Yeah, you triggered me so hard I smashed my glass and went on a monkey rampage crying about your privilege.

No, I just mean everybody always has opinions about why this thing in math is whatever, and half of them are contradictory. Quoting the age-old sage wisdom, "opinions are like *******s, everybody has them and they all stink".

Wow, somebody's bitter.

I am pretty sure Danny was just finding a way to humorously justify his own research program.

Anyway, when you expressed frustration with group theory, I was reminded of Danny's characterization of the study of finitely presented groups as "full of monsters", with almost all interesting questions unsolvable, and how he contrasted the dire situation along that line of research with his own study of "typical" groups, which are much nicer (he says that a typical group is "hyperbolic", which apparently is "the best thing a group can be").

And as a joke, he named these two visions GGT: "Grumpy Group Theory", and "Generic Group Theory".

At any rate, I probably could have avoided setting you off on your little rant about how "people shouldn't have opinions about math" (and apparently not even about their own research program???) by starting the video a few seconds earlier, where he says:

And yes, I think triggered is the right word here. It sounds like studying group theory has left you with some kind of PTSD (which is where feminists got the word from).

It's more just boredom of meta-level discussion of math. I've been around it for a few years now and used to participate in those kinds of discussion online.

Math is becoming slightly more "job-like" in a sense, so I'm starting to avoid it more in my casual conversations.

Apologies for being snappy.

Once quals are done, which will be tomorrow, every ounce of stress will shift to preparing to instruct calculus next week. I have to prepare a course plan and write a syllabus. ****tt.

I completely sympathize. But I still think your response had absolutely nothing to do with what Danny says in the video. And if you really did watch it and still think that, I think you've become a little bit jaded.

The one that you started.

Annnd topology is done. I was stressed beyond worry about differential topology because I often struggle with it. I also understudied it. Fortunately, just knowing the IFT's, Sard's theorem, stuff about regular values, and related definitions helped me stumble through the proofs at least somewhat competently. I made a few false assumptions, so that's going to hurt a bit. Hopefully that plus my generally on point algebraic topology answers will be a passing grade.

hang in there

Thank you. All I have to do now is begin preparing for instructing a class and begin reading for my other classes. Oh, and get drunk. Much lower stress.

Alcohol is detrimental to memory, but I suppose that's sometimes what we want.

Not alcohol is detrimental to sanity. It's a balance.

So that's what I've been doing wrong all these years...

https://perens.com/2018/08/22/new-intel-microcode-license-restriction-is-not-acceptable/




I guess Intel is pretty scared. This is, what, the 3rd or 4th microcode patch in the last year that drags down Intel CPU performance? Obviously someone is going to publish these benchmarks anyway, and I'm both excited and terrified to see how bad this is going to be given they're threatening to sue anybody who finds out.

Sounds to me like either the lawyers at Intel have no clue about what they are doing, or this is purely an intimidation tactic made in desperation.

The only aboveboard explanation I can think of is that the license was accidentally repurposed from a pre-release version. But honestly, given Intel's recent choices and the kinds of customers most affected by this side channel attack, I would be surprised if it weren't the latter.

I've been reading that recent AMD CPU's have a better performance/price ratio.

I've always been partial to AMD because motherboards for AMD processors sometimes make it possible to run them completely without any proprietary BIOS code at all. I don't think there have been any recent Intel processors where this is even possible.

Basically short version of the story is, certain Intel customers (the Googles and Amazons) have been using their monopolies to pressure Intel to deliver exclusive advance access to their upcoming Xeon chips. They want a perf/watt advantage over their smaller competitors and have been threatening Intel with developing their own ARM server chips to get it.

(Intel's also been pushing this exclusive advance access up the supply chain, which is why I heard about it. It's not secret now but it was when I first heard.)

So here's the problem for Intel.

Most of these recent vulnerabilities - this L1D side channel exploit, Meltdown, etc. - don't really affect we prole end users. Home users will probably never need or want to install this most recent microcode update, because our machines are single user or trusted multiple user. Previous patches only affected syscall heavy workloads, like I/O intensive stuff. The kinds of workloads where you might, just hypothetically speaking, need to purchase a Xeon Platinum for the extra PCI-E lanes. Just for example.

So for the kinds of customers who've done this back door, maybe-antitrust-regulators-ought-to-look-at-this-please deal, it's getting worse. These microcode patches aren't slowing the chips down like... it just takes longer for a certain instruction to finish, or it needs to wait longer to get stuff from memory. They're making the processors do a lot more work and consume a lot more energy to do the same work they used to do. Throughput down, latency up, perf/watt down down down.

No idea what this'll all mean for Intel. Probably nothing good though. Xeons are obscenely overpriced, but for the kinds of high volume companies like Google and Amazon, the price basically does not matter - they want performance/watt, they're limited much more by waste heat than by petty cash, so they're willing to spend any amount to get it. $16-$20k per CPU, quad socket motherboards, whatever. They just don't care. Intel currently dominates the server market and therefore can charge basically whatever they want to Googles and Amazons as long as they continue to cut down perf/watt. The server market is absolutely critical for Intel's survival as a company.

The more algebraic **** you learn in math the more you learn to appreciate category theory. It really boils down to this:



Math without categories



Math with categories

Not even going sky-high with the meta-level category theory BS, it's really convenient to have universalized notions. It's nice to see, for instance, the universal property appear in a category. You just know how it works.

A d j o i n t F u n c t o r

You're the morphism now, dog!

Did Halmos do work in category theory?

(That's Sammy Eilenberg.)

Oh, he looks like Halmos. Eilenberg is definitely the category theory guy..

Halmos would have probably preferred to work directly with matrices: "We [he and Halmos] share a philosophy about linear algebra: we think basis-free, we write basis-free, but when the chips are down we close the office door and compute with matrices like fury."

http://www-history.mcs.st-andrews.ac.uk/Quotations/Kaplansky.html

Also, that's a picture of Sean Category from the little known movie Finding Formalism

I stumbled upon the obituary of Peter Landin, of "The Next 700 Programming Languages" fame (a paper which seems to have stimulated the invention of modern, non-Lisp functional programming languages, such as ML and Haskell).

I was amused to read the following bit, though:



...BUT (emphasis added):



Instead leaving him time to tend to more important matters:

People who have developed an open-source project, I have a question. There's a package for SublimeText that has a feature I really like and would love to use outside the context of the that package (specifically it's the live preview of math for LaTeXTools, outside of .tex files). However, LaTeXTools has no license on the repo, so I don't know what the etiquette is for reusing the code from the package. Obviously I could do it locally and no one would care, but if I can get it working, I think it might be a useful tool for other people.

My larger vision is to use the inline math preview with the PlainTasks plugin and have a nice todo/notes file that I can write my daily notes in, which frequently include a few equations/mathy expressions.

Send an email to the owner of the repo?

The short version is, they catastrophically ****ed up.

There is no license. By default that means all rights have been reserved by the original authors, which means technically you don’t even have the right to use or redistribute this software. You’d need to contact every single past contributor to secure those rights.

Likewise, in order to apply a license now they’d need written permission from everybody who’s ever contributed to this repository. For a project with no CLA that actively solicits anonymous contributions, this is impossible.

https://github.com/SublimeText/LaTeXTools/issues/1175

lol **** me

Holy ****, lmao

Looks like we lost the author of TempleOS to mental illness.

(N.b.: strong / racist language in that thread is a reference to Terry Davis' frequent paranoid outbursts about those who were out to get him, probably due to his schizophrenia).

I never tried it, but it looks like a neat little OS in the spirit of Amiga and other minimalist 80's home computer systems.

Relevant insight to Eversor's point about racism being about intent (contrary to the now popular definition of prejudice based instead on words rather than actions):



https://news.ycombinator.com/item?id=17913367

Actually, on the other hand, I'd be interested to hear what Eversor thinks of this guy's idea about how to respond to racist comments:



It sounds like he's saying that intent doesn't matter, because we can never really know what the person intended. But then that it is still necessary to chide them for having made a racist remark. But then, is there any room for comedy? And who defines what is acceptable comedy, if we can't trust ourselves to correctly interpret their intent?

maybe the tl;dr is that we can always let people get away with saying racist things so long as they are either schizophrenic or simply assure us it's part of their comedy act

Site says "don't reveal your power level", contains dozens of people being "ironically racist"

yuh huh