Hacker News new | past | comments | ask | show | jobs | submit | fortenforge's comments login

One of the cooler features in Windsurf is Cascade, sort of an agent powered by an LLM that you can talk to and can make changes to your codebase.

Here's an example of me using Cascade to solve a CTF challenge: https://youtu.be/LbYepFmVB20


Eh not quite. Famously, you can fork VSCode, but you can't use the VSCode Extension Marketplace if you do, which loses a lot of the network effect benefits of the VSCode ecosystem. (As far as I know Cursor is flat out violating Microsoft's terms of service with respect to the extension marketplace).


And a lot of the licenses for flagship Microsoft VSCode extensions for languages like C/C++ and Python don't allow using them outside of VSCode/Extension Marketplace so open source forks are crippled by default.


I believe this also blocks you from using Microsoft's proprietary language extensions, and they have been steadily switching the default language packages from OSS to proprietary.


Yes. You famously cannot use the C/C++ language server bundled in the C/C++ extension or Pylance. Who knows what other development tools they will lock behind their fork to the detriment of open source communities. Also you can't use their Remote Extension suite.


Correct


OpenVSX. Again this is just the same as RHEL repos behind license login.


Red Hat provides support for their packages. If you're not paying for support, you don't get access to the repos. That makes sense to me. What does Microsoft gain by creating a walled garden? They don't provide support. All that they provide is hosting. The Eclipse Foundation provides hosting for free for OpenVSX, which is an amazing service to the community of people using VSCode forks that aren't allowed to access the VSCode Marketplace. Microsoft should either relax the ToS on the Marketplace or acknowledge OpenVSX as the one and only marketplace for extensions.


Codeium has extensions for all the major IDEs, including IDEA.

https://plugins.jetbrains.com/plugin/20540-codeium-ai-autoco...


These emails were released as part of the antitrust lawsuit against Google currently being pursued by the FTC. It seems to me that contrary to the FTC's claims about how Google's monopoly power leads it to stop innovating, exactly the opposite is true. If Google had stopped innovating it's clear that Bing eventually would have caught up in terms of quality. As these emails make clear though, Google kept its lead by continuing to invest in cutting-edge AI research.

Indeed, if anything it's Microsoft who should be scrutinized.


Both are acting with monopolistic power in numerous areas and both should be scrutinized. Just because one argument against monopoly power is that incumbents can sometimes rest on their laurels and fail to innovate doesn't make this the sole reason to pursue antitrust action.

Not on you though, this kind of reasoning failure is so common I think it deserves a fancy "cognitive bias" name. Off the cuff maybe something like "Single Rebuttal Fallacy"?


> These emails were released as part of the antitrust lawsuit

I never understood this: Why would hot shot, high powered people risk putting such things on email? Things that could backfire when released into public domain like from a lawsuit or a leak. They know this well, and still keep doing it. Why not setup an in person meeting, or just pickup the phone and talk ? Why email?

I almost feel they actually want it to happen but could never point my finger to how this could cover their asses, or an exit strategy?


Ultimately, if the people at the company have to do a job, they have to communicate somewhat, and putting it on an email or another recorded medium is the best way to prevent endless meetings and have the perspective documented in a single place.

Everything has a legal liability, it doesn't mean that it's worth it to move everything to an undocumented medium.


why were these emails released as part of the FTC's lawsuit against google.

How did the FTC get microsoft's email over an antitrust lawsuit against google


These emails might have been offered by Google to show the opposite right, they just are part of the trial.


How would Google have access to Microsoft's emails?


I think this airplane is not the best way to experience full totality—especially if it's your first time seeing a total solar eclipse. There still should be plenty of opportunities to see the eclipse from the ground in April!


The author mentions at one point that he was unable to solve a problem because he didn't memorize the formula for the Euler totient function in order to count the number of numbers relatively prime to 9999.

...but its actually an interesting (and not super difficult) exercise in its own right to figure this out even if you don't know the formula. Encourage you all to give it a shot.

SPOILERS: 9999 = 3^2 * 11 * 101, so first subtract out the multiples of 3 (3333 of them), the multiples of 11 (909 of them), the multiples of 101 (99 of them). Note that we've now double-subtracted multiples of 33 (303 of them), multiples of 303 (33 of them), multiples of 1111 (9 of them) so add these back. Finally subtract 1 to not count 9999 itself.

phi(9999) = 9999 - 3333 - 909 - 101 + 33 + 303 + 9 - 1 = 6000

I guess my point is that the purpose of these problems is not to separate out people who know specific tricks from people who don't—its to separate out people who can reason their way through difficult mathematical problems and people who can't.


The difference for you is that you're doing it as a fun exercise. With contest math, you're drilling these formulas and tricks so you can reproduce them quickly on a timed exam. If you know both of the facts listed in the essay then you can knock off this question in a minute or two.

Trying to come up with everything from scratch could take a lot longer and be very frustrating when you've got other problems waiting for you to solve.


> Trying to come up with everything from scratch could take a lot longer and be very frustrating

If that is frustrating instead of thrilling to you then you probably shouldn't do these competitions.


I’m trying to put myself into the shoes of the blog’s author. I just finished my math degree. It wasn’t at an elite school. I enjoy math for recreation though I’ve never been a contender in contest math.

I think I can understand why the author complains about drilling for math contests. I think it’s the same reason high level chess players complain about opening book. Drilling is not fun and it reduces the creative element of solving the actual problems.

The fact that he wasn’t doing it for fun but for college admissions plays a lot into it as well. That kinda pressure can really kill the fun.


Except for the intense speed pressure to create artificial rankings for competition, "drilling" is "practice for understanding".

OP says in the article that he didn't understand what he was doing, but was just trying to imitate people who did to try to appear brilliant. And he claimed that most students in club were like this, due to parenting and teaching that was focused on resume building for ignorant Admissions Offices, instead of focused on learning.


"drilling" is "practice for understanding"

It really isn’t. The essay makes this point as well. The acquaintance of his who was most interested in math dropped out of the contest team. He went on to become a mathematician because he was actually interested in understanding higher math and how things fit together.

Drilling is just practicing computational tricks to be able to execute them by heart. This would be true regardless of the amount of time pressure or lack thereof. I’ve taken many math courses that had exams in this style and I always hated them. I much prefer trying to figure out a proof I’d never seen before. For any real work, those computations would be done by a computer anyway.


You need to memorize things in order to be effective You need to be able to match patterns to things you've seen before. You need to develop an intuition.

LeBron James much prefers to slam dunks, but he still has to practice the game ad condition.


The problem (and linked solution is here)

https://artofproblemsolving.com/wiki/index.php/2022_AIME_I_P...

The totient formula isn't the hard part of the problem.

The test has a very short time limit (for the difficulty of the problems), and has many gruelingly complicated problems,so if you dont have the formulas down cold, you'll burn out during the contest.

Of course, if you don't care about silly speed-mathing contests, you can enjot the problems at a leisurely pace.


Agree, and to add a bit more context:

* The contest has 15 problems and a time limit of 3 hours, giving only an average of 12 minutes per problem.

* The average score is 4.83/15, from the official statistics [1].

* The statistics noted that only 5.17% of test takers answered this problem correctly, making it the second-hardest problem out of the 15 on the exam.

[1] "AIME 02/08/2022" https://amc-reg.maa.org/reports/generalreports.aspx


What's especially fascinating is that the core of the problem is a generalization of the totient computation, so understanding the inclusion-exclusion construction of totient is very helpful to the problem, while simply memorizing the formula is a misdirection.

OP missing this point shows that he really was doing this for all the wrong reasons. He should have done FIRST or science bowl instead.


Completely agree, I also liked the problem and thought it was conceptual as far as these things go.

Asking for N mod 1000 was another cute twist that was meant to get you thinking about the divisibility properties of the totient function - "hmm, so (p-1) always divides \phi(n) for all prime factors p of n, how convenient..."


The mod 1000 is actually a consequence of the test format: all answers are integers in [0, 999], you fill in 3 digit-bubbles.


That's a cute coincidence that 1000 divides (11-1)(101-1), but I doubt it was intentional. The test writers didn't have much choice unless they chose a different base than 5+5.


More direct calculation:

For every distinct prime factor p (so, of 9 is a factor, use 3 not 9), only (p-1)/p of the natural numbers ≤ n are relatively prime to it. Pime. This overcpunts nothing, since prime factors are relatively prime to each other. (Proving this requires some analysis of remaimders / modular arithmetic. But working an example shows the pattern).

This gives a formula, phi(n = Sum (p_i ^ e^i)) = n • Product ((p_i -1)/p))

This also shows how OP (and maybe the coach) missed the point of math team.


Cryptographers often tell software engineers that they shouldn't roll their own crypto. I think lawyers would tell software engineers that they shouldn't roll their own license. If you really intend or want other people to be able to use the software you wrote, for the love of god please pick a sane, well-known license so that people can use your software with full knowledge of the legal implications.


I don't want to be overly rude, but this is nonsense. The reason to learn calculus is that it's incredibly useful in several domains and never learning it prevents you from become a skilled practitioner in those domains which in turn reduces your future earning potential.

Basically: https://www.smbc-comics.com/comic/why-i-couldn39t-be-a-math-...


That is a good comic. But I am not sure "reduces your future earning potential" is really right. If you are going for future earning potential, then there are other things. Probably along the lines of - learn to code - find a way to migrate to the US - live in SF/NYC - learn algorithms and data structures - learn leet code - emotional control/resilience for putting up with those kinda jobs ... etc

I reckon there are plenty of PhDs earning less than $100k around the world, who know calculus and matrix algebra like their ABCs.


Everything can be incredibly useful in several domains, but we don't teach everything. Instead, people learn what they need when they start working in that domain.

The point of the article is that calculus is not taught because it might be incredibly useful for a small percentage of students, but because it measures their ability to pick up a hard subject and ace it.


If you make it harder for startups to achieve profitable exits does that incentivize or disincentivize innovation?


In my person opinion, in the grand scheme of things, the impact on innovation in this particular case is negligible because only under exceptional circumstance, anti trust regulators intervene to block a merger of a startup.

If it becomes more of a common occurrence, I agree it could stifle innovation.


Actually "working overnight" is "probably carcinogenic" which is a higher evidentiary standard than Aspartame which is "possibly carcinogenic".

Basically these WHO cancer guidelines are entirely useless in terms of public health and have probably done as much harm as Prop 65 has in California.


Guidelines | FAQ | Lists | API | Security | Legal | Apply to YC | Contact

Search: