Hacker News Board Hacker News, browsed like a classic message board
Live data from Hacker News

"Software is getting slower more rapidly than hardware becomes faster."

techslang.com

Post reply on HN
Print view View original

151–160 of 193 posts

Re: "Software is getting slower more rapidly than hardware becomes faster."

#151

Post by Sohcahtoa82 »

Tommah wrote: post #140
sandoze wrote: post #90

Why am I doing all these interviews that grill me on big O(n) if they’re not even using it?!

It is important to understand asymptotic complexity, but the places where it's useful in the real world are very different from textbook examples or Leetcode problems. You'll rarely have a triple loop in your code that makes you think, hmm, maybe I can reduce this to a double loop. That can happen, but these are more frequent problems I've seen: * Application code is issuing too many queries to the database. The clas…

> * Missing database indexes. A lot has been written about this, and it can cause a slowdown of N or N^2 times in some cases. Adding an index in the right place can change a linear scan into a much faster B-tree search.

I can give a personal piece of anecdata on this.

I was using a static application security testing tool on a huge repo (over 2M lines of code). It reported a couple thousand issues, nearly all of which were issues on code style rather than actual security issues.

Generating a report would take literally 2 DAYS and would grind the database server so hard that any other use of the SAST suite had a noticeable impact. I decided to run a profiler and see what queries were hitting it so hard. Turns out there was a SELECT that was being used a lot, but the columns being searched weren't indexed. I don't remember the exact query, but I'm guessing it was searching all the code in the repo that was stored in the database repeatedly.

I manually ran a CREATE INDEX. It took about a day for it to complete, and it added a couple gigabytes to the database. But now, those reports went from 2 days to about 20 minutes.

> If you’re writing Java code that has to make calls ten levels deep to get anything done, those levels of indirection aren’t free.

I'm of the opinion that any time you're performing type introspection, reflection, and ".invoke()", it's a code smell, but I suppose middleware in Java is impossible without them due to its draconian type system.

Re: "Software is getting slower more rapidly than hardware becomes faster."

#152

Post by nikau »

butz wrote: post #66
myself248 wrote: post #2

I've been saying for a while now, that developers and their managers should use a low-spec machine, let's say the 5th-percentile of whatever their users are using in the field, at least one day a week. This might force some reprioritizing of performance and efficiency goals. As long as devs spend their days with bugattis while writing software for corollas, this will continue to get worse.

In some big business companies this is solved by pre-installed "security software" that usually takes up about half of the RAM, sometimes even more, and is using a measurable chunk of CPU all the time.

The problem is all the bottom tier machines also have all this crap installed

Re: "Software is getting slower more rapidly than hardware becomes faster."

#153

Post by 1vuio0pswjnm7 »

myself248 wrote: post #2

I've been saying for a while now, that developers and their managers should use a low-spec machine, let's say the 5th-percentile of whatever their users are using in the field, at least one day a week. This might force some reprioritizing of performance and efficiency goals. As long as devs spend their days with bugattis while writing software for corollas, this will continue to get worse.

As a user of "low-spec" computers, I try to avoid software written by teams of developers working under managers.

Most of the programs I use day-to-day are (a) decades old, from the distant past before so-called "tech" companies existed and/or (b) written primarily by single authors, often as a hobby or as a result of itch-scratching. Some of them are written by me.

When I occasionally compile and run the scripts and small programs I write on more powerful computers, I am amazed at the speed. Everything I write is compiled on "low-spec" computers. Today's hardware is amazing.

Re: "Software is getting slower more rapidly than hardware becomes faster."

#154

Post by MichaelZuo »

nebula8804 wrote: post #115
MichaelZuo wrote: post #98

Earlier quoted context omitted.

Maybe it is part of their secret strategy to force in-house developers to develop for the 5th percentile worst case. As the parent alluded to.

No its a mentality that has been in corporate for decades. The best way for IT to not get in trouble is to lock everything down and define strict policies that prevent any ambiguity and define peoples work requirements in strict categories. This used to be easier with the advent of timeshared terminals but the advent of desktop computing caught them off guard. Now they have done their best to seize back control so mu…

It was a joke/sarcasm.

Re: "Software is getting slower more rapidly than hardware becomes faster."

#155

Post by earthling8118 »

myself248 wrote: post #2

I've been saying for a while now, that developers and their managers should use a low-spec machine, let's say the 5th-percentile of whatever their users are using in the field, at least one day a week. This might force some reprioritizing of performance and efficiency goals. As long as devs spend their days with bugattis while writing software for corollas, this will continue to get worse.

Where are all of you people working where you get crazy machines? I've gotten a laptop at every job that didn't let me use my own computer. They're never fast at all even when they use supposedly higher end hardware. Not even a fraction of the speed of my regular workstation.

Re: "Software is getting slower more rapidly than hardware becomes faster."

#156

Post by dietr1ch »

When developing there's tension between multiple things, and the first that's let go is performance because velocity is pushed too hard, particularly by management since their goals are rarely performance (unless the project is truly performance-critical (still second to correctness unless it's NoSQL stuff).

Even if people say they value performance it's always below velocity since they assume that it can be simply fixed later, and only if it's a big issue. To make things worse, hardware keeps getting better anyway, even if it's at a slower pace.

Re: "Software is getting slower more rapidly than hardware becomes faster."

#157

Post by Gigachad »

troupo wrote: post #121
Gigachad wrote: post #105

Earlier quoted context omitted.

Businesses and programmers listen to money. When they don’t, they go out of business. Users primarily care about features first and speed second.

Users do care about speed. They complain about it all the time , you just have to listen to them. It's most of the "features" that users couldn't give a crap.

So why aren’t the users picking the software that’s faster instead of the ones with more features. Why does everyone use the “bloated” Gmail webui instead of some barebones desktop client that’s a few ms faster?

Re: "Software is getting slower more rapidly than hardware becomes faster."

#158

Post by up2isomorphism »

Because being fast is no longer the main reason you make money from software, the platform makes money.

I recently wrote a slowest piece of software I have ever written since I found a job and it is making 1M per day for the company since it owns the damn platform.

Re: "Software is getting slower more rapidly than hardware becomes faster."

#159

Post by illiac786 »

foofie wrote: post #138
rahkiin wrote: post #132

Earlier quoted context omitted.

I get that, just sad npm becomes 100x slower (We measured this), and build times and indexing times increase so much too.

You're telling them that your convenience is worth more to the company than plugging a security hole. The company can live with your loss of productivity because it is living with the loss of everyone's productivity. The company cannot live without security. On top of that, if you are the only one making noise about slow computers and complaining about stuff and things then you eventually end up being the problem. I…

No the statement is "it's worth investing in making the UX less crap to increase productivity". Either improve SW (of it's in house crap scripts, as often), switch to more performant AV, etc. etc.

It does not have to be crap actually, I have seen companies with usable machines.

O365 on the other hand, I don't see a solution, it's horrible.

Re: "Software is getting slower more rapidly than hardware becomes faster."

#160

Post by hulitu »

butz wrote: post #66
myself248 wrote: post #2

I've been saying for a while now, that developers and their managers should use a low-spec machine, let's say the 5th-percentile of whatever their users are using in the field, at least one day a week. This might force some reprioritizing of performance and efficiency goals. As long as devs spend their days with bugattis while writing software for corollas, this will continue to get worse.

In some big business companies this is solved by pre-installed "security software" that usually takes up about half of the RAM, sometimes even more, and is using a measurable chunk of CPU all the time.

And when the ransomware hits, nobody notices, because "computer slow" is the new normal.
Post reply on HN