Every time I see Sam Bankman-Fried walking around free of any charges and giving interviews to major media outlets after defrauding people out of billions of dollars, I'm thinking of this guy.

In a month from now, it will be 10 years since Aaron Swartz died. He was sentenced to 35 years in prison and $1 million in fines for wire fraud, because of downloading "a large number" of academic journals. With the prospect of spending most of his life in a cell, he took his own life.

Back in the day, New York Times didn't ask him whether he's sleeping OK and if he's playing video games to clear his mind. News outlets didn't try to justify his actions as just a "foolish mistake". He didn't go around talk shows fake-crying and claiming he didn't understand what he was doing.

</rant>

People who target printers at #Pwn2Own, do you get to keep them if you successfully exploit them?

Good luck to all the participants.

Every time I see someone online overlaying candle graphs from different time periods and predict bull runs, I also wonder how many times they parroted the phrase "past performance is not indicative of future results" and felt smug about it.

As they say, technical analysis is just astrology for men in their 30s.

Back in 2014 I was doing web operations at a large SaaS provider. Besides the core application, customers also had the option to have custom integrations, which were built for them by the Professional Services team. These integrations were often used within our customers' intranets, as opposed to the core application which was used 24/7 by our customers' customers.

One Monday, out of the blue, the database servers were on their knees from the load, and so were the memcached servers. High network utilization. The environment was multi-tenant, so the degradation was not localized only to a single customer.

We quickly pinpointed the issue to a specific customer. After we confirmed that the infrastructure was healthy, we started the long process of trying to figure out what was the root cause of this extreme resource utilization.

The volume of the external traffic was high as always. There were no recent updates in the core application. The PS developer swore he didn't push any code changes. I confirmed.

On Tuesday we eventually spotted a table in the MySQL database that had a handful of rows added to it on Friday afternoon.

I contacted the developer once again, and he confirmed that he updated the table on Friday. But he didn't understand how adding this minuscule amount of data could create such extreme load on our infrastructure.

The exact details escape me, but it turned out that due to this minor database update, each URL invocation resulted in a memcached query, a consistent cache miss, a query to MySQLd and an attempt to cache the fetched data once again.

This integration would receive millions of hits per day from 8am to 4pm. The change was done on a Friday after 5pm, when the usage was dying down, so any minor usage spikes flew under the radar.

Anyway, here's my face when reading takes about how easy it is to run infrastructure at scale from people who can barely operate their own phone.

Yugen Blakrok's "Anima Mysterium". Moody and groovy.

"Monatomic Mushroom" is probably my favourite off the album:

https://youtu.be/M1yyH3XRV3w

In April 2013, the official Twitter account of Associated Press was hacked and tweeted that the White House was bombed. For the next few minutes the markets moved wildly and billions were wiped in market cap. FBI and SEC opened an investigation.

These days it takes about 8$ to pull off such a stunt. As we saw today, it wasn't long until someone would put 1 + 1 together.

The exodus of users and advertisers was bad, but wait until most publicly traded companies realize how massively exposed they are due to Twitter's mere existence in its current form. This is not a problem they can solve by just leaving the platform. 🍿