** UnhandledExceptionEventHandler :: OFFICIAL LC / CHATTER THREAD ** - Page 1525 - Computer Technical Help

Two Plus Two Forums Other Topics Computer and Technical Help

** UnhandledExceptionEventHandler :: OFFICIAL LC / CHATTER THREAD **

Post Reply Subscribe

...

Page 1525 of 1603

First

525 1025 1325 1425 1475 1505 1515 1520 1521 1522 1523 1524 1525 1526 1527 1528 1529 1530 1535 1545 1575

Last

Page 1525 of 1603

First

525 1025 1325 1425 1475 1505 1515 1520 1521 1522 1523 1524 1525 1526 1527 1528 1529 1530 1535 1545 1575

Last

04-05-2019 , 12:24 AM

#38101

jjshabado

Carpal Tunnel

Join Date: Jul 2006 Posts: 22,732

I guess I also don’t see the ability to piss on your bosses desk as a great measure of career success. It’s a fragile form of success and security where you’re betting on the success of that company and product area.

I’d aim for the success of being highly valuable and where people really want to work with you. If your current company or product goes under (or you just want to move on) you’ll have opportunities up the yin yang.

Quote

$25m Guaranteed WPM on CoinPoker

Join the action now

Daily Rewards • Splash Pots • CoinRaces

04-05-2019 , 12:59 AM

#38102

suzzer99

Save the Cheerleader, Save the World

Join Date: Nov 2005 Posts: 101,163

You take things to literally. It's not about the ability to be an *******. It's about the job security of knowing you could be an ******* and get away with it - vs. the zero job security of most PMs I've known no matter how hard they worked. Maybe what I've seen isn't the norm.

Quote

04-05-2019 , 03:25 AM

#38103

jjshabado

Carpal Tunnel

Join Date: Jul 2006 Posts: 22,732

Quote:

Originally Posted by suzzer99

I understood that’s what you were saying. The ability to be an ******* is almost always about specific high value skills/knowledge/experience that you have that makes you hard to replace in a specific situation. But those are almost always tied to specific technologies or products a technologies that can go out of flavour quickly or products that can be killed off.

Job security across products, technologies, companies is way better. And in my experience not something that good PMs are lacking.

Quote

04-05-2019 , 03:29 AM

#38104

jjshabado

Carpal Tunnel

Join Date: Jul 2006 Posts: 22,732

To some degree this is also just a management vs non-management thing. If you’re responsible for the output of other people you can get thrown under the bus and will almost never have the ability to be an ******* to other people.

If you go the pure tech career choice you have much more control over your destiny. But it’s absolutely more limiting then management.

Quote

04-05-2019 , 09:18 AM

#38105

suzzer99

Save the Cheerleader, Save the World

Join Date: Nov 2005 Posts: 101,163

I'm guessing this is really a merge sort and the idea is to break the list down until you get to chunks of 2, then sort those chunks in parallel, then start combining back and sorting as you go - still in parallel?

So for the bottom few rows - you're basically looping over both chunks to sort them back together - but because both chunks are sorted you can be more efficient in merging it with the other chunk as you loop.

Do I have that about right?

Quote

04-05-2019 , 10:30 AM

#38106

kerowo

lolcat

Join Date: Nov 2005 Posts: 37,183

Seems to be missing a lot of steps, if you can sort groups of four why dick around with groups of two?

Quote

04-05-2019 , 10:57 AM

#38107

Chips Ahoy

help me help you

Join Date: Apr 2007 Posts: 28,869

The base case for merge sort is one item.
You do this because merge is touching each element once -- n -- but sorting four or more is n squared.

It's log n merge steps to divide in half so the whole thing is log n times n.

Quote

04-05-2019 , 11:04 AM

#38108

jmakin

banned

Join Date: Jan 2008 Posts: 27,798

Quote:

Originally Posted by jjshabado

Yea and the skills I could develop here, imho, are not useful. I am not good at C programming or writing C for complicated distributed, highly multithreaded systems. The only real experience I am getting is a comfort level using tools like vim/git/docker etc and a stronger understanding of unix based systems (even then specifically centOS). Idk if there’s really high demand for the things I’m good at here.

Quote

04-05-2019 , 12:18 PM

#38109

suzzer99

Save the Cheerleader, Save the World

Join Date: Nov 2005 Posts: 101,163

Quote:

Originally Posted by Chips Ahoy

So the point isn't to divide and conquer and run tasks in parallel but to get the n steps down?

How does this differ from the famed bubble sort I've heard happens at every interview? Is there a nice Marge-like graphic of bubble sort for easily distracted visual learners like me?

Quote

04-05-2019 , 12:55 PM

#38110

kerowo

lolcat

Join Date: Nov 2005 Posts: 37,183

Quote

04-05-2019 , 12:57 PM

#38111

Chips Ahoy

help me help you

Join Date: Apr 2007 Posts: 28,869

Quote:

Originally Posted by suzzer99

The point is divide and conquer to get the time complexity down.
Bubble sort is n^2. In bubble sort you put one more item in place each pass but you also get some things closer to the right spot. There are some youtubes of bubble sort in action you'll like.

Mergesort's n*log(n) is much better time than n^2. Merge sort requires allocating which makes it take up O(n) space, sometimes undesirable.

I think being able to quickly implement quicksort, heapsort, mergesort in your favorite language has some interview value.

Quote

04-05-2019 , 01:20 PM

#38112

RustyBrooks

Carpal \'Tunnel

Join Date: Feb 2006 Posts: 24,646

No one has ever asked me to implement a specific sort in an interview. Maybe once or twice I've been asked to write or at least sketch out a sorting algorithm. I always choose heapsort, because it's very easy to implement and understand. And if they'll let you, you can just show a few core principles of the implementation and then say "you just keep doing this until it's sorted"

Quote

04-05-2019 , 01:57 PM

#38113

jjshabado

Carpal Tunnel

Join Date: Jul 2006 Posts: 22,732

I remember being told at one point that efficient mergesort implementations don’t actually recurse to the base case of 1. Instead you pick a different sort algorithm for anything smaller than some N that might have a higher big o runtime but a much smaller constant coefficient which makes it much faster with small Ns.

Quote

04-05-2019 , 01:58 PM

#38114

jmakin

banned

Join Date: Jan 2008 Posts: 27,798

Has anyone in the last ten years ever had to implement their own sort?

Quote

04-05-2019 , 02:00 PM

#38115

suzzer99

Save the Cheerleader, Save the World

Join Date: Nov 2005 Posts: 101,163

Sure!

[40, 100, 1, 5, 25, 10].sort(function(a, b){return a - b});

Quote

04-05-2019 , 03:24 PM

#38116

jjshabado

Carpal Tunnel

Join Date: Jul 2006 Posts: 22,732

Most of the motivation for teaching sorting is that it’s a really good way of teaching algorithmic complexity. You have a trivially obvious thing (sorting) that actually has many different implementations. You can focus on the theory and not worry about the problem getting in the way.

It’s the same reason why people ask interview questions about it. It’s a trivially simple problem to explain and all of the effort can be on the important stuff.

Literally nobody is teaching this or interviewing on it thinking that the actual sort function implementation is itself useful.

Quote

04-05-2019 , 05:45 PM

#38117

RustyBrooks

Carpal \'Tunnel

Join Date: Feb 2006 Posts: 24,646

Quote:

Originally Posted by jmakin

Has anyone in the last ten years ever had to implement their own sort?

No, and I also haven't had to prove any mathematical theorems either, I guess all those math classes were just a waste of time.

ETA:although I don't think anyone should ask me to prove theorems in interviews either. But we've discussed exhaustively how dumb programming interviews are

Quote

04-05-2019 , 07:03 PM

#38118

Victor

Carpal \'Tunnel

Join Date: Jul 2003 Posts: 81,658

Quote:

Originally Posted by jmakin

So why do you insist that leaving is not an option?

Seems you are in a great spot. You have great salary for your experience level and job security since you fulfill a desperate need.

So go to work and learn as much as you can while shipping your resume everywhere. Now you can be super picky about where you work next.

Quote

04-05-2019 , 08:16 PM

#38119

suzzer99

Save the Cheerleader, Save the World

Join Date: Nov 2005 Posts: 101,163

My work is hiring a PM - $50-$55/hr. They've got one narrowed down who's supposedly really good.

Quote

04-05-2019 , 09:19 PM

#38120

kerowo

lolcat

Join Date: Nov 2005 Posts: 37,183

When I was looking I was getting some head hunting offers from companies contracting for Nike and that's the range they were in as well, probably into the low 60's, and that was for my very weak PM resume.

Quote

04-06-2019 , 05:25 AM

#38121

KatoKrazy

Pooh-Bah

Join Date: Jan 2009 Posts: 4,623

Quote:

Originally Posted by suzzer99

My work is hiring a PM - $50-$55/hr. They've got one narrowed down who's supposedly really good.

As a full time employee or a contractor?

If FTE with good government bennies then that's not bad, but if contractor hell no.

Quote

04-06-2019 , 11:46 AM

#38122

blackize5

old hand

Join Date: Sep 2012 Posts: 1,358

Quote:

Originally Posted by Victor

He says conflicting things about the salary. In one post it's 50th percentile nationwide while he lives in one of the highest CoL markets. In another it's great for his experience.

I suspect he's actually underpaid by 10%+

Quote

04-06-2019 , 12:17 PM

#38123

jmakin

banned

Join Date: Jan 2008 Posts: 27,798

I make $90k should i make 100? I dont think so

Probably could in a few months

Quote

04-06-2019 , 01:40 PM

#38124

KatoKrazy

Pooh-Bah

Join Date: Jan 2009 Posts: 4,623

Quote:

Originally Posted by jmakin

I make $90k should i make 100? I dont think so

Probably could in a few months

Remind me when you graduated and started working?

A little over 4 years ago for me and my base salary is about 50k more than that and also fairly decent bonus, RSU and ESPP money. I think we are both in the OC area right? So you can use that as a decent barometer.