Live data from Hacker News

Ask HN: What were the things you did that made the biggest impact at your work?

news.ycombinator.com

101–110 of 216 posts

Re: Ask HN: What were the things you did that made the biggest impact at your work?

#101

Switching my company to remote. It was a gradual process but it worked and now we’re fully remote and everybody seems happy with no exceptions. I can’t take full credit but we also switched to doing 6 week project cycles[0] and pitches inspired by Basecamp. [0] https://3.basecamp-help.com/article/35-the-six-week-cycle

Sort of did this at my previous job by being senior and in a good position and communicating that I’d be working from home first 1, then 2, then 3 days a week. Other team members with long commutes followed (it made sense to avoid in-person meetings the days I wasn’t there, so then other devs would also wfh..) and the company invested in better tech, software and processes for remote working. Talked with a former colleague who is now 4 days a week remote - when I started he spent 3h/day in the car.

Re: Ask HN: What were the things you did that made the biggest impact at your work?

#104
I worked in an old non-tech LargeCorp IT department for a year and the most impactful decision was simply giving a damn about things. Everything was so abandoned that it was hard to break the inertia. I started trying to care about the warnings thrown by the compiler, and eventually managed to upgrade the internal apps to be able to work on IE8 instead of 6, and from them moving to IE11 was simpler, and afterwards everything was working on Chrome and modern browsers, which in turn allowed some coworkers to push to moving up from Win XP, and eventually by the time I left the inertia had turned into forward moving improvements. Felt really proud about that and left in good terms.

Re: Ask HN: What were the things you did that made the biggest impact at your work?

#105

I wrote a little tool that used perf profiles from our production fleet to generate a custom linker script that reordered our main server program’s binary to be significantly more cache friendly. The heuristic I came up with for reordering was one of the few (maybe the only) genuine “eureka” moments I’ve had in my career. And the performance win was extremely nice :-)

Interesting, got a blog that goes into more detail?

We wrote this paper about it for CGO: https://research.fb.com/publications/optimizing-function-pla....

But the paper actually describes a significantly more sophisticated heuristic. My initial implementation simply used the number of perf samples divided by the size of the function, which helps make sure you’re getting the most out of your I-TLB. It worked shockingly well for its simplicity.

Re: Ask HN: What were the things you did that made the biggest impact at your work?

#106

I wrote a little tool that used perf profiles from our production fleet to generate a custom linker script that reordered our main server program’s binary to be significantly more cache friendly. The heuristic I came up with for reordering was one of the few (maybe the only) genuine “eureka” moments I’ve had in my career. And the performance win was extremely nice :-)

For anyone curious about using something like this technique, Facebook has a similar tool. https://github.com/facebookincubator/BOLT

Yes! In fact that’s from the same team I was working on at the time (HHVM). My tool predated BOLT by a few years, I think. Obviously BOLT is way more sophisticated and does an even better job.

Re: Ask HN: What were the things you did that made the biggest impact at your work?

#109

I played a really small role when looking at some caching data for a system another team run (their logs are ingested and processed by us) and noticed that a lot of the time the content requested is only requested once - this is sometimes referred to as a "one hit wonder" Once I told them this and showed them some graphs etc they did an experiment to set it to only cache on the 2nd hit, and this reduced the write rat…

One hit wonder - thanks I’ll introduce this concept to my boss and see if it helps us: we accidentally lost some caching due to a bug but there was no measured performance penalty on our production performance tests. Hence we didn’t notice until someone looked at the code. Could be a one hit wonder! It’s a nice concept to think about in any caching code.

Re: Ask HN: What were the things you did that made the biggest impact at your work?

#110
I changed a boring linear telemarketing script into a flow chart (with lots of clever jokes in it) that instructed the marketeer to thank people and terminate the conversation slightly prematurely. (in stead of trying to force people though the linear script) The mood in the call center changed from wanting to die to perpetual laughter. It seemed to increase productivity slightly but people called back years later when they needed insurance.
Post reply on HN