As the famous, slightly modified from original, saying goes:

"There are only two hard things in computer science. Cache invalidation, naming things and off-by-one errors."

I always start with a codename. And for the longest time I've used the surnames of foreign scientists:

http://en.wikipedia.org/wiki/List_of_German_scientists

We use Pokemon. The reason? Nobody gets too attached to the name Jigglypuff.

That's a great idea I might steal. I'll go through the names in dex order, as I happen to know at least the first 151 in order...

Also I'm pretty sure that they make new Pokémon faster than anyone starts projects...

Don't bet on it.

http://ajf.me/imagedump/screenshots/ihavewaytoomanyprojects....

You can switch to Digimon if you run out of Pokémon.

I guess you haven't seen this... http://evilbrainjono.net/pages/startup-or-pokemon.py

There's always Finnish lakes: http://en.wikipedia.org/wiki/List_of_lakes_in_Finland Might be difficult to pronounce for some, though.

I've used this list multiple times for personal project codenames: http://en.wikipedia.org/wiki/List_of_rivers_of_the_Americas

Funny. I use mythological figures. Right now it's "metatron", and "delphi"...

These are so bad for a software project that they might even work for a codename. And this coming from a Finn.

Even better: http://en.wikipedia.org/wiki/List_of_lakes_in_Minnesota

I'm torn between Operation "Big Spunk" and "Big Sucker".

I once came up with a great name for a product I was hacking on with my roommate, so we used that name internally. By the time we'd been coding for a few weeks, though, what the product actually was had evolved and the name was no longer relevant (despite being scattered throughout the codebase, the github repo, the directory structure, etc).

Ever since then I've been a codename fan. I go for liquor names (since I work at home across from the booze shelf)- I'm currently building Frangelico.

Norse gods here:

http://en.wikipedia.org/wiki/Norse_gods

Having to type in Gerðr or Þrúðr would get old pretty fast. ;)

Gerthr. Thruthr.

now I know where tumblr and flickr got their names.

I generally use either the first thing that pops into my head (sometimes it's just obvious) or the song I'm listening to at the time, either name or a snippet of lyrics. SteelBreeze, ClubThing, Tain, Ronin, Arienette, etc. Works wonders.

I pick hostnames from this: http://en.wikipedia.org/wiki/Sumerian_King_List

Characters from Bananaman. Not my choice, and I can't wait for the list to exhaust... though I like the idea of calling the devs VM test server 'zookeeper'.

http://en.wikipedia.org/wiki/Bananaman

The Apache foundation beat you to it!

http://zookeeper.apache.org/

It's actually a clever name for what they're doing

Just be careful which scientists you use and in what context. Picking the wrong one could get you sued: http://rationalwiki.org/wiki/Carl_Sagan#Apple_Computer

When I was with HP calculation division in the early 2000s, the project names came from Star Wars / Star Trek characters. Why? Because with we were working SciFi (SCIentific FInancial) calculators.

I use coffee shops or espresso company names, often local to Seattle. The list is quite limitless. Zoka, Vivace, Rococo, Velton, Olympia, Lighthouse, Ladro, and so on it goes.

To take the head-scratching out of the start of a project, and to give apps personality, we use surnames of film directors.

If I have to work on another project named after Johann Christian Polycarp Erxleben, I'll quit this damn company!

You'll never run out of microbrew beers. I mean for naming. I do run out of the real thing while coding.

I used to use Japanese foods, so I've had directories named sushi, sashimi, miso, karasumi and so on.

LittleBigPlanet, our PS3 game, was (is) called 'ps3test1'. the sequel, LBP2, is... also called ps3test1. that project really was our first attempt to bring up a devkit, probably with a rotating cube.

the project, and compiled output, on every platform, is called 'pc.elf' (or .vcproj or .exe or whatever) SIGH

there's an inverse correlation between awesome-ness of directory name and chance-of-shipping, in my experience.

> there's an inverse correlation between awesome-ness of directory name and chance-of-shipping, in my experience.

Sackboy's law? It's absolutely true.

Thanks for sharing that about LBP: it's great to hear that my nominative inertia is shared by such high company.

I know what you're thinking. You're thinking: "How can CPAN help me with this problem?"

Answer: http://search.cpan.org/search?query=Acme%3A%3AMetaSyntactic&...

Without knowing my way around perl, how would I use this to generate a name from one of those categories?

Perl modules (almost) always come with a _Synopsis_ section in the documentation, which describes how to use them. Here's a direct link to the one for `Acme::MetaSyntactic::soviet`:

http://search.cpan.org/~jforget/Acme-MetaSyntactic-soviet-0....

I've gotten asked a lot where the name Miranda IM came from and I honestly have no idea. I was creating a new solution in Visual Studio and I needed a name. I remember it took under 10 minutes to come up with, I was browsing a bunch of name lists and I came across 'Miranda' and I thought that sure is a strange / unusual name, I'll use that. Turns out it's a fairly common name. I still think it's a good name for the product.

I remember when I was working on ACDSee, the original author said, "If I knew it was going to become popular, I would have picked a better name." The company originally made catalog software, and a Co-op student made ACDSee as a side project. It's sustained the company for almost 20 years.

There are only two hard problems in Computer Science: Cache invalidation, and naming things.

And off-by-one errors.

Off-by-one errors only affect good programmers. Otherwise you just add enough of such errors that all the under- and over-summing cancels out.

7 years ago, when I was a new programmer, I did use a technique like that to fix some JavaScript. I feel brave confessing that.

Just I case FYI: the off by one errors is part of the joke of the statement.

The reference: http://martinfowler.com/bliki/TwoHardThings.html

I usually start projects on github, and allow the random name chooser to make that decision for me..

    "Great repository names are short and memorable. Need inspiration? How about drunken-nemesis."

Good translate is your friend.

Pick a language you don't know, and have at it:

"picture site" => "irudi_gune" (basque)

I find that Basque, Portuguese, and Welsh produce very good names.

As someone from Brazil, I wonder what kind of words would sound weird to an English speaker like that. I would have thought that most words would be recognizeable, given that English also has its fair share of Latin words.

Portugese is far enough from Latin that many of the English cognates are not immediately recognizable.

"Sangue bom," for example, is not visually translatable to English.

I find just calling it app_number and moving on helps. I used to go in circles with names, until I realized how much time and sanity I was wasting. Spending more time on developing the product helps me understand it's core value better and, consequently, potential names to communicate that value effectively to new users.

I just give projects a plain, descriptive name, not worrying too much at first about using the exact, correct words. I have projects with names like “Boggle word list”, “cost matrix solver”, “def_init-initializer-type DRY enabler”, “Ghost Assistant”, and “recursive spiraling dots animation”. I don’t publish my projects until I’ve worked on them for a while, so I don’t mind if they’re not perfect at first.

I don't get to name many projects. But when I'm naming procedures I want the name to be self-documenting without being too long. Frequently I get exasperated and name it blah or sdfbdhfs!

foo? bar? qux? metasyntacticvariable? potato?

name it something. mv works if you need to change it.

it also should be somewhat unique so that sed -ie can do its job

You could steal ideas from Debian, Ubuntu, Fedora, and several other Distro for giving codename to their release.

For example, in Debian, they take a character name in Toy Story for their codename. In Ubuntu, they take an animal name in alphabetical order.

I have designated an in-prog folder and call it 'dev'. Everything in dev is strictly modifiable.

Working in a folder called dev feels wrong to me. It's like having a folder called 'bin' that is a trash bin.