3 ms·
A screencast is not representative of actual coding. They already know what they're typing, and a few typos seems to dramatically slow them down. But, in real
by SwellJoe 18y ago
A screencast is not representative of actual coding. They already know what they're typing, and a few typos seems to dramatically slow them down. But, in real programming, you spend 80-90% of your time looking up library functions, flipping through your own code, reading comments, A/B testing various algorithm ideas, fixing the rendering in IE (actually this is probably 90% of your time...the other 10% is divided amongst all of the other activities), etc.
If you're spending more than 10% of your development time typing, you're probably coding things in the most naive way, and producing a lot more code than you should be. But even so, in six months, you'll probably be spending 90% of your time reading your code, because it's so verbose.
- stevecooperorg 18y ago* But, in real programming, you spend 80-90% of your time [...]* Well, yeah. But the question was about text editors -- how do you optimise that 10% of your time when you're actually cutting code? My point was, ultra-terse command syntax like vi's don't amount to much if what you're really typing is; arr(<back>[3<back>2];<back> to get the string arr[2] into your source code. Also, typing matters more in dynamic languages or uncompiled source like html, where spelling mistakes can go undetected more easily. The real benefits, as you note, are in things like automatic intellisense-style function- and class-lookups, easy integration with source control and documentation and refactoring/lint tools, and ease of navigating the call graph. The big win, obviously, is live IE and firefox previewing ;) http://www.sublimetextwiki.com/cgi-bin/moin.cgi/BrowsersPlugin http://www.sublimetextwiki.com/cgi-bin/moin.cgi/BrowsersPlug...