I’m going to give a HARD disagree with this type of hand waving for a bug in pow().
First, how was a change allowed in such a fundamental function without significant justification and validation. It’s exponentiation. That’s like saying you want to fix multiplication. Sure, but it better get higher scrutiny that a higher-level function.
Second, where is regression testing. It’s a trivial function to write tests for, which means MS never created real test plans for basic UCRT functions. This is where the outrage should placed.
Precisely. I have been told I have a knack for reading other people’s code. Turns out it was years of reverse engineering as a hobby in my youth that gave me the skills to sit down and understand someone else’s code very quickly. A hobby that gives useful real-world skills.
Exactly true. Was a teenage software hacker, reverse engineered many games mostly because I was a bored teenager with free time and many games I legitimately bought. Mostly it was to cheat a little, like tweaking player stats or number of lives in a save file.
All that reverse engineering experience gave me a particularly useful talent as a developer: an uncanny ability to sit down and grok other people’s code almost like magic.
Solid experience with reverse engineering really is a good skill to have.
Depends on when the people did this and how complex the mechanism were at the time. I do remember just learning about JMP in assembler and that you could bypass a lot of stuff by changing a true to false or vice versa. Not exactly rocket science. (Not saying this to downplay anything, but what seems magical to many non-techies was actually very easy to do if you knew the trick, and it already worked for like 30% of the games - I'm not talking about the real stuff later on, not even sure how complicated NoCD was.)
On the other hand, probably 15-20 years after I had last used a crack for a game I had to implement some basic protection like this into a piece of software (more like: make sure the customer will not run this demo version forever) and it was fun to discuss in the team what we all remembered from back then and which measures were adequate to implement.
Have recently started seeing “Xitter” as a reference to the site. Possibly because it can be pronounced like a crude slang for a restroom, or just to make sure everyone realizes “Twitter” isn’t going away.
This sounds interesting!
The main reason I moved to perl instead of sed or AWK is for better control. Shell scripts don’t differentiate between variables and data (anyone else still having nightmares from trying to escape quotes and meta chars in a shell?).
Perl gives you all the simple features from sed or AWK and adds useful (maintainable) foreach iterators, arrays, hashes, etc.
Recommend using Strict mode if you are new to Perl. It gives good guardrails against silly mistakes like not declaring or misspelling a variables, or accessing strings or numbers that aren’t the correct datatype.
Thank you for the tip re: strict! Escaping - that's definitely a sore point. I feel like I get tripped up by embedded double quotes way too often - does perl have a good way to handle them?
Apple tried. Recall the huge IP licensing battle from a year or two ago where Apple challenged Qualcomm on base price for royalties and what was considered FRAND given some of the patents were incorporated into the communication protocol standards.
I have a fairly sophisticated user registration system that can't handle people without family names. We had to jump through some hoops to be able to enroll people who just have given names.
I use strips from lightdims.com also available from Amazon and other places. Comes in a sheet of precut circles and squares in different sizes so you can stick it over just the LED. Seems to have better sticking power than painters tape. They also have sheets of white / translucent if you want to dim the brightness but still have some color pass through.
Seconded. My daily walk to Starbucks is my time to meditate on my coding issue of the day, free from distractions of emails, office visitors, and meetings. Many coders I’ve met say a change of scenery can help you solve a problem you were stuck on.
Also a chance to say hi to the local crows who appear to recognize me nowadays and don’t fly away when I walk by :)
Managing that risk is what the customer CFOs were supposed to do. Your average retiree or semi wealthy person knows the FDIC limits and manages their portfolio around them.
This is not a new risk for businesses in any industry and financial services exist to help manage that risk for companies with cash on hand.
I would instead ask why the SVB customers didn’t manage this risk. Others on HN said that banking with SVB was a requirement placed by many VC companies, which does tie the customer’s hands.
Are you really expected to spread out all your day-to-day business cash around tons of different accounts just to stay under the FDIC limit? How is that meant to work, logistically?
Let's say you're a company with like 50 well-paid devs, they make 200 grand each on average, so monthly payroll for them is 50 x 200,000 / 12 = ~830,000, so you'd need accounts with four different banks to stay under the FDIC limit just to manage payroll for engineering. What about everything else?
