Hacker News Reader

That'd be true if we weren't talking about an iPhone app that users will install without ever actually going to the web site behind it.

  1st User Experience:
  - open App Store
  - search for "Color"
  - install

  2nd - nth User Experience:
  - click the "Color" button on your iPhone

Notice how we never visit the actual site. TheColorApp.com would work just fine for them.

Only true if they planned to exit on the app alone. Maybe the plan is ambitious enough to warrant the expense.

In unrelated news: a friend of mine casually referred to her boss as a 'green' on Saturday... apparently it is the result of popular personality test that everyone in the company has to take. The result (one or more 'colors') is part of their email signature for internal correspondence.

Totally Agree. I think photo sharing is a classic "Thin Wedge" that Dixon and Wilson have written about before. Think about how powerful Color could be:

Business Networking - Go to a conference and create an ad hoc social network. You are basically automating what Fridge and Lanyrd are doing now.

Neighborhood Social Networking - Need to organize a carpool or soccer practice schedule? Moms can set up a network with their phones while sitting in the stands watching their kids play.

Same for schooling, retailing, and dozens of other opportunities where you might want to let people communicate then sell them stuff. If nothing else it might be so it is easier for advertisers to remember and more easily access the self serve ad mechanisms the founders described.

I could see all of those eventually having a traditional "Web" page though it might be assembled and interacted with very differently than a traditional soc net.

I can remember a couple of nights in my youth when I might have enjoyed putting together a location-specific ad hoc social network... right around closing time.

The neighborhood social networking example is not as straight forward as it sounds. The barrier to use isn't just the user experience, but the metaphor behind the application. Most regular people have a hard time learning new metaphors to interact with technology.

Do you really think that an entire group of middle-aged moms would all download an app on their smartphones to co-ordinate carpools? No! They don't want to deal with technology. Most likely they will just talk person to person (what a novel idea) or use their existing email list from the team.

Don't mean to get too personal, but this kind of thinking is one of the disadvantages from living in Silicon Valley. You become disconnected from how the majority of people live and interact with technology

Exactly. Normal people won't use technology unless it does something they can't, but need to do, or makes something significantly easier. They don't just use tech for its own sake like I do.

Green, red, blue, and with rainbows in the middle?

I think I know the test.

But if you ever want to be more than just an app, lets say a BRAND, then IT IS important.

>Most mainstream web users can remember 7 +-2 web services

I seriously doubt this. AFAIK most people have 7+-2 "registers" of short term memory, which has strong implications for stuff like the maximal # of digits in a phone number - but it doesn't affect the number of phone numbers you can remember, or names of people, or ball players etc.

As for "real company name": Google, Yahoo, eBay, Skype, Paypal, Wikipedia, MySpace, Hotmail, Amazon, Apple, Dell, Starbucks... Company names can become meaningful without starting this way.

I actually had a hard time trying to find the app in the Android Market, because there's only one search term and it's really generic. Colr would probably have been easier.

Even if it was Colr, Android Market could find a way to make it impossible to discover via search.

This seems like a somewhat trollish comment, but serious question: is it easier to find in the iOS App Store? How is that possible?

Similar problem (tons of "color" apps), but Apple managed to rank them highly in results when I looked.

I assume they key off of selection criteria from the users (ie. most users installed X after searching for Y, so rank X higher for Y), but it could also be that their company name and app name are both "Color".

This is exactly true, and the worst mistake, imo. I have no idea how to find it in the Android Market, and I suspect many others don't as well.

> 1. Easy to say/spell....having to explain your site is Colr.com or TheColorApp.com instead of just "Color" will impede consumers.

> 2. It makes you seem like a real company with the resource to own a short, properly spelled name...

This is not a very hard problem to solve.

    kragen@inexorable:~/devel$ cat shortnames.py
    #!/usr/bin/python
    # Pick unused domain names. Kragen Javier Sitaker, 2011.
    import socket, random, itertools

    def lookup(domain):
        try:
            return socket.getaddrinfo('www.'+domain, 80)
        except socket.gaierror:
            return None

    # We try an unlikely-to-exist domain in case our ISP engages in a
    # man-in-the-middle attack on DNS traffic.
    failing_dns_response = lookup('jaoiwgjaioawjio'+'aioj.org')

    words = [line.strip() for line in file('/usr/share/dict/words')]

    short_words = [word for word in words if 0 < len(word) <= 6]
    print "Considering %d short words." % len(short_words)

    for ii in range(20):
        while True:
            sld = random.choice(short_words) + random.choice(short_words)

            if len(sld) <= 2:
                continue

            domain = sld + '.' + random.choice(['com', 'org', 'net'])
            if lookup(domain) != failing_dns_response:
                continue

            print domain
            break
    kragen@inexorable:~/devel$ time ./shortnames.py
    Considering 22221 short words.
    LippiSeurat.org
    crownseagle.net
    JoleneMable.org
    anorakbarf.org
    baitdepend.org
    opticbib.org
    elopevictim.org
    maniacAbner.org
    glandhanker.org
    nthprissy.com
    dumpsSandra.org
    measlyhafts.com
    cassiaLaud.net
    avowstarots.com
    Elvianoun.org
    duchyfeisty.net
    dualgales.net
    hikingpinup.com
    dirgesBarnes.net
    tillchars.net

    real	0m15.749s
    user	0m0.140s
    sys	0m0.028s

How about dualgales.net, opticbib.org, napstiff.org, ticketgorge.org, damnharem.net, bargehutch.org, ravagepunk.net, jigglemutant.net, EliRx.org, jobunsafe.org, vendspanda.com, lathedkid.com, poppedcopula.net, triadsvia.com, civicsdiary.com, greyvalet.com, queenfigs.org, pagansultan.net, tacofoci.com, odderzoos.net, fleatablet.org, and so on? The whole blob of them all together looks like spam, because spammers register domains using a very similar approach, but any one of them is a perfectly respectable, easy-to-say and easy-to-spell, relatively short domain. This program can generate about 1.5 billion different domain names, modulo possible weaknesses in random.choice.

It's so easy to spell that they had to buy the variant spelling of the domain as well.