Comments on: Memory Allocations and Automation: Generalism versus Optimization https://bloggeek.me/memory-allocations-automation-generalism-vs-optimization/ The leading authority on WebRTC Sun, 09 Feb 2020 08:01:47 +0000 hourly 1 By: Ori Albin https://bloggeek.me/memory-allocations-automation-generalism-vs-optimization/#comment-116123 Mon, 20 Feb 2012 08:31:06 +0000 http://bloggeek.me/?p=98#comment-116123 This seems to be a classic Pareto principle problem (or refers to the 80/20 rule) . Are you willing to design your product relaying on the 20% that may encounter problems in the future or the 80% that will supply you with a working system.

Although the answer isn’t clear, as the last 20% usually encounter 80% of your time.

I usually relay on working elements and reuse as much as possible than re-invent the wheel. Inventing things that work even for an improvement won’t be the best solution every time. Mostly you will encounter the sane problems that the previous team ran through. You’ll be wasting your time on fixing bugs that the “all ready” system has eliminated already.

You should of course design your code as fluent as possible that the memory allocation mechanism could be easily changed if these problems that you fear of pop-up.

The key to this issue for my opinion be thought through the design phase. The optimization for these problems must be in thought at the documentation level as therefore it will be a big mess to fix if needed.

]]>