The main piece of advice of the article is finish it, it being whatever piece of software you're working on.

I think "finishing" is generally bad advice, because the problem for many developers is that "finishing" implies perfection and makes completion seem insurmountable. The author discusses various possibilities for what "finishing" may mean to you, but I would simply tweak the wording a bit to be release it. Get it out to the people who you think will use it and see if (and how) they do. Then release again.

Agreed, this goes hand in hand with the whole "artists ship" thing pg wrote about. If you're too caught up in the perfection of something start slicing of limbs of it until it's releasable and go from there.

i like your "limbs" idea. of course programmers get frightened by 'finishing' in ryanwaggoner's context. we know what it's like to want perfect.

i'm only just learning how to slice and chop my projects into pieces... find what i hav to do as bare minimum to make the software 'useful' and put the big black period there.

work in very small iterations - 2 week-mini projects. set good but minimum goals. have a "later list" so you can stop coding that next great idea u think of.

..and after you 'finish' each mini project. hav a beer.

i loved this article.

I agree. But I didn't get the same meaning of "finish" form the article that it seem you have. I think his meaning is very close if not the same as "ship it".

Yeah, to be fair to the author, the explanation he gives is very close to "release it" but my point is more about which makes a better slogan.

My advice would be "finish it and make sure it work".

This article is a bit... crap.

Writing software, no matter what level, or how much, is very difficult.

...

but there are document after document written on how to write good software, so implementation details can eventually be solved.

???

Then, when I saw he spelled "alma mater" "alma mata", I skimmed through the rest of the article and closed the window. Sounds like he has a very newbieish perspective on writing software.

A "newbieish" perspective can sometimes help even the best professionals in whatever industry to become more aware of flaws or improvements that they may have become to accustomed with to notice.

It's hard because he is writing a framework. Since he's not solving a real problem, he doesn't know when it's done or not.

So true. I cant even count the number of projects that I've started, only to have them fall into the background and die, never reaching completion.