Coffeeshopped

How to Disconnect Completely

Chadwick Wood — Thu, 29 Dec 2011 16:50:36 +0000

On January 8th, I'm flying to Guatemala. I'll be there for about two months, studying Spanish and sightseeing. Over the past month, I've been going back and forth on a question: should I take my laptop? It's a big decision.

If I don't take my laptop, then the trip is basically a two-month vacation. I can use internet cafes to correspond with clients and take a thumb drive to do some minor work on websites. But without a Mac laptop, I can't do any updates to my iOS apps (e.g. Pouch and C2G). The thought of something going wrong with one of those apps (someone finds a serious bug, or car2go or 37signals makes a server-side change that breaks things) is a scary one.

On the other hand, if I take my laptop, I'm going to be worrying about it the whole time. Theft is common in Guatemala, and carrying around a device that costs the equivalent of several months' rent might not be a good idea. I would have to keep the laptop on my person when I'm doing things like hiking around Tikal or Semuc Champey. And then there's the mental aspect to it... constantly carrying a laptop on an exotic trip is about the best symbol of never being able to escape work that I can think of.

So, I think I'm leaving the laptop in Austin. I really care about being responsive to customers' and clients' problems and needs, but if anything comes up in the next couple of months, I'm going to have to respectfully ask people to wait. It's time for me to get away and disconnect for awhile. When I get back, you can be sure my first priority will be taking care of anything that happened while I was gone. Thanks for understanding.

-- Chadwick Wood

Tags:

Against Flat Rate Projects

Chadwick Wood — Wed, 02 Nov 2011 15:35:46 +0000

If you're a freelance web worker, then you may have run into this scenario before:

Prospective Client: I think I'd like to work with you! What's your hourly rate?
You: Great! My hourly rate is $X.
Prospective Client: Oh, that's a bit more than we can afford. Would you consider doing the project for a flat fee?
You: ...

First off, let me point out the obvious error in reasoning here: you can't say that an hourly rate is too expensive until you know how long the project is going to take. What if I can make your Groupon clone in 2 hours? The cost of the project could end up being lower than the cost of someone who charges a lower hourly rate but takes longer to get things done. So, I gotta call bullshit on that one.

But more importantly, I want to talk about the evils of working on a project for a fixed amount. In short, I think someone always gets screwed. It's the nature of the agreement.

Flat Rate == Stand-off

When a project is agreed upon to be completed for a fixed amount of money (and it supposedly has a fixed scope, yeah right), the worker and the client have set themselves against each other.

The worker is given an incentive to complete the project in as little time as possible. The faster you work, the higher your effective hourly rate is! So, screw quality, let's get this sucker done (not that you'll think this way, but the agreement promotes thinking this way).

And consequently, the worker also will end up wanting to fight tooth and nail to not let the client change anything whatsoever about the scope of the project. The thought process there goes something like,

"The client wants a little lightbox widget to view larger versions of photos? Well, that doesn't take very long, but I won't get paid any more for it, and it wasn't part of the initial agreement. What do I doooooo?"

At that point, as the worker you either stick to your guns and say no, or you say yes and effectively give away a little of your time for free.

On the other side, the client now knows exactly how much they're paying, and they will want to get as much as possible for that amount of money. When they change their minds about aspects of the project (as all clients do, hey it's natural), they'll be confronted with a worker that either says "no, no, no", or a worker that says "yes", begrudgingly, and resents them just a little more with each new request.

So by the end of the project, either the worker has given more time than they planned to, and effectively devalued their time, or the client paid more than they would have if the worker had been paid hourly. Or maybe, just maybe, it worked out just right. That's rare.

How I Think You Should Do It

Freelancers:

Do your work for an hourly rate.
Give your clients good estimates, with a low and a high number as a range for the cost.
Find parts of the project scope that are vague, and talk them out and make them more specific.
Tell the client that you won't exceed the estimated price so long as there isn't a significant change in scope to the project.
Communicate about cost and progress over time as the project goes on.
If the scope is creeping, talk to the client about it and give them an idea of how much more the upper-end of the estimate needs to be to accommodate these changes.

Clients:

Stop asking for a flat fee for your projects (I know, you won't). Yes, it's a little scary to not know exactly how much your project is going to cost. But we can give you a range! And we can give you a worst-case scenario.

I think it mainly comes down to trust. A flat fee for a project feels safer. But it's an illusion, and it makes for a bad relationship.

Full Disclosure: I'll still do a fixed-rate project, once in a blue moon. But I feel dirty and a little ashamed of myself every time I do, and I always tell myself "this is the last time..."

I know I'm taking a bit of a hard line on this. And if you disagree, I'd love to hear it in the comments.

Audiences:

Tags:

C2G 1.2: Now for iPad, Too

Chadwick Wood — Fri, 30 Sep 2011 18:38:57 +0000

C2G 1.2 is now out, and with it comes support for the iPad. I've also added a button to contact support (me) if you ever run into any problems. More improvements are on the way!

Tags:

Products:

Lessons from Pouch at 1000 Sales

Chadwick Wood — Mon, 26 Sep 2011 19:54:34 +0000

A couple of weeks ago, Pouch passed the 1000 sales mark! And so far, I think I still have a happy (and growing) customer base. I couldn't be happier about how the app is doing, and I figure that this milestone is a good occasion to tell you some things I've learned during this process.

It's Hard to Get Ratings

The vast majority of people who use Pouch have never rated it. Out of over 1000 sales worldwide, Pouch has been rated 73 times, with about 53 reviews: about 7% of customers bothered to rate it.

But, this percentage would have been even lower if I hadn't included a ratings request screen in the app; the 3rd time a person launches Pouch, a picture of me pops up with a request to rate the app in the App Store. The user can tap "Yes", "No", or "Maybe" (which means the request will come back again later).

I use KISSmetrics to track people's response to this request. In the past 30 days, the ratings request screen has been shown 164 times, and 21 of those people tapped "Yes" (13%)! And, 46% of people tapped "Maybe", so I think the inclusion of that option is very important. I guess that if people were presented with only a Yes/No option, most Maybe's would be No's.

Right now, the average rating for both the current version of Pouch and all version is 4.5 stars. I don't think I could ask for better, as it's just about impossible to avoid the occasional unhappy customer who gives you a 1-star rating. My most recent 1-star rating came with this review:

"I did a search in the app store for 'basecamp' and saw great ratings but did nit [sic] notice it was for backpack. Yes, bad user but why did it show up in the search with that keyword? Not sure who's fault this is but certainly this is not the key driving all the success on this platform."

So, the guy searched for a Basecamp app, found mine (titled "Pouch, for Backpack"), and decided I was to blame for his confusion. Oh well, can't please everybody.

And next, here's what I think has been the key to good ratings (aside from having a good product!):

Make It Easy to Get in Touch

Early on, I included a button within Pouch titled "Support" (now it's got an icon instead). When tapped, it brings up an email form addressed to me, pre-filled with info about the customer's version of Pouch, iOS, and whether they're on an iPhone or iPad. I get an average of 1 email per day thanks to this button.

This Support button has served so many purposes: Some customers have told me about bugs. Many customers suggest additions to Pouch. A few people write to just tell me how much they like Pouch (and then I ask them to leave a review!).

Without that button, I probably wouldn't have heard half of this stuff. When I first considered adding the button, I was concerned about how much email I'd have to deal with. And I have to admit, there have been a couple of times where it was hard to keep up with. But I think it's been crucial to the success of the app, and it's taught me a lot.

It's Both Wonderful and Frustrating to Sell a 3rd-Party App

Pouch is a 3rd-party app for using Backpack, made by 37signals (if you've read this far, you probably know that). Being 3rd-party has been a blessing and a curse:

I've done almost no marketing. Pouch has a built-in customer base: everyone who uses Backpack and has an iPhone or iPad. And the Backpack website has an Extras page where most of that customer base is going to look when they need a Backpack app. I think the Pouch link on that page is responsible for most of the sales I've made. So I have to say a big thank you to 37signals not only for making Backpack and its API, but for giving me so much traffic! It's been awesome. On the other hand:

My hands are tied with regard to many features. This is the "curse" side of things. People would like to be able to upload files, manage picture galleries, leave comments on things, read their Writeboards, and more. And hell, I would love to make that possible! But I can't, because none of those features are available via the Backpack API. That means that a lot of requests I get from customers have to be answered with "I'd love to add that too, but I can't, because...". Luckily, most customers are understanding about that (even if they don't know what an API is). But it is a little frustrating.

Moving Forward

I hope I can get to 2000 sales! Moving forward, I'm going to keep working to make Pouch faster and more full-featured. And along the way, I'll be happy to share what more I learn with you. If you have any questions, leave them in the comments.

Audiences:

Tags:

Products:

C2G 1.1 Fixes Crashes for Vancouver Customers

Chadwick Wood — Fri, 23 Sep 2011 17:54:50 +0000

I'm a little late on posting about this, but last week the version 1.1 update to C2G became available, which fixes the crashes that Vancouver customers were seeing for a couple of days.

The crashes were happening because the data coming from car2go about available vehicles in Vancouver contained some invalid location information for some vehicles. I wrote to the car2go tech people about this, and they actually fixed the problem on their end, which stopped the crashes in C2G. But in case this kind of thing happens again, version 1.1 of C2G includes error checking code to prevent crashing.

Now, if I could just get in contact with Jeffelliot, who experienced such a crash right after he bought C2G :( Jeff, I fixed it! Please give me a better review!

Tags:

car2go

Products:

C2G

App Store Customers Having Troubles Leaving Reviews

Chadwick Wood — Tue, 30 Aug 2011 16:30:18 +0000

Yesterday I received an email from the second customer informing me that they've bought one of my iPhone apps (Pouch, in this case), tried to leave a review in the App Store, and got an error message stating "To rate this item, you must own it." This customer is in the UK.

A few days before that, I heard from a customer in the US who had purchased C2G and was getting the same error message when trying to leave a review. So far, I have no idea what to do about it.

Two customers in two different countries buying two different apps and getting the same error? This got me wondering, so I tried something myself: I have purchased copies of C2G, Pouch, and Glory Math 1, and guess what? I can't rate any of them! I get the same error message about not owning the app.

So apparently all of my apps can't be reviewed right now. I've contacted Apple about it, so we'll see what they have to say.

Audiences:

Tags:

Now Available: C2G, for car2go on your iPhone

Chadwick Wood — Thu, 25 Aug 2011 21:01:25 +0000

The latest iPhone app from Coffeeshopped is now available: C2G, for car2go. C2G allows you to view a map of available cars, parking spots, and gas stations in your city, and to make short-term reservations. It currently works in Austin, Vancouver, Ulm, and Hamburg (and I'll add more cities as car2go makes them available).

When I began work on C2G, there were already 3 (or more) iPhone apps out there for using car2go. I've used two of them, and frankly they just weren't very good. The most popular one available has a cluttered interface (e.g. who needs satellite view on a map when you're looking for a car?), and a strange feature where, once you reserve a car, there's no way to see where it is from within the app. Anyway, I don't want to talk a bunch of trash; the point is that it was pretty straightforward to make an app that better serves the basic needs of a car2go user. I hope C2G is exactly that.

See some screenshots on C2G's App Store page.

Tags:

iPhone

car2go

Products:

C2G

Pouch 2.3 Now Available, with a New Icon

Chadwick Wood — Sat, 09 Jul 2011 19:03:21 +0000

It's Saturday, but I just don't want to wait until Monday: Pouch 2.3 is now available on the App Store!

In terms of features, this release includes the addition of the Newsroom, letting you see an overview of the latest activity on your Backpack account. Also, by popular request, you can now turn on and off individual calendars in the Calendar section. You can add and delete Dividers on your Pages now, too. And there are a few other improvements here and there, as well.

This release also includes a new (vastly improved) icon for Pouch, by the wonderful Lauren Herda. Lauren did an amazing job and was a pleasure to work with, and I think everyone using Pouch will appreciate having something pretty to look at on their Home screens. Here's the new icon, in its 512-pixel glory:

Tags:

Products:

Pouch Update: Now with HTML and Textile Support

Chadwick Wood — Fri, 29 Apr 2011 16:09:06 +0000

Pouch 2.2 is now available on the App Store. This update includes support for HTML and Textile in your Page content, so any fancy formatting you've done on the web should now also show up in Pouch! Give it a try.

Tags:

Products:

Increase the PHP Memory Limit on a Dreamhost Virtual Private Server (VPS)

Chadwick Wood — Tue, 26 Apr 2011 17:52:26 +0000

If you've ever seen a PHP error like "Fatal error: Allowed memory size of xxx bytes exhausted..." on one of your websites, then you might need to increase the memory limit for PHP on your web server. Depending on your web host, this may or may not be possible. If you have a Dreamhost VPS, then luckily it is! Here's how you do it.

Create an Admin User for Your VPS

To edit the PHP settings on your VPS, you'll need an admin user so that you can get root access via SSH. To do this, go to your Dreamhost Panel, and select "Manage Admin Users" under the VPS menu on the left. Fill out the short little form, and you're done. It takes a few minutes for this process to complete, and you'll have to wait until it is complete to do the next steps. So, give it 5 or 10 minutes before moving on.

Manage Your Own PHP

The next step is also in your Dreamhost Panel: turn off Dreamhost management of PHP on your VPS. By default, Dreamhost auto-generates the configuration file that PHP uses. Since we want to manually edit that configuration, we need to turn Dreamhost management off. Go to "Configure Server" under VPS again, and find the section of the page for "PHP Configuration". Un-check the "DreamHost Managed" box, and save. Again, you'll have to wait a few minutes to complete.

Find out Which Configuration File to Change

My VPS had multiple versions of PHP installed on it by default, and yours probably does too. So, you need to figure out which configuration file (php.ini) needs to be changed. The easiest way to do this is to have a page on your website somewhere that calls phpinfo() to show you how PHP is set up for your website. I'm going to assume you can figure that part out. Once that's set up, visit the page that calls phpinfo() and look for the line (near the top) that reads "Loaded Configuration File". That's the path of the configuration file you need to edit.

Edit Your Configuration File

Now that you know where to look, SSH into your server as the new admin user you set up, and edit that file. I use nano for command-line editing. So, if your phpinfo() page says your config file is at /etc/php5/cgi/php.ini, then just type sudo nano /etc/php5/cgi/php.ini at the command line to get in there. Note that you need to include "sudo" at the beginning (and then enter your admin password) to get permission to edit this file.

Once you're in, find the line that starts with "memory_limit", and change the number on that line to what you want your new PHP memory limit to be. It'll most likely end in "M" (for megabytes). Leave that "M" in there. Save your file, and you should be done.

Check Your Changes

After you've saved your config file, go back to your page that calls phpinfo(), and check to see that the line that reads "memory_limit" now shows the new value you entered. If it does, you should be all set! (And if not, check to make sure you didn't mess any steps up).

I hope this helps! If you're like me, you might know most of this stuff already, but it's nice to have it all laid out step-by-step. Leave any feedback in the Comments section.

Audiences:

Web Developers

Tags:

PHP

Dreamhost

Pouch 2.1 Now Available in the App Store

Chadwick Wood — Mon, 11 Apr 2011 19:08:41 +0000

Pouch 2.1 became available on the App Store today. This update includes several improvements and bug fixes, mainly for the Calendar and Reminders. Get Pouch 2.1 from the App Store.

Tags:

App Store

Products:

Pouch

Colored Images: Documented Source Code

Chadwick Wood — Wed, 09 Mar 2011 20:36:45 +0000

$3.00

ColoredImages is an XCode project that demonstrates how to dynamically color a grayscale UIImage just using code (as described in iPhone: How to Dynamically Color a UIImage), and also includes a bit of basic animation. The source code is well-commented and compiles into an example app to help demonstrate what's going on.

SKU: coloredimages

List price: $0.00

Price: $3.00

Weight: 0 lb.

Dimensions: 0in. × 0in. × 0in.

Taxonomy upgrade extras:

CacheExclude: Disable Caching for Individual Drupal Pages

Chadwick Wood — Wed, 23 Feb 2011 17:52:53 +0000

On a recent Drupal-based project, I needed to disable caching for just the home page, because of some dynamic content it had that the caching was preventing from updating. CacheExclude makes that dead simple. It's available for Drupal 6, but it's not clear whether there's going to be a Drupal 7 version.

URL:

KISSmetrics for iOS on GitHub

Chadwick Wood — Sat, 29 Jan 2011 01:31:47 +0000

Today I was finishing up work on version 1.4 of Pouch, and one of the things I wanted to add was some basic analytics tracking, to see how often the app was getting used (basically, just a count of how many times people launch the app). Since I'm already a KISSmetrics user, I figured that would be the best way to go. Unfortunately, there was no existing Objective-C library for interacting with the KISSmetrics API that I could find... So, I wrote one. Fortunately, it didn't take very long, as the API is pretty simple. If you're also a KISSmetrics user, and you work on an iPhone/iPad app, you might want to check out KISSmetrics-for-iOS, which I published on GitHub today.

If you use it and/or have any feedback on it, please let me know.

Audiences:

Tags:

Faster Scrolling for UITableViews

Chadwick Wood — Wed, 26 Jan 2011 19:13:53 +0000

A useful article by the creator of Tweetie on how to get better scrolling performance for UITableViews in your iOS apps. In short: just use one custom view within each UITableViewCell (instead of composing your cells of multiple subviews).

URL:

Fast Scrolling in Tweetie with UITableView

LROAuth2Client: OAuth2 for iPhone and iPad applications

Chadwick Wood — Mon, 24 Jan 2011 23:22:18 +0000

LROAuth2Client is a great little library put together by Luke Redpath to enable iOS developers to use OAuth2 in their apps. I'm using it in Pouch.

URL:

Coffeeshopped RSS Feed Has Moved

Chadwick Wood — Tue, 11 Jan 2011 18:19:42 +0000

This is just a quick note for my RSS subscribers: the Coffeeshopped RSS feed has moved. You can subscribe to the new feed here.

Saving Sessions with the Facebook iOS SDK

Chadwick Wood — Wed, 05 Jan 2011 20:06:01 +0000

Recently I added Facebook image sharing to Glory Math 1, and it was my first experience with the Facebook iOS SDK. I used the newer Sign Sign-On feature, which briefly takes the user to the Facebook app to ask them whether your app can access their Facebook profile. The docs are good, and the whole process was pretty straightforward, but one aspect of the authentication process really annoys me: Every time my app launches again, the user must go through the authentication process with Facebook. The first time the user authenticates, it'll ask for access, but every time after that, it just tells the user that they already approved the application for access, and they just have to tap "Okay". This process seems pointless to me. So, I tried to find a way around it. The cause of having to go through the sign on process every time the app is re-launched is that, after successfully signing on the first time, the Facebook SDK doesn't save the access token it receives in a persistent manner. So, when the app quits, your token is gone. This is easy enough to get around. Rather than create my own example project to demonstrate, I'll just specify how you would change the Facebook Demo App to store the access token. All it takes are changes to a couple of methods in DemoAppViewController.m:

// around line 25
static NSString* kAppId = nil;
// use these as the keys for storing the token and date in user defaults
#define ACCESS_TOKEN_KEY @"fb_access_token"
#define EXPIRATION_DATE_KEY @"fb_expiration_date"


/**
 * Called when the user has logged in successfully.
 */
// around line 193
- (void)fbDidLogin {
    [self.label setText:@"logged in"];
    _getUserInfoButton.hidden = NO;
    _getPublicInfoButton.hidden = NO;
    _publishButton.hidden = NO;
    _uploadPhotoButton.hidden = NO;
    _fbButton.isLoggedIn = YES;
    [_fbButton updateImage];

    // store the access token and expiration date to the user defaults
    NSUserDefaults *defaults = [NSUserDefaults standardUserDefaults];
    [defaults setObject:_facebook.accessToken forKey:ACCESS_TOKEN_KEY];
    [defaults setObject:_facebook.expirationDate forKey:EXPIRATION_DATE_KEY];
    [defaults synchronize];
}

/**
 * Show the authorization dialog.
 */
// around line 88
- (void)login {
    // on login, use the stored access token and see if it still works
    _facebook.accessToken = [defaults objectForKey:ACCESS_TOKEN_KEY];
    _facebook.expirationDate = [defaults objectForKey:EXPIRATION_DATE_KEY];

    // only authorize if the access token isn't valid
    // if it *is* valid, no need to authenticate. just move on
    if (![_facebook isSessionValid]) {
        [_facebook authorize:_permissions delegate:self];
    }
}

Making these few small changes will store the access token and attempt to re-use it.

Better, But Still Annoying

There's a big gotcha, though: the expiration date that Facebook gives the token appears to be about 24 hours from when it was issued. So, this technique will only save the user from repeat authentications in a 1-day period. It's an improvement, but still pretty annoying. Maybe down the road Facebook will start issuing tokens that work for longer. Also, I've only tested this when requesting the publish_stream privilege for an application. I'm wondering if maybe the token you get when requesting offline access might have a longer life.

If you've had any different experience, or know of a better way to not force the user to authenticate over and over, let me know in the comments.

Audiences:

Tags:

Products:

Pouch 1.0.1: Bug Fixes

Chadwick Wood — Tue, 28 Dec 2010 22:46:10 +0000

This is just a quick note to Pouch users out there: I just submitted the Pouch 1.0.1 update to the App Store, which includes a fix for errors that said "Can't connect to the internet" or "Authentication needed". Sorry about that. The update will be available as soon as Apple approves it. Thanks for hanging in there with me!

Tags:

App Store

Update

Products:

Pouch

Pouch, for Backpack, now Available

Chadwick Wood — Tue, 21 Dec 2010 16:55:27 +0000

Today my new iPhone app, Pouch, is available. From the product page:

Pouch is an iPhone app for use with 37signals' Backpack. Use it to read your Backpack content in a simple, familiar format, and to create and edit your Backpack Notes.

Tags:

Products: