-m.
[Response still to come. -R.]
20.1.C.(b): What does generally choosing mean here? In the usual sense, i.e., an open locus in the total parameter space, we will in fact have a regular sequence thanks to Cohen-Macaulayness. However, not having C-M-ness at this stage, a less general approach is to choose these hypersurfaces one by one, each arbitrarily in an open locus in its own parameter space determined by the previous ones. This will be less general because open subspaces in fibers in a fiber bundle don’t necessarily give an open subspace of the fiber bundle, however it will be more doable to readers at this stage. I hope there could be some clarifications here.
[Response still to come. -R.]
Hi, Professor Vakil! Thank you very much for helping! Here’s further explanation for question about 20.2.19:
Consider the ‘If A * L <= 0, then we are done without even using H'. Can this be understood that if A * L < 0, then without using any conditions about H, we will have the result that L * L <= 0? But this is not true, for example take L = -A .
Hi, Professor Vakil! Thank you very much for helping! Here are some further clarification requests to your response:
21.6.E: Here do we have the condition that X and Y are smooth? It isn’t mentioned in the statement, but appears in the hint.
[Right! Here is a further hint: they are “mostly smooth”. – R]
[[Is it true that in this exercise X and Y are smooth? If so, why do we need the hint? If not, what does ‘Right!’ mean?]]
18.4.L. How could one write L into the form of O(sum n_j p_j) where are p_j couldn’t lie in a specific finite set of points? After counselling one strong friend of mine, I got this approach:
a) The target is to construct a rational section invertible at each p_j.
b) Divide L into quotient of two very ample line bundles. WLOG we may assume L is very ample.
c) If k is infinite, we may choose a suitable global section not vanishing near these points.
However, this approach requires the hypotheses that k is infinite. Even if we could base-change to infinite k in first place while preserving degree, since it is unmentioned in the hint, I don’t think it is the intended solution. What’s the intended way to do this?
[Here is a hint. If you have finitely many points in projective space over a finite field, you are right that there need not be a hyperplane missing them all. But (you can show that) there is a positive degree hypersurface missing them all! -R]
[[I’ve thought about this idea previously. However, positive degree hypersurfaces don’t give rational sections on the line bundle we care, which is assumed to be O(1). I met serious obstacles trying to overcome this. Moreover, I think a hint might should be better added: There has been many pages since the idea of factoring a linear bundle as quotient of very ample bundles and this exercise. And the idea of factoring as quotient of very ample bundles hasn’t got much usage, so readers might forget this is a thing at this stage. It happened to me and I had to seek help from a talented peer student. ]]
Thanks for the rewriting of 17.3, it is much easier to follow now! Regarding your note about focusing comments on mathematical errors – please let me know if any of my suggestions fall outside that scope. I'm happy to adjust my approach based on your guidance. Thank you again for this wonderful book!
March 9, 2024 at 7:29 am
Song Ye wrote to me: “I believe that the two relative tangent spaces upstairs are mislabeled in Figure 21.2 in your notes. I think you want T_{X/Z}p to be the whole thing and T_{X/Y}p to be the linear subspace.” Thanks! That’s completely right, and is now fixed. I hope I didn’t make mistakes of similar quality in too many other re-drawn figures!
March 9, 2024 at 11:19 pm
In 4.5.L, the third condition is “√I = S₊”. This is certainly false, e.g. take a graded ring for which S₀ is not reduced and take I = S₊.
This exercise dates back to the very early days of the book, and until the August 29 2022 version the corresponding condition was “√I ⊃ S+”, and it is straightforward to show that this *is* equivalent to the other two. So I suspect this got changed inadvertently somehow.
March 9, 2024 at 11:29 pm
Oh, and in part (b), I think technically you should specify that the fᵢ are homogeneous.
March 10, 2024 at 11:29 am
I’ve also been reading Görtz and Wedhorn’s Vol. II recently and they seem to have a pretty neat way of handling errata: https://www.algebraic-geometry.de/errata-vol2/
@ravivakil, since you’re going to be stepping away from this project for the next few years, something like this might be useful to keep the errors readers highlight in one place (and avoid different people reporting the same error).
March 14, 2024 at 11:20 pm
Hi Ravi. After the sequence (6.1.2.1), you say
“(Equation (4.1.6.1) shows how the map should be defined, with approrpiate signs”
But 4.1.6.1 doesn’t point to an equation.
March 26, 2024 at 10:18 pm
Nit: p. 125, first line, “irreducible subsets” => “irreducible closed subsets”.
April 3, 2024 at 4:41 am
Dear Ravi,
my impression is that in 16.4.1. (universal sequence of the Grassmannian) a reference back to “universal object” (i.e. 1.3.Z.b)) would be beneficial to understanding. Reading in a non-linear fashion made me unaware of this small definition, and had me spend quite some time scratching my head.
Best, Clemens
April 13, 2024 at 5:33 am
I’m late to the party, but there’s a typo “R” in place of the ring A in the proof of Lemma 12.3.9.
Oh, I see you’ve caught the other error in the proof from my older version (shouldn’t q_{j-1} contain, rather than be contained in, (f, q_{j-2})
April 15, 2024 at 7:03 pm
Small nitpick: in 4.5.13 you say (“Here k can be replaced by any ring A…”). You probably should also say that in this case V is assumed to be a *free* A-module on n+1 generators.
April 19, 2024 at 8:59 pm
Potential typo in ex. 14.6.K: I think “Y'” should be “X”.
April 25, 2024 at 11:47 pm
Another small nitpick: in 4.5, when you define homogeneous elements as elements of some S_n, I think you probably want to include the condition that homogeneous elements are nonzero.
April 30, 2024 at 9:54 pm
In Figure 4.1(b) and (c) on p. 133, the exponents l₁ and l₂ on g₁ and g₂ shouldn’t be there (they’re part of the definition of g₁ and g₂).
May 9, 2024 at 11:02 am
Hi there! Noticed a small detail on p. 108 of this draft: in the third last line, where it says “can be said to depict k[x,y,z]/(x^2+y^2-z^2)”, I believe it should instead read “can be said to depict Spec k[x,y,z]/(x^2+y^2-z^2)”. I’m enjoying the book so far! 🙂
May 13, 2024 at 2:50 pm
Thanks, fixed! I can’t believe that one lasted that long…