4 ms·
Due to context switching costs, I'd think the progress made working 30 minutes a day would be nowhere close to what you would achieve if you did it all at once.
by bostonpete 11y ago
Due to context switching costs, I'd think the progress made working 30 minutes a day would be nowhere close to what you would achieve if you did it all at once. For me, I think my output would be closer to 2 weeks worth of work than 4.5...
- toupeira 11y agoIf you don't do it in-between your other work you can avoid the cost of context switching. And consider that you spend a lot more time refining your ideas (whether consciously or not), so I bet you'll be a lot more focussed during those 30 minute sprints than on average during your work days.
- khalilravanna 11y agoI think this is where planning and leaving notes for yourself comes into play. If you can spend one hour long span planning out what you're going to work on for the next 6 days of 30 minute sprints, it should lessen the time loss due to context switching. That's what I found worked best during my year of writing code every day. I would sit down on a weekend day and plan out a bunch of bite sized tasks. That way I could look at the list, pick something that could be done in 30 minutes without much thinking and get right to the task of writing code. (That being said, I didn't restrict myself to 30 minute sprints during that year and there were certainly things I did that either could not be done in 30 minute sprints or would require quite a bit of extra planning to be able to be broken up into subtasks and done in 30 minute sprints.)
- adrusi 11y agoOTOH full work days involve fatigue, which you won't run into only working for 30 minutes. And there's a good amount of context switching that goes on at the workplace as well.