3 ms·

Until 4PM, you've also been working on things that are value-adds on some level. Your choice not to make use of the small time windows (which aren't actually so

Until 4PM, you've also been working on things that are value-adds on some level. Your choice not to make use of the small time windows (which aren't actually so small) is your own choice, not a direct function of the office.

mcantor 15y ago
You're right! However, it's inefficient for a developer to always be doing that much non-development work. And it is possible to get real work done during the small windows, but only certain types of real work. Occasionally you can fit in a bunch of good bug-fixes. Other times you're able to organize things into discrete small tasks up-front, and you can get a few of those things done with each small window. But, it's impossible to really achieve flow in those circumstances. The lack of flow is really the gist of my argument.
- mkopinsky 15y ago
  With the schedule you described, it doesn't sound like you're purely a developer. Either a lot of this stuff needs to move from your plate to someone else's, or you need to accept that you do a lot of non-development work. Note that of the distractions you describe, a lot of them are things that you would have still had if you were working from home. The emails and the IMs are still there, although I guess they might be more ignoreable since you don't have the threat of the person showing up at your desk. :-)
Retric 15y ago
Depending on what you are working on 15 minutes may be enough time to make some progress, but anything with a moderate level of complexity is going to take longer than that before gain enough context to start.
- mkopinsky 15y ago
  If you take notes, you can gain context from 9:00-9:30, and rev up again quickly from 10:00-10:03. The time doesn't need to be wasted time. (Not that this is necessarily easy to do, or perfect, but I think we can accomplish more than we think in this manner.)
  - Retric 15y ago
    Notes like all other forms of documentation is suspect. If it's a complex problem that you don't already have a solution to you really do need to sit there and think "Why did I think this when I wrote it." otherwise it's easy to create flip flopping bugs. AKA fix bug 357, fix bug 372, hey wait bug 385 looks like 357 ahh that's right I need to set the flag to X, hmm wait bug 392 looks a lot like bug 372 etc. PS: And when you have 15 minutes to work it's next to impossible to figure out how to fix 392 without also creating 357 yet again.
    - mkopinsky 15y ago
      What I meant by "notes" was just scribbling down your thoughts as you go along - things like "$property modified between file.php:500 and :600? WTF". Enough data to jog your memory when you get back to your desk and then throw away once you solve the bug and document the fix in a real fashion.
detst 15y ago
http://www.paulgraham.com/makersschedule.html http://www.paulgraham.com/makersschedule.html

mcantor 15y ago

You're right! However, it's inefficient for a developer to always be doing that much non-development work. And it is possible to get real work done during the small windows, but only certain types of real work. Occasionally you can fit in a bunch of good bug-fixes. Other times you're able to organize things into discrete small tasks up-front, and you can get a few of those things done with each small window. But, it's impossible to really achieve flow in those circumstances. The lack of flow is really the gist of my argument.

mkopinsky 15y ago
With the schedule you described, it doesn't sound like you're purely a developer. Either a lot of this stuff needs to move from your plate to someone else's, or you need to accept that you do a lot of non-development work. Note that of the distractions you describe, a lot of them are things that you would have still had if you were working from home. The emails and the IMs are still there, although I guess they might be more ignoreable since you don't have the threat of the person showing up at your desk. :-)

Retric 15y ago

Depending on what you are working on 15 minutes may be enough time to make some progress, but anything with a moderate level of complexity is going to take longer than that before gain enough context to start.

mkopinsky 15y ago
If you take notes, you can gain context from 9:00-9:30, and rev up again quickly from 10:00-10:03. The time doesn't need to be wasted time. (Not that this is necessarily easy to do, or perfect, but I think we can accomplish more than we think in this manner.)
- Retric 15y ago
  Notes like all other forms of documentation is suspect. If it's a complex problem that you don't already have a solution to you really do need to sit there and think "Why did I think this when I wrote it." otherwise it's easy to create flip flopping bugs. AKA fix bug 357, fix bug 372, hey wait bug 385 looks like 357 ahh that's right I need to set the flag to X, hmm wait bug 392 looks a lot like bug 372 etc. PS: And when you have 15 minutes to work it's next to impossible to figure out how to fix 392 without also creating 357 yet again.
  - mkopinsky 15y ago
    What I meant by "notes" was just scribbling down your thoughts as you go along - things like "$property modified between file.php:500 and :600? WTF". Enough data to jog your memory when you get back to your desk and then throw away once you solve the bug and document the fix in a real fashion.

detst 15y ago

http://www.paulgraham.com/makersschedule.html http://www.paulgraham.com/makersschedule.html