I didn't have any structure to the "hour" rule beyond: Have something to show for your efforts in an hour, and then move on to something else. Stuff was either

I didn't have any structure to the "hour" rule beyond: Have something to show for your efforts in an hour, and then move on to something else. Stuff was either "Never done this before" or "Done a version of this countless times before" - and either way, in the past, I've gotten caught up in trying to be perfect, and then not getting anything done. The hour limit took perfection out of the equation for old and new tasks. So, for example, who knows what is the perfect regex for grabbing websites and their associated comment pages from HN? Not me! But I whipped one up, and v1.0 was enough to make a graphical version of the front page of HN. Spin up a Linux Azure VM? No experience with it - but I could probably spend a day reading about other people doing it. Or, I could give myself an hour (or maybe more if Azure takes longer than an hour to create a vm(!) ) and then see what I have after that. I think the most valuable tool I used was my brain :^) By that I mean if you are great at coding in a particular language, and you are trying to "get stuff done" then use that language to help you learn the new stuff. For me, shell scripts work perfectly to glue multiple languages and utilities together for projects like this. The last part I'll mention is that I tried my best to keep the solution simple. So there's no elaborate backend here, no super-cool caching of anything, nada. Yes, I can do all of that stuff if I have to, but I can also spend days or weeks perfecting (messing around with) all of that and still have little more than what is here. Sixty minutes is plenty of time to get something done if it's new or old hat. That was a pleasant discovery. I can go into even more detail about the path that led to "Hey, this is actually becoming a thing" if you wish - I sense I have started to ramble.