AugustoCAS

To add - I agree that payments is great question for systems designs.

Merchants need to receive the money and being able to transact with it. Imagine the merchant receiving the payment is something like Amazon or Ebay (sorry, I don't know what similar large online retailers are in India).

With the above, the problem becomes harder. Imagine receiving 2-3k TPS just on one account during a black Friday or similar day.

Now your system has to perform fine for accounts that do 30 transactions an hour for a retail customer and 3k a second for a merchant.

I'm waiting!

Thank you!

Worth mentioning that setting this via effortLevel in .claude/settings.json does not work. https://github.com/anthropics/claude-code/issues/35904

This is common across all corporations. My go-to example is Unilever or Nestle pushing products that are 100% unhealthy.

In Asia, it's not uncommon to see healthy drinks for children that are sugar+artificial flavouring with huge marketing campaigns targetting the parents . The corporation makes millions and then advertises how they donated $10k to an obesity charity.

Are 'project rules' a thing?

.claude/rules/.md Project rules

~/.claude/rules/.md User rules

or is it just a way to organise files to be imported from other prompts?

Dora released a report last year: https://dora.dev/research/2025/dora-report/

The gains are ~17% increase in individual effectiveness, but a ~9% of extra instability.

In my experience using AI assisted coding for a bit longer than 2 years, the benefit is close to what Dora reported (maybe a bit higher around 25%). Nothing close to an average of 2x, 5x, 10x. There's a 10x in some very specific tasks, but also a negative factor in others as seemingly trivial, but high impact bugs get to production that would have normally be caught very early in development on in code reviews.

Obviously depends what one does. Using AI to build a UI to share cat pictures has a different risk appetite than building a payments backend.

This was easy because it's a Chinese company.

The largest companies in this space that do similar this (oxylabs, brighdata,etc) have similar tactics but are based in a different location.

This is not an issue for me due to my workflow.

I have a script for each of my projects that I run when I open a new terminal window (Alacritty). The scripts set up tmux with 3-8 terminals, each terminal launches a components, utility or just sits in a folder from which I later run commands.

Having said that, I use only a few zsh plugins, and have a theme configured to not run commands that add extra latency.

A chunk of the internet is down for me. So far Perplexity, AWS (VPN) and vercel.

Seems AWS is limping: https://news.ycombinator.com/item?id=45640772

Something that I find amusing in the Java community is that a good number of senior developers, with anything from 5-20 years of experience, who do 'tdd' have never heard of the concept of test doubles and religiously think that a class must be tested in complete isolation mocking everything else.

The saddest one I saw was a team trying to do functional programming (with Spring). The tech lead was a bit flummoxed when I asked why mocks are not used in functional languages and continued to think that 'mocking functions' is the correct way to do TDD.

A quick google shows this for FF (taken from a thread in StackOverflow):

In Firefox you can completely disable beforeunload events by setting dom.disable_beforeunload to true in about:config. Extensions may be needed for other browsers.

A word of caution: I'm not 100% sure, but I wonder if some web collaboration tools might use this to ensure data has been synced with a server.

[posted this in another thread, but maybe the author can clarify this]

I wonder how this works when one runs test in parallel (something I always enable in any project). By this I mean configuring JUnit to run as many tests as cores are available to speed up the run of the whole test suite.

I took a peek at the code and I have the impression it doesn't work that well as it hooks into when a thread is started. Also, I'm not sure if this works with fibers.

I wonder how this works when one runs test in parallel (something I always enable in any project). By this I mean configuring JUnit to run as many tests as cores are available to speed up the run of the whole test suite.

I took a peek at the code and I have the impression it doesn't work that well as it hooks into when a thread is started. Also, I'm not sure if this works with fibers.

A side comment, I have found that configuring a few live templates in IntelliJ helps me to write a lot of the repetitive code just a handful of keystrokes regardless of the language.

Structural refactoring is another amazing feature that is worth knowing.

Would love to know if you found it's a real key or a fake one.

I find a bit odd that MS didn't have a couple of people in the board of directors, who I assume accepted Sam Altman resignation (and signed the severance package).

Edit: I just read he was fired, but the point remains.

I had the same problem (there's a github issue about this: https://github.com/canonical/microk8s/issues/2186).

I swapped to k3s and the usage was half of what microk8s used.

I have the strong impression that most people confound cloud to be public only. Private clouds are fine and deliver a lot of the flexibility at a big % discount of a public cloud.

The big gain from clouds is the flexible infrastructure, especially in the microservices world we are now. In the past, one needed to procure, provision, etc a new server to run a service (times X per environment). With a cloud, regardless if it's public or private, provisioning a VM (or container) to run a new service is a few clicks away.

I only skimmed through the article, but if I understood correctly the scammers made $23m, now have to repay $3.3m (which probably will never happen) and some of their assets were taken. They were given 4 and 6 year prison sentences (which might be cut short for good conduct I assume).

So in 3 years these blokes come out, with probably ~5m-10m stashed away? Not a bad ROI :(.

Thanks for sharing this! I'm just getting my head around category theory (in Haskell).

I would appreciate if anyone knows of a small project out there that uses category theory to learn how to design and architect a solution (even better if it's in Haskell!)

Just saw this on LinkedIn posted by the author of the paper. He claims that with this, RSA 2048 can easily be broken.

There's not much information about this yet, but if true it will accelerate the end of RSA.

Post to the linked in post: https://www.linkedin.com/posts/edgerck_today-we-could-announ...

https://archive.ph/G11qB

Sharing this as I experienced a lot of friction over the years with startups/scaleups that did the transition to (heavy) management too early and as a consequence there was a lot of wasted effort as inflexible plans had to be re-planed into new inflexible plans. All these organisations had the idea that a manager cannot manage more than ~6 people, so in a company with 100 people there would be 2 (sometimes 3) layers of managers.

Link for those who are curious: https://github.com/IAmStoxe/oracle-free-tier-wirehole .

Thanks for sharing that repo. I have a similar setup, but with each service in its own compose file as I added them over time. I would definitely use WireHole if I didn't have everything running already.

It's a good question!

Coaching is about helping people find their own answers. One needs to know about a topic when giving specific advice / guidance.

Having said that, for me building technology has 2 challenges: people and technology. Initially in my career I focused on the technical aspects of it (clean code, good design & architecture, etc) and later found the people challenge is a lot harder than the tech. I learn by researching topics so I read quite a lot about teams, management and lean. Then I pieced together what good and bad managers did and how that related to what I had read.

From my experience most managers (and product managers, engineers, etc) perform their work in the same way they learned from other people while doing their job and rarely do any research or go to the source to learn how to be good at what they do. They might be sent to a 2-5 days course on how to manage, but that's it. The knowledge they have is often from 2nd+ hand, and very diluted. If you fancy doing an experiment, grab a couple of managers and ask them what mix of managerial theory or practice they follow. Ask for names of authors and thought leaders. Ask for any book they have read in the last 6 months about good management. There's a big chance you won't get anything or something very fluffy. What I found is that one doesn't need to know a lot in order to make an impact, and going to the sources helps a lot.

About a nun being a sex therapist, I do know a few Buddhist monks and nuns who joke about how much relationship advice they are asked when they never had a romantic relationship or when they weren't very successful at them before being ordained.

I would love the Linux Foundation to add a clause on the licensing of the kernel saying "This software cannot be used inside the EU and applies retroactively" and see the surprised pikachu face of the MEPs who voted on this.

I can only dream.

How much would cost to replace Linux with a proprietary, commercial OS on all devices inside the EU? 10s or 100s of billions?

A variation of "what are you going to do?" or "how would you solve X?" helps the IC to come up with a plan of how to address the feature/issue/whatnot. This creates space for them to take the initiative and after a while they grow and get used to do it, but it can take 1-2 years.

For some people who've been told what to do for most of their careers it's important to not leave them in the deep end on their own. It takes a mix of coaching and guiding them. The aim is that at each step they are faced with something that is both challenging and achievable.

Peopleware is a gem of a book, yet most managers and leaders in the tech world haven't heard of it (or at least never read it).

A recent example from my last gig: the way of working was to shuffle people relentlessly between teams to address tight deadlines.

Just to emphasise how bad this was, team names were "Feature Team X" (where X is a number), because management wanted teams to shift very quickly. This was in a company which was building a product, but treated people as in a body shop. What the managers didn't realise is that encouraging that detachment between teams and what they were building, also made teams not really care about what they were building. The message was clear: teams don't own any part of the product.

Absolutely bonkers.

Being an awesome manager is not about delegating, it's a about

* Creating the right environment for teams to do their best (See book Leading Teams below).

* Being skillful at challenging pre-conceptions.

* Coaching people and teams to work in the most effective way.

* Communicate, communicate, communicate.

* Hiring people who are smarter than you.

* And, a difficult one, having honest 1-to-1 with people when they are not delivering to know what is going on in their lives and, in the worst case, to help them go (aka fire/made redundant).

My favourites resources are:

* Turn the Ship Around by David Marquet

* The Five Dysfunctions of a Team by Patrick Lencioni

* Leading Teams by Richard Hackman

* The Goal by Eliyahu Goldratt

I'm not a manager (I don't like managing people), but I've been working for ~13 years coaching engineer managers, tech leads and engineers. You have to let the baby go so the people you hired (which should be smarter than you!) can do their job. Obviously if you see something that is wrong, by all means raise it, but be careful that it's not one of your own pre-conceptions!

Thanks for the tip!

For a while I've been thinking of setting a VM that navigates the interwebs clicking on random links to generate tracking noise. I need to find out how to do that :D.

Google makes money, so they are happy to look away. The scams (e.g. Mr Beast fake video) are identical so it would be trivial to automatically kick them from the platform. The crypto bro's ads are also pretty much identical.

The crypto bro ads are hilarious and it looks like they all use the same platform. I wonder if all of them are part of a ponzi scheme in which someone sells a course to sell crypto courses.

Something I do, which is a bit silly, is to click the ad just to burn money from the scammers.