The Only Truly Failed Project | Hacker News Reader

The Only Truly Failed Project

codinghorror.com

62 pointsfogus28 comments

How many people know that the lead developer for Bob 2.0 was also the co-founder of Valve and the development lead for Half-Life [Gabe Newell]

Worthwhile read just for that snippet of knowledge!

Half-Life was almost completely redesigned from scratch because they realized well into the project that the game they were making was not fun. You can read about it at http://www.gamasutra.com/features/19991210/birdwell_01.htm , I've posted it here before.

The rest is history.

From that article:

Avoid all one-shot technical elements. Anything that requires engineering work must be used in more than one spot in the game. Engineers are really slow. It takes them months to get anything done. If what they do is only used once, it’s a waste of a limited resource.

I'm not sure I 100% agree with this but it is something to keep in mind for many of us: if we're programming it, it should scale. Not scale to a hundred thousand users, that is easy. No, it should scale hilariously disproportionately to time invested.

Getting a new landing page on my site took me about an hour of programmer time. That is... well, OK, it performed well so it wasn't a waste of time, but it is something that I don't want to have to repeat for every landing page.

A little bit of Rails magic, though, and that one landing page became seven hundred. And additional ones can be created by my freelancers, whose time is very, very cheap relative to mine.

your doing a 700 split test? can't wait to see the blog post on that!!!

Valve practices like... anti-agile. They take something and they say, "No, it's not PERFECT, let's start over."

And they're heroes.

It's not all that different. The biggest distinction is they keep the bulk of their iteration private. They use themselves as the ultimate customer and iterate until they have a product they love.

See Also: Blizzard, Apple

Pixar's the same way. They halted production of Toy Story for the better part of a year (if I recall the time correctly) so that the writers could improve the story.

Valve is terrific. I'll love them forever not for Half-Life, which is great but flawed, but for Steam, which lets me buy games quickly and has incredible weekend deals. (Two weeks ago they sold 10 brilliant indie games for $30. Best game-related deal since the Orange Box, which incidentally was also Valve.)

Eight games, IIRC. I bought that pack too since I hadn't played Goo and Braid yet past the demo.

is that a bad thing?

The description he makes of how they cut the game into one level and had a team essentially scrum on it for months bis kinda agile.

I've seen GN do a talk about developing HL1 - one of the things he said was that he learned a long time ago to come up with your own design process. Because "otherwise you spend as long getting a team into someone else's groove as you do making the product"

(Valve IMO are a masterclass in how you can spend years developing a game and still maintain a huge number of fans: mostly because everything that leaves the door is a quality product and so much fun).

"We set up a small group of people to take every silly idea, every cool trick, everything interesting that existed in any kind of working state somewhere in the game and put them into a single prototype level. When the level started to get fun, they added more variations of the fun things. If an idea wasn’t fun, they cut it. When they needed a software feature, they simplified it until it was something that could be written in a few days. They all worked together on this one small level for a month while the rest of us basically did nothing. When they were done, we all played it. It was great."

Nice illustration of Gail's Law, from a comment [1] on HN today:

"A complex system that works is invariably found to have evolved from a simple system that worked. The inverse proposition also appears to be true: A complex system designed from scratch never works and cannot be made to work. You have to start over, beginning with a working simple system." http://en.wikipedia.org/wiki/Gall%27s_law

[1] http://news.ycombinator.com/item?id=775134

That's a great article. I see some of the techniques that made free software so successful; like only allowing people to suggest ideas if it's something they could reasonably implement. ("Patches welcome")

I remember MS Bob being enjoyable when I was a kid. I loved the characters, the rooms, and the feeling it gave me. I'm installing it in a virtual machine right now so I can relive it. I'm sure I'm going to be disappointed, things just aren't as magical when you grow up.

Okay, I just decided I was a dumb ass kid. MS Bob is lame!

Microsoft Bob is quickly becoming a pseudo Godwin's law for tech bloggers. There hasn't been anything new written about it in years. The only sort of interesting points were written in the article codinghorror quotes. (http://www.techflash.com/microsoft/Innovation_The_lessons_of...)

The final line is great: "The only truly failed project is the one where you didn't learn anything along the way."

If the project is original or interesting, you almost have to go out of your way not to learn anything.

I do love that last line. Overall, a good post from what I find is an often stale blogger.

I believe it's impossible to succeed without failing.

Not anymore.

OP overlooks the key to avoiding failure: release early and often. Customer feedback during development, not afterwards, is the rudder that keeps your project from going down the abyss. Unless you're an ocean liner. Like Microsoft.

Well, isn't it the things wrong in the release (e.g. failures) that drive the next release, that is to say learning from them?

Some projects don't have that possibility.

Examples?

(I've run many projects and have always been able to get some consumer feedback during development.)

Games.

Complex games take a non-trivial amount of work to get them to where they're even remotely playable.

"Release early, release often," would, for instance, likely be a poor philosophy for a team working on the computer vision systems used by armed autonomous military vehicles.

Could test early and test often be substituted successfully?

for all you know, we do... the U.S. has stopped measuring/publishing collateral damage for wars.

"Experience is what you get when you don't get what you want."

Deleted comment

Not my downvote but it may be because you're in the wrong room - this thread has nothing to do with libertarians. Perhaps the above is a perfect example of some libertarians' abject lack of failure to argue the merits of their ideas at all times and places, appropriate or otherwise.

Reply on news.ycombinator.com