Carving Statues Of Elephants

Published
by
James Thomson
on December 19, 2008
in Development
. 7 Comments

Ok, since it’s the festive season, I’ll do the good news first before ranting a bit:

  1. You can now download PCalc Lite, a free and only-slightly-less-awesome version of PCalc for the iPhone and iPod Touch. It’s available on the App Store here.
  2. Despite being a new app, was approved by Apple in a mere three days from submission to being on the store.

And the bad news:

  1. Despite being a new app, the release date was set to the day of submission, the 16th, not the day it actually turned up on the store. So, it’s already on page 3 of Utilities and won’t get the “new app bump”. I don’t know if the release date trick works for new apps, but I’ve given it a try.
  2. PCalc 1.3, an update to an existing app, is nowhere to be seen, lost in the submission queue. Well, so much for publicising them together.
  3. It’s almost impossible to plan a product release in advance with the App Store. I mean, don’t get me wrong, a three-day turnaround on a new app submission is really good to see. It’s just that it could have easily turned up next year. I was taken by surprise this morning and was scrambling to update the website at 8am.

Ok, let’s back up a little bit and talk about PCalc Lite. My PR spiel for the release says:

Available for a limited time only, PCalc Lite is a fully functional and free taste of our popular scientific calculator for the iPhone and iPod Touch.

It includes an optional RPN mode, multiple undo and redo, unit conversions and constants, as well as two stylish themes and our highly praised design.

If you like what you see, the full version has many more settings and themes, a paper tape, engineering and scientific notation, and full support for hexadecimal, octal and binary calculations.

In the continuing absence of demos or trial periods in the App Store, the idea behind PCalc Lite is to provide something useful and free that’s better than the Apple calculator, but which will also drive interest in our full version. PCalc Lite is a great calculator, but PCalc is even better!

So yes, it’s basically an advert for PCalc, but it’s actually a pretty solid calculator in its own right and miles better than the default Apple one. And did I mention it was free? So, everybody should go and download it. I can say it’s awesome all I like, but just go try it and see what you think yourself.

This is of course, another in a long line of big marketing experiments. I’ve given away something significant with the hope of making more money by doing so. The real question is, have I got the right balance between removing functionality and making it useful, so it doesn’t completely kill sales of the full version?

As the old line goes:

How do you carve a statue of an elephant? Start with a block of marble, and remove everything that doesn’t look like an elephant.

So, I took PCalc and chipped away at all the features. I decided early on to keep the RPN mode which is probably the most dangerous decision. It’s something that makes it quite distinct from the Apple calculator, but it’s also an important selling point to a lot of people. I figured if I took it out, people might say that there’s no reason to download PCalc Lite over the Apple calculator in terms of features. I’d disagree politely, of course. But some of my beta testers expressed the concern that with the RPN mode in there, PCalc Lite does everything they need and they wouldn’t need to buy the full version.

All of the other options are gone, with the exception of the “easier to read digits” option and two themes. I made a new “Twilight” theme to be the default for PCalc Lite, mainly because I needed to create a distinct new icon for it, and wanted them to match. Attention to detail and all that. In the end, I ultimately added the theme as an option in the full version of 1.3 as well, as it looks pretty sweet.

I also trimmed the constants and conversions down by half, taking out all but the basics, and adding a small polite note in those sections saying you can get more in the full version.

Hex, octal and binary modes gone. Tape and stack / register display gone. Two-line display gone. Engineering and scientific notation gone. Even thousands separators. I took a positively Jobsian approach to the removal of options and features.

But the core user interface is all intact, so you should be able to get the feel of the application which I think is very important. Like the way undo and redo is implemented with swipes on the LCD, or the way the parentheses display works. The subtle details that make the difference, or so I hope.

Many of my testers commented that I should keep all the features in there, but if you tapped on anything outside of the Lite stuff, I should throw up an annoying alert saying that it’s only available in the full version.

That’s tempting, but I don’t think Apple would allow it into the store. As it is, I thought I was sailing close to the wind, given that @chockenberry had already told me about the problems they’d had with the free version of Twitterific. I think it might be different for PCalc because my Lite version isn’t ad-supported, it’s just plain free. Or I might have just been lucky.

But it was important to me to make something that was a complete product in its own right, with some gentle upselling to the full version. I’m hoping that decision will make PCalc Lite very popular, perhaps even Top 100 popular in the free apps section if I’m lucky, and I’ll make the sales up due to the sheer number of people exposed to the wholesome PCalc goodness.

I’ve said it’s only available for a limited time, but to be honest, that depends on how popular it is, and what it does for sales. If sales nose-dive from the get go, it’s going to be pretty darn limited. If sales go up significantly, then I don’t see any reason to remove it.

So, today will be spent promoting PCalc Lite, trying to give it a good start in life, and hoping this form of altruistic capitalism pays off. Wish me luck!

UPDATE - 2pm GMT

Yes, the release date trick works even for new apps! Currently PCalc Lite is on the front page of the new releases in the Utilities section of the iPhone store, though the iTunes view hasn’t updated yet. I expect it will as data gets pushed out to the servers. Ok, now to try and get some press coverage…

UPDATE 2 - 8pm GMT

Oh, App Store, you are so loveably unpredictable! I just got an email there saying that the full PCalc 1.3 release is also now approved and “Ready for Sale”. I’ve never had one of those emails at this time of day before. I wonder if somebody in Apple reads this blog?

Anyway, it’s not appeared on any of the stores as yet, but presumably it will be there soon. I’ll wait before updating pcalc.com though.

Of course, I sent out all the PCalc Lite PR emails about four hours ago, and Macworld has run a story at least so far. I can’t really send out another one now. Will assess the situation tomorrow I think…

Share/Save/Bookmark

Go Towards The Lite

Published
by
James Thomson
on December 8, 2008
in Development
. 4 Comments

Apologies for my silence over the last few weeks gentle reader, I caught the man-flu that’s going around these parts, and haven’t felt up to much until now. Two weeks ago, I started working on a PCalc 3.4 release, and in less than a day of highly productive work I had ported the themes from the iPhone version back to the Mac. 24 hours later had a temperature of 39°C and I haven’t touched the code since… That’ll teach me. At the moment, I just have the tail end of a cold, so it’s back to work.

Anyway, one of the things I was mulling over in my feverish state was the possibility of doing a free “Lite” version of PCalc for iPhone.

In the continuing absence of demos or trial periods, I was thinking I should create something useful to give people a taste of the full awesome PCalc experience and encourage them to buy it, but not quite so full or as awesome as to cannibalise our sales. The App Store is filling up with such little tastes at the moment, so it certainly qualifies as a trend. And I’ve heard at least some anecdotal evidence from other developers that lite versions can increase sales.

This is more a “pre-mortem” rather than post - I haven’t started working on it yet, but I thought I’d encourage some debate before I do.

There are two big questions really. The obvious one is “what functionality should I remove?” and the slightly less obvious one is “what will Apple actually let me put in the store?”.

Erica Sadun wrote a nice guide on Ars recently entitled “App Store lessons: creating demos for fun and profit“. She says:

“Demos help sell products for very little overhead. Developers need only cut down their feature set, change a few options and ship the result out to App Store. From a financial point of view, demos are made of win.”

But I’ve also been hearing from illustrious people like Craig Hockenberry that Apple is insisting that “free and paid versions have to be feature equivalent” and that explicit upselling language isn’t allowed. The Iconfactory has a free ad-supported version of Twitterific as well as a premium version at $9.99. They are pretty much identical in terms of functionality, with the exception of no adverts in the premium version and an additional theme. Apparently that’s all Apple would allow them to do.

So which is it? Clearly, I can’t just produce a feature equivalent version of PCalc and give it away for free. Ed Voas wryly suggested I just remove the “9″ button, which isn’t actually a bad idea in terms of letting people try everything without giving away the farm, but I suspect that Apple wouldn’t find it quite so droll. 

So far, I haven’t been able to track down a definitive policy statement from Apple on the matter. It’s easy enough for a game - you just include the first couple of levels and get people wanting to play more. And you can do that while staying within the (apparently unwritten) guidelines.

So what is PCalc Lite going to be then?

I could ship it with some very basic layouts for a start - remove functionality that way. If you think of a layout as a game level, that kind of works. I’ll lose the binary, octal and hex, some of the fancier scientific stuff.

Likewise with the themes - just ship with the one, probably the original “Blue Sun” one. Most people seem to prefer the fancier coloured ones, so it’s a chance to convince people to upgrade there. Likewise, get rid of the LCD colours, and the two-line display option.

Conversions and constants, also should go - or perhaps just include one or two as a demonstration.

I should probably keep the RPN mode though, that means I’ll have something to offer over the built in calculator.

So yes, I think the idea should be to provide a very simple RPN calculator that hints at what you will get with the full version of PCalc, but is still quite useful in its own right. And if it proves to be really popular at the expense of sales, I guess I can always kill it!

Any thoughts, please add them below.

Bonus trivia, Apple fact fans. I made a DragThing Lite many moons ago, back in the days when I was working at Apple. I was asked by management to put one together - to potentially ship as part of Mac OS 8.5 - when it didn’t look like the in-house app switcher would be ready in time. I made a quick prototype over a weekend, but in the end I think I remember the development time was extended for the OS in general and it wasn’t needed. So I just shipped the Lite version myself with DragThing 2. But yeah, DragThing was nearly part of the OS.

Share/Save/Bookmark

Just A Big Popularity Contest

Published
by
James Thomson
on November 21, 2008
in Development
. 3 Comments

With every positive thing Apple does, there’s always something else that seems to annoy me. Just installed the 2.2 OS which is great, and have also been playing around with the new phone version of the App Store. Some interesting changes, but not all good from a developer perspective.

The main thing I noticed was that in the updated “Categories” view, there’s now a new switch at the top to sort apps in that category by “Top Paid”, “Top Free”, or “Release Date”.

Previously, categories were only sorted by release date, but now the default view for a category is to show off the top paid apps. And if you set the category to list by release date, it will forget that setting the next time you go back into the store, and present you with the top paid apps again.

What does this mean? Well, on the phone at least, it means that most people when they go into a particular category will first see the top 25 paid apps, and then will probably look at the top 25 free apps. Possibly they’ll look at the new stuff, but it’s no longer the first (and only) thing they’ll see. So, if you have a brand new release or an update, there’s less opportunity for people to see it. Which is of course not going to be good for your sales or indeed you chances of ever getting into that elusive top 25 club…

Ok, James, but how is giving the user more control over what they see in the store a bad thing? You were only just complaining about that last week. Make your mind up.

Well, part of the problem with the store at the moment in my opinion is the strong emphasis on the top selling apps. It’s like an echo chamber - the top selling apps are prominently displayed in the store, so they sell very well. If you’re not a best seller, then you get relegated to the dark corners of the store and thus don’t sell very well. That’s some catch, that Catch-22.

Something like PCalc which is pretty specialist - calculators just aren’t sexy I’m afraid - isn’t ever going to make it into the top 25 bestseller list for a big undefined category like “Utilities”, so will have a hard time getting noticed if that’s the focus. Of course, if you actually made a separate “Calculators” category, it would probably be close to the top.

In iTunes, the default category view is still by release date, and the other options are for “Most Popular” and “Name”. I do wonder if the new category sorting options are due to appear in iTunes too. It would make sense for the two stores to be consistent. But if they do, then I think the “Availability Trick” isn’t going to be worth much, as people just aren’t going to see what’s new anyway unless they actively go looking for it.

I’m sure this is perfectly normal retail economics, and I wouldn’t actually be complaining if I was actually in the top 25, being featured by Apple in commercials, keynotes, and their new iPhone Your Life site. But I’m not, so I am.

I want a store sorted by quality, not by popularity. Is that so hard?

Share/Save/Bookmark

A PCalc 1.2 Post-Mortem? Not Yet

Published
by
James Thomson
on November 18, 2008
in Development
. 3 Comments

Okay, it’s a little early for another post-mortem, it’s only been out just over 24 hours. But day one sales of PCalc 1.2 were ten times that of PCalc 1.1 - which was arguably a bigger deal in terms of new features than the admittedly cool additions of 1.2.

So far, press coverage has still been somewhat limited despite my plea, although Mr Snell and Mr Gruber were most kind yesterday. Early days, anyway.

More analysis soon, but I’d have to conclude that getting PCalc onto the front of the Utilities section on the store was clearly very important.

Share/Save/Bookmark

A PCalc 1.2 Press Release

Published
by
James Thomson
on November 17, 2008
in Development
. 4 Comments
Dear Member of the Press,
I know what you are thinking, “Not another iPhone press release, and for a calculator at that!”. But, please bear with me for just five minutes to hear my impassioned plea.
This isn’t some throwaway application I put together in a weekend - it’s an application that’s had over six months of solid development love over five releases, itself built on over sixteen years of coding. It currently has over sixty 5-star customer reviews on the App Store, and thousands of die-hard fans on both the iPhone and the Mac.
So, yes, it’s a calculator. But it’s not just another calculator.
I think we got off to a bad start last time with version 1.1. It was my fault entirely. I came out with this big release I’d spent months on, sent out my press releases, and not a soul ran a story with the exception of that nice Mr Gruber fellow. Sales were quite poor and we both said some things we regretted in the morning.
But I understand, don’t worry. You’re tired of the deluge of iPhone press releases, and everything you see just starts to look the same after a while. I didn’t do anything to make it an interesting read and, let’s face it, I’m kind of fighting a losing battle with a calculator which - I’ll be the first to admit - is next to the flashlights and to-do list apps in terms of reaching saturation point on the store.
So here’s the thing. Send me your iPhone’s device ID (*) and I’ll send you back a copy of PCalc 1.2. Have a play around with it and if you don’t think it’s the best app you’ve ever used on the phone then that’s cool. But if you like it, how about some coverage? We can put the whole 1.1 incident behind us, and move on together.
Hugs,
James
(*) The “Ad Hoc Helper” app on the store is a good way of finding it.
PS, I nearly forgot, here’s the press release:
TLA Systems Ltd are pleased to announce another significant new release of their advanced calculator PCalc for the iPhone. 
It’s available to download today from the App Store. Review copies are available on request.   

What’s New?
Version 1.2 includes an optional two-line display for use with the RPN mode, so you can see the first two items on the RPN stack. This was actually the number one feature request after the initial release of PCalc on the iPhone, and we’re happy to add it for all the engineers out there.
There are two new calculator themes included too - “Seventy Three” a sleek dark retro style, and “Rough Draft” a fun drawn blueprint look which will be rather familiar to DragThing users.
The full list of changes is here:   

   <http://www.pcalc.com/iphone/history.html>

Also, for those people who want to see what goes on behind the screenshots, our brand new developer blog is available here:
Screenshot
Here’s a shot of the two-line display, complete with the new “Seventy Three” theme:
More Information   

PCalc for iPhone is available to buy today at $9.99 from the iTunes Application Store at the following link:

   <http://phobos.apple.com/WebObjects/MZStore.woa/wa/viewSoftware?id=284666222&mt=8>

More information and screenshots are available at:

   <http://www.pcalc.com/>

Thank you for your interest.

Share/Save/Bookmark

Predictably Unpredictable

Published
by
James Thomson
on November 17, 2008
in Development
. 0 Comments

So, PCalc 1.2 was submitted to the App Store last Thursday, and the Apple approval process before it shows up on the store has taken exactly six days for the last four releases. So, I figured it would appear on the store around Wednesday this week. Plenty of time to write some PR and plan my marketing strategy then.

Apparently not.

Woke up this morning to find the “Ready for Sale” email from Apple sitting smiling in my inbox. So, four days from submission to being on the store this time, and that included a weekend.

Given that developers have been so unhappy about how long the submission process takes, it seems somewhat churlish to complain about the process getting faster, but it’s still annoyingly unpredictable. At least it didn’t hit on a Sunday morning I guess.

I think Apple should at least give you an ETA when you submit something to the store, and perhaps daily updates as the submission queue changes. It would be nice to get a heads-up about when your software is going to be available to the public.

I’m currently scrambling to update the website with the 1.2 information. I’ve also submitted new screenshots via iTunes Connect and I’ve done the “Availability Trick” to hopefully move PCalc onto the front page of Utilities on the store. So far the store hasn’t updated with any of my changes, but fingers crossed that will happen soon and I can do a bit better from this release than I did with 1.1.

I’ll send out my PR this afternoon, as soon as I’ve actually written it. Wish me luck!

UPDATE

Ok, App Store has updated with metadata changes, and we are currently the second app listed on the front page of Utilities on the store on iTunes. Excellent! Oddly though, still not listed on front page of the phone store. I’ve seen changes happen on one before the other, so it might just be lag between different servers. We’ll see. Anyway, stage one completed - PCalc is up and visible on the store.

UPDATE 2

PCalc now listed as the first app in the Utilities section on the phone store. Result! Now all I need is to get some PR out this afternoon and we’ll see what difference this all makes to sales today. Feeling good…

Share/Save/Bookmark

The App Store Crunch

Published
by
James Thomson
on November 16, 2008
in Development
. 0 Comments

Interesting article by Andy Finnell about iPhone app pricing, and how the current race to 99c isn’t sustainable.

PCalc started out at $9.99, and is still at $9.99, because I think that’s a fair price for the amount of effort I’m putting into it. It’s actually a little on the cheap side even, but I figure I can write off a good chunk of time spent learning the iPhone so I can hit the ground running for my next app. Or just sell my consultancy services to the highest bidder of course…

Having said that, I did also laugh at this satirical article on iPhone Footprint that references PCalc:

Apple has done it again. Recently the company announced that it has become the number one distributor of flashlight and calculator applications through the App Store. Scott Forstall, Senior Vice President of iPhone software said that 20 million flashlight and calculator applications were downloaded in October, 2008. He termed this as a rare feat for the Company.

Well played sir, well played.

Share/Save/Bookmark

Attack Of The Clones

Published
by
James Thomson
on November 12, 2008
in Development
. 0 Comments

One neat Twitter trick to try is using RSS feeds with http://search.twitter.com/ to track the use of keywords that are important to you. For example, I’ve got feeds on “DragThing” and “PCalc“, and also on “@jamesthomson” to catch any tweets I miss that reference me.

And yes, @jamesthomson is me - feel free to follow me to boost my vast programmer ego even further.

Anyway, first thing this morning as I was heading out to the gym, Twitter flagged up the following tweet by @MDMstudios:

Xcode has been crazy all day. At least we got PCalc ( the calculator ) to Apple finally.

“What? What?! WHAT!?!” I said, channelling an alarmed David Tennant. “I’m just about to submit my 1.2 release today, but I’m pretty sure I didn’t actually do it yet…” Reading on, it looks like these folk have written a programmer’s calculator for the iPhone, called it PCalc, and just submitted it to the App Store.

OK… Well, that’s going to be interesting… It was only a month or two ago, somebody said they were working on a new application for Windows called DragThing and I had to send them a little note.

I’ve sent a polite email to the first address I found, and hopefully I can convince them nicely that pulling the submission and renaming it to something else might be best for both of us, without any threats of lawyerly smiting with trademarks required. Luckily, I caught it at the submission stage, rather than having it appear on the store itself.

Anyway, it got me thinking - I wonder if the Apple submission process actually checks such things? Could I just make a new app, call it, say, Twitterific and submit it and get it onto the store. Yes, of course The Iconfactory have a trademark on the name, and there would be legal recourse, but would Apple reject it before it reached the store?

Or does Apple consider that’s not their job, and will leave it to the lawyers to sort out afterwards? I wonder.

Hopefully there will be a reasonable ending to this particular story anyway. Some further investigation suggests that the developer in question is actually 14. Which depressingly means that they are younger than some of the code in PCalc, which I started writing in 1992 when I was at university…

UPDATE

I just got an email back from them, and they are going to change the name, and apologised for any trouble. Phew!

Share/Save/Bookmark

Relativity

Published
by
James Thomson
on November 11, 2008
in Development
. 1 Comment

Apparently, I haven’t written anything here for nearly two weeks which comes as a bit of a surprise to me… When I am coding, time seems to flow quite differently. Anyway, the one of the two main reasons for my silence is that I’ve been working away on a new 1.2 release of PCalc.

It’s got an optional two-line display mode for the RPN folks, and a couple of new themes. It’s a nice update I think. The two-line display has been a very common request since the iPhone version came out, but there was a large number of things I needed to do to make it possible. Most of that work happened in 1.1.

In order to support a variable height display, the button area below it needed to be resizable rather than having all the positions hard-coded as in 1.0.x. That led to writing the dynamic layout code which as a bonus meant that I could support multiple different layouts.

Also, in 1.0 the buttons were all drawn using pre-rendered PNG files. That worked fine, but it meant that every time I introduced a new button, I needed to create a new background PNG file of exactly the right size, as scaling the graphics made them quite blurry. A better way of doing it would be to draw the buttons on-the-fly in code, so they could be any size at all. Luckily, I had just the technology for doing that in the DragThing theme code, so I used that. And, as a bonus I could now support multiple calculator themes.

Point being that all this started with the requests to do a two-line display, and that simple request led to rewriting large chunks of PCalc to make it much more flexible, and now the original feature that started it all is finally done. This makes me happy.

PCalc 1.2 isn’t as big an update as 1.1, but I’ve spent less than a month on it rather than the two months I spent before. I think that’s a reasonable goal for doing iPhone application updates. In two months, you slip so far down the listings that people forget about you.

Hopefully, the release will also improve sales, and I plan to try every trick in the book to get it noticed this time round! It should be submitted to the App Store within the next few days, and I expect it will be on sale in about a week.

Now, while I wait for my beta testers to give me the ok, I will return to the second reason I’ve been a bit quiet - exploring the wastelands of Fallout 3 :)

Share/Save/Bookmark

Interview With A Developer

Published
by
James Thomson
on October 28, 2008
in Development
. 0 Comments

Here’s a quick interview I did yesterday with Rene Ritchie at the The iPhone Blog - he wanted to chat about the App Store and how it could be made better for both customers and developers. I think the key is more intelligent filtering to separate the wheat from the chaff.

Currently thinking about PCalc 1.1.1 and what to add next. More soon.

Share/Save/Bookmark