beckingz

31 ms·

361.
by beckingz 5y ago
Implying that tumblr is over. The golden age continues.
362.
by beckingz 5y ago
If one person you meet is an asshole, that is unfortunate. If every person you meet is an asshole, you may be the asshole.
363.
by beckingz 5y ago
I just left a relationship for this reason, and it is so deeply painful. I wish you the best. I think it is the right decision and that making it sooner rather than later is best, but it is not easy.
364.
by beckingz 5y ago
You get a data lake per team/service.
365.
by beckingz 5y ago
FastAPI probably does everything that you're looking for?
366.
by beckingz 5y ago
spicyusername has it right. There's a concept called COST - Configuration that Outperforms a Single Thread, which basically asks the question: "how many cores/servers does a system need to outperform a single thread/core
367.
by beckingz 5y ago
Microservices do allow scaling the development of an application on the human side. This is true! A kickass CICD setup and good developer tools probably should be done first for the typical company, otherwise the majority of the added engin
368.
by beckingz 5y ago
So staff engineer is Senior + Product Manager effectively?
369.
by beckingz 5y ago
Except that state on kubernetes seems really hard to do well, and data is inherently state. Also the complexity cost significantly adds to the 'configuration that outperforms a single thread'. 99% of the time you'll be better
370.
by beckingz 5y ago
Knowing how to run software makes you better at writing it!
371.
by beckingz 5y ago
I really like that this stores all data in the data warehouse, unlike Monte Carlo.
372.
by beckingz 5y ago
MLOps is radically overhyped. I've seen two companies in the last year start twisting themselves into circles to worry about 'productionizing' a data science project. The intent is right to focus on end use case and business
373.
by beckingz 5y ago
I really like obsidian for writing text files. Technically they're markdown, but I almost never look at them as markdown files.
374.
by beckingz 5y ago
This. Sometimes holding information until things are ready is the kindest thing to do for anxious people.
375.
by beckingz 5y ago
Agreed that getting more specific with tasks is helpful. Unfortunately, it's hard to leave the engineering approach for larger projects to the developers when their productivity (and apparent satisfaction) increases significantly with
376.
by beckingz 5y ago
And what should a PM do when their engineers are not effective at shipping?
377.
by beckingz 5y ago
The system works! In reality, we will likely see some aspects of software development become less well compensated, while other aspects will continue to be rare and highly valued.
378.
by beckingz 5y ago
Metabase is fantastic for this use case. Create a read replica, deploy metabase, connect it, start building visualizations. Or use the X-ray auto analysis feature to generate a bunch of mostly useful visualizations with a single click.
379.
by beckingz 5y ago
Apparently this was the move for a while on amazon: buy upvotes for the people who are listed higher than you so they get removed.
380.
by beckingz 5y ago
'BizOps' is usually a mix of internal consulting, operations, and data analysis. There is no standard definition and I think we should keep it that way. With the number of 'Ops' that people are trying to do soon companie
381.
by beckingz 5y ago
All the examples I can think of didn't actually start a new industry. They did get in early on a new industry, but they didn't start it even if they ended up defining it.
382.
by beckingz 5y ago
They just work if you dont' care about money.
383.
by beckingz 5y ago
The mean age at founding a 1 in 1000 fastest growing organization is 45. Which makes sense, because industry connections for talent and sales are incredibly valuable.
384.
by beckingz 5y ago
You can buy them off amazon. Even if they are not allowed?
385.
by beckingz 5y ago
A lot of them still have an ICE powertrain unfortunately
386.
by beckingz 5y ago
And this is why Product Engineers are a thing now.
387.
by beckingz 5y ago
I attended a talk a few years ago given by one of the blockchain VPs at IBM. It turned out that under the hood they were just using conventional distributed database technology. All marketing in order to get multiple companies to work toget
388.
by beckingz 5y ago
I assume that they'll pay one senior person this and otherwise it's just marketing.
389.
by beckingz 5y ago
All abstractions are ultimately leaky, but you still need to have them to build large things.
390.
by beckingz 5y ago
Frank - the autoresponder gpt-3 bot is fantastic.

More ›