This is not true. Ask Facebook, who have rewritten things multiple times explicitly because this is not true, but someone assumed it was
>maintain ability and first to market are usually more important
Maintainability and first to market are not trade offs for performance in most cases, no matter how much you want to propagate this ridiculous propaganda.
But the question is, would Facebook still be around if they didn't "just ship this turd lol"? I don't really have any insight in Facebook engineering over the years and it's a "what if" type of question that's essentially unanswerable, but the answer to that being "no" is very plausible.
And Facebook really does have unique(-ish) scalability problems, and I bet rewrites would have been inevitable even with the best possible engineering, because who can write an application to deal with a billion users (2012, currently 3 billion) right from the start? When Facebook launched in 2004 this was pretty much unheard of, and even in today it remains pretty rare (much less 2012).
This type of thinking is what has turned everyone's "desktop" app into an Electron piece of shit. It turns software into a race to the bottom where as long as it's just good enough for users to not drop it, companies say "ok let's do it". It's not good advice to give, imo.
> Don't forget, you, the dev is 99% of the time the most expensive resource.
That boils down to not valuing time of the people who use your software. Despicable attitude imho.
Developer time is wasted once. User time is wasted for every user, each time the software is used, for as long as it remains in use.
All these can be large: billions of users, many uses each day, software in use for decades.
If you're the only user, no point to spend time on optimizing anything, unless you perceive it as too slow.
If software has billions of users, then almost any effort in optimizing the tiniest bottleneck in it, is worth the effort. A good example of "the needs of the many (users) outweigh the needs of the few (developers)".
A lot of software sits somewhere in between though (not that many users & performs at acceptable speed).
This is not true. Ask Facebook, who have rewritten things multiple times explicitly because this is not true, but someone assumed it was
>maintain ability and first to market are usually more important
Maintainability and first to market are not trade offs for performance in most cases, no matter how much you want to propagate this ridiculous propaganda.