TJ’s excellent interpretation of my release notes:
Perhaps second on my list of “Favorite Release Notes To Read” is Instapaper. What is impressive is not the ongoing development and improvement of an application that is already widely considered “essential” by many people, but also what gets removed from each version.
“X was a bit of a hack to deal with Y. I’ve dealt with Y so X is now gone.”
“Q was a bad idea so I’ve replaced it with R.”
“Previously there was an option to do G or H, but G was buggy and so H is the only option from this point forward.”
And so on…
It takes either a lot of arrogance or self-confidence (which of course is a very fine line to begin with) or to do this. On more than one occasion I have found myself thinking, “BUT I USE THAT!” and soon afterwards thinking “Yeah, but I don’t really miss it.”
Exactly. What Instapaper does, in both the website and the iPhone app, is a small, solid foundation under a massive collection of hacks and just-barely-working features to do things that you’re not supposed to be able to do. That’s why they need to be hacks.
Parsing text out of a web page? Easy. Trying to figure out which parts of the input are not important with enough confidence to remove them? Pile of hacks.
Showing HTML content in an iPhone/iPad app? Easy. Making it behave like a book and tilt and paginate and adjust fonts and save your position and work offline and not look like hell? Pile of hacks.
Any time a pile of hacks is involved, it’s going to need to be updated and improved frequently. Sometimes, that means removing some of the hacks. But there’s a bigger point, even for simple features.
Making a product better often requires removing features.
I have to give credit to Brent Simmons for helping me get over my fear of feature removal, by his example and expressed philosophies.
Dealing with the negative feedback is tough. Every feature removal, even if minor, is greeted with an initial barrage of emails from people whose lives I have just completely ruined by this change to my free website or my $5 iPhone application. I still get an email about once a week from someone who wants to burn down my hose1. It’s especially tough with web and iPhone apps, for which there’s no good way, or no way at all, for the offended customers to just keep using the old version.
But the result, once the fire has died down, is a much better product for the majority of customers.
If I could never remove features, I’d never add any.