lostdog

46 ms·

601.
by lostdog 6y ago
Often all the context you load into your head is other people's code. Sometimes you don't have control over how complex the codebase becomes. Debugging also requires a lot of context. I didn't downvote the comment (yet), bu
602.
by lostdog 6y ago
Scifi authors have definitely been writing about smart computers for a while now. Maybe some of the things that are newly specific to ML can be fodder for writing: - When building an ML model, you think much less about the structure of the
603.
by lostdog 6y ago
The bombs could be ejected at high speed along with the outer hatch door, kind of like a magazine, and then a secondary inner hatch closes to cover the hole. ...but I'm also not an aircraft designer.
604.
by lostdog 6y ago
Where are you even going to take your video calls from? Everywhere I've worked constantly runs out of conference rooms. Are you going to take your video calls from your desk, and disrupt the work of 30 people around you every time you
605.
by lostdog 6y ago
Let them serve a 16 year term on the supreme court, and then return them to a lifetime appointment on a circuit court.
606.
by lostdog 6y ago
Steer the conversation to your strengths and accomplishments. Sure, refusing on-call is a reason not to raise your salary, but what are the reasons that you are valuable to them?
607.
by lostdog 6y ago
I work in an area where high variance is very expected and unavoidable. Here's what we do: In your PR, you link to the tool showing the performance diff of your PR. The tool shows the absolute and relative differences of performance fr
608.
by lostdog 6y ago
The high variance is another problem. Good software has low variance in performance. Especially if you're sampling in the tens of thousands. The high variance does give you two tactical problems. First, how do you keep performance from
609.
by lostdog 6y ago
Maybe you'll get lucky and they'll ban you from using their products!
610.
by lostdog 6y ago
Dear Esteemed Colleague at Atlassian, I also use Confluence and JIRA regularly, and can confirm that they are the slowest most terrible software that I use on a regular basis. Every single page load and transition is slow and terrible. Aski
611.
by lostdog 6y ago
I worked in Python for a while, and once you have a production bug because the indentation shifted around on one line when you moved some code around, you start to get really cautious and nervous every time you cut-and-paste. I much prefer
612.
by lostdog 6y ago
Which do you think is more likely: A) There was a bunch of money allocated to climate change. We then took the money away and used it to bring an asteroid rock sample back to Earth. B) There was no money allocated to climate change, because
613.
by lostdog 6y ago
Anecdotally, I definitely came across more startups being located in the East Bay during my last job search. The founders already lived there, so they located the company there too.
614.
by lostdog 6y ago
I don't mind working for less on something I'm passionate about. But, if we do get lucky and the startup goes to the moon, I expect to do well! And that's why I can't work at startups anymore. It's fundamentally unf
615.
by lostdog 6y ago
> If there was a time to have options at a growing, revenue generating startup it’s today. AirBnB, DoorDash, etc. have been giving RSU's, not options, for years. RSU's are still worthless without an exit, but stock options have
616.
by lostdog 6y ago
Companies will shovel money to their lawyers to set up an ownership structure that benefits the investors and founders, but don't bother to develop a workaround to the AMT exercise trap, and that tells you everything you need to know a
617.
by lostdog 6y ago
Think about this: If the number of faculty positions isn't growing, how many PhD's should a faculty member graduate in total? Probably around 1-2 in their entire career, otherwise there is a glut of PhD's with no faculty. But
618.
by lostdog 6y ago
Many companies aren't ready to succeed at ML until they have failed at least once. Leadership doesn't understand that building a good ML models takes more time and effort. It takes a failure of an ML project (and sometimes multipl
619.
by lostdog 6y ago
Give them a term limit, and then send them back to their life appointment on the district courts (with a higher salary if you want to sweeten the deal a little).
620.
by lostdog 6y ago
But the payout to consumers should never be in coupons or discounts. Always cash. (Or conversely, the payout to the lawyers should be the same form as the payout to consumers. Good luck spending $100 million Wells Fargo Free Checking Accoun
621.
by lostdog 6y ago
It seems like the key to surviving is to grab the "I care about this" knob and turn it all the way down to as close to zero as possible. When I look back at times I was burnt out, I wish I had cared even less than I did. Every par
622.
by lostdog 6y ago
That bill also bans dropboxes and makes it harder to get a mail-in ballot. It's just a normal voter suppression bill.
623.
by lostdog 6y ago
Not if there's only one post-hole digger. And if they have a standup meeting beforehand to give status updates and "identify blockers," it's going to take even longer.
624.
by lostdog 6y ago
Thank you! I will take a look at that.
625.
by lostdog 6y ago
Glad you asked! I don't usually do GUI-related work, but each time I do I typically want to do something complex like plotting charts or drawing diagrams. Each time, I look for a tutorial, but I never find something helpful. The result
626.
by lostdog 6y ago
If only grants were big enough (and permissive enough) to hire professional SWEs. It's as if we sent a bunch of theoretical physics PhD's out to Geneva, gave them garden trowels, and let them break ground on a particle accelerator
627.
by lostdog 6y ago
The circle drawing and spreadsheet tasks are very interesting, because most GUI frameworks and tutorials just focus on laying out standard widgets and make it difficult to create custom components. Gtk and QT help you when your interface is
628.
by lostdog 6y ago
The problem is what happens if a state that signed the compact decided to break it. Because the federal government considers the compact illegal, it won't enforce it, so any state can break it at will.
629.
by lostdog 6y ago
Given the number of comments and downvoted comments, it almost certainly set off an automated flamewar detector.
630.
by lostdog 6y ago
Caplan is very right wing ("the homeless still smile!"). The talk was still okish, but the points of view come from one side and could use some balance. Blaming poverty on "impulsive sex" and leaving blameless the polici

More ›