WebIS, Inc. Forums: WebIS, Inc. Forums -> Alex's Blog

The iPad is a new market. The market for the iPad is not necessarily iPod Touch or iPhone users, though of course many of those users will buy one. Its not for tech geeks who need something as close to a real laptop - just smaller - as possible. Its for everyone else. And that means its going to be popular. The price point and capability will frankly mean it will easily outsell all Macs combined. There are so many people out there who never need a full sized computer, but they have one to check email, web browse and do cool things. The iPad will run iPhone apps and iWork. It'll have AAA games from EA and others. It'll run iPhone apps that do 150,000 different things. Honestly for the vast majority of consumers the iPad will be all they ever need. No mess. No fuss. No anti-virus. I already know of two families thinking about ditching their PCs for an iPad. Everybody who has thought "bah, what will it do for me?" and then touched it came away amazed.

The best part about the iPad is that it still uses the iPhone OS. That means great iPhone apps will run it now and optimized iPad apps will run on it later. It means developers like us can keep our codebase strong and tight and optimized to work with both. It means that we can keep a core app and really blow people out of the water for a large and small screen UI. Its not enough to simply let PI grow bigger on the iPad screen. We really have to redo a few things to make it work. But that's OK because that will make PI stronger on the iPhone and iPod Touch. Truly this is great news for iPhone users because it gets our creative juices flowing and we believe that some of the new UI elements on iPad will come down to the iPhone where it makes sense. Plus things like landscape iPhone will be able to match some iPad features and well, I'm just dancing with excitement here!

We will be bringing Pocket Informant to the iPad. We plan to bring a few of our other app ideas to the iPhone and iPad as well. We've been planning on some great new apps for the iPhone OS in 2010. The iPad just makes it that much greater.

At 2:30pm this afternoon Apple approved PI 1.21.11 for the App Store. Now it took about 3 hours after that point before people started to see it shown as an update on their phones. However, for many people if they hit update at that moment they would then get the OLD version instead of the new. This simply occurs because Apple has many many servers (hundreds) and the information has to be copied (propagated) from server to server and that doesn't happen instantly. Some servers handle the update info while other servers handle the actual app binary itself. So when some people saw that they had updates they hit an "info" server that had the new updated info, and then clicked update and that went to a server that gave out the binary files which had not been updated. It takes Apple a full 24 hours to get all their servers up to date (according to them, not me - I asked just today and got that answer).

As such we've been getting a lot of people yelling at us that the update didn't fix the problem. I realize that this is frustrating and everyone has been more than patient so I'm carefully asking that people be a bit more patient and wait for Apple's servers to fully update. If your update doesn't work, just wait awhile longer and try again.

Think about this - if the update didn't work, how would it have passed Apple's review? They wouldn't have accepted an app that by this point would simply point to our website. Finally, we are seeing many many people who got the old version after an update, waited a bit longer, continued to try updating (PI still shows up as needing an update) and finally getting the correct version. So please be patient and wait for Apple's servers to finish getting the update.

Thank you sincerely and very much.

I realize this issue was a result of my own mistake. I made the problem. But I think Apple is being difficult unnecessarily and that's saying it nicely.
December 9th: Beta code that was in an #ifdef section was being moved around. We suspect that I got interrupted and forgot to finish moving the code around to keep the beta stuff in the #ifdef. Then it was merged into the release branch. This should never have happened, but it did.
January 15th: The beta code caused PI to expire. We removed PI from the App Store completely worldwide.
January 15th: We submitted to Apple a new build with no expiration date.
January 18th: We are told our urgent review request was approved.
January 19th: PI is in review
January 19th: PI is rejected for sync text that was at the bottom of the sync "First Time or Reset Sync works best with a freshly started device to provide as much free RAM as possible."

This text was within PI for the last year and is there because at the time we found that some people would have better chances with a first time sync when they had a fresh device and many other apps we had used had the same text. We don't really need it anymore and I had actually removed it from our next release already.

But the point is that we have tens of thousands of users who can't use our app right now and Apple decides that its worth rejecting the update for something so inconsequential. For the record, we've responded to the rejection (we don't get many of those), we've uploaded a new build without that text, and now are waiting again.

Here is the link to the WDS Public Beta info.
http://www.pocketinf...showtopic=15981

I'm really liking my Nexus One. I still end up using my iPhone because the Nexus One just feels too complicated for some things (but compared to Android 1.0, 1.5, etc... its a ton better so I expect Google to be making huge strides), but I do love that screen and I'm a sucker for new things and animations.

I've had a few people tell me that PI for Android should be even easier than iPhone or BlackBerry to do because its so open, right? No, unfortunately not. Stuf like calendar and tasks are still off-limits on Android. We can use the open source system to get to their code and get access to it, but that doesn't do us much good as Google makes constant changes to that area and as we saw we didn't get Android 2.1 source until a week after the devices were out. Not good for that kind of thing.

But one thing that will make a dramatic difference is betas. I admit we rely on betas quite a bit for testing. A big part of that is simply that dealing with people's personal data means that we hit a ton of use-cases we simply never thought of or experience and betas are just the main answer to handling that.

Android will allow us to conduct public betas more easily than on the iPhone and for that I'm really happy.

We got a Nexus one in-house today. In general, there are aspects of this phone I love and prefer over my iPhone greatly, and aspects that I feel are greatly unfinished, unpolished, or I just don't like.

Loves:

• Screen. Fantastic. In most apps I didn't really see much difference from my iPhone, but then I went into Maps and....oh wow. Its incredible. I mean its....I don't know if I can ever use the iPhone maps app again.
• Android notifications
• The new Nexus One keyboard is FAR better than any other Android I've seen. Not as good as iPhone, but very very close.

Not so loves:

• Back Button. I just think its bad UI.
• No multi-touch in maps or anywhere else but keyboard. This is almost enough to negate the love above of the high res screen.
• I'm not really in love with the idea that I have to hit a menu button to get my menus and actions. I really prefer the iPhone way where stuff is always on screen. if there are menus, I get a control that shows that. I just feel like there are hidden parts to the apps. It doesn't make me feel right.
  Example is Amazon app. I open it and I get a screen like the first screen on the iPhone Amazon app. But what now? On the iPhone there are tabs at the bottom. I have by now figured out to hit the menu button and then hit a tab...but this just feels wrong again. Not liking that at all.
• Navigation on this phone just does not feel intuitive. Even with the 1Ghz snapdragon it also feels - even though its smooth - it feels like its just barely hanging on to my finger.
• App-wise it just feels really limited. Not much that it comes with. No clock, no weather, no stocks, no lots of stuff. EDIT: OK I found this is wrong. I didn't realize the little matrix of dots in the middle was the actual app list. I thought it was the dialer until I hit it. Sigh...not a big fan of that really. Two ways to launch apps. I think iPhone does it better and I don't think that's very good anymore either.
• I downloaded the Amazon app. But I never got a notification it finished downloading and I can't find it. Where is it? I go to Marketplace and go to Downloaded - and boom there it is, but its no where on my Android launch apps page. Why??? EDIT: Its in apps folder. Not in love with that.

To me the major difference between the iPhone and Nexus 1 come from a philosophical difference of user interface. iPhone uses UI elements at the expense of screen real-estate to show the user what they can do. Contextual back button, on screen menus. Android has maximum real-estate with an ever-present context-less back and menu buttons. Perfect for power users who can memorize what's there. As a sidenote, on upgraded apps, you can see what's new on the iphone. On Android I surmise you'd have to explore.

This philosophical difference is a big deal. As big a deal as physical keyboard vs soft keyboard to many. Neither is right or wrong. Best thing to do is know the difference, pick the phone you want, and let us write the best software /to/ that phone's strengths. I've come to really like Android in the form of the Nexus One. and could use it as my only phone. I prefer the iPhone's philosophy though.


Edit: For the most part I am really liking the Nexus One. Its smooth, easy to use in most areas. Don't get me wrong, the above were my impressions as a new user without any prior experience. I do not intend to say one thing is BAD or GOOD. I am simply giving my impressions. The fact is that the Nexus One is the first Android phone I could use as my primary, even though I prefer the iPhone.

For an up to date list of what I'm thinking of the Nexus One:http://twitter.com/webis_mobile

Please vote I'd greatly appreciate it. http://bestappever.c.../tmap/302503702

So 1.20 was approved by Apple in near record time (one week and during the holidays). With 1.20 hitting the streets we found a few bugs that as usual never were seen during beta. One issue is due to the fact that all localization testing is done by Chris and myself manually because we have no German beta testers.

So one major bug for German users is that we accidentally left this code in (we missed it during testing and code checks for this very issue):

newController = [[StartEndTimeChooser alloc] initWithNibName:NSLocalizedString(@"StartEndTimes",@"") bundle:nil];

Any developer will see immediately the problem. The NIB is a resource that holds the view info and its a hardcoded string. Its not supposed to be localized, yet here it is. It should be:

newController = [[StartEndTimeChooser alloc] initWithNibName:@"StartEndTimes" bundle:nil];

So what happens is you try to edit a task's dates and voila it crashes because the OS is looking for a NIB called "Start-/Endzeiten" instead of "StartEndTimes". We typically find these sort of things in our own tests as we did quite a few extensive tests in localization - but we missed this one.

We have a few other bugs reported, but so far its just a handful from the automated system. We are going to fix these top issues and release a 1.21 in the next couple days. Lets hope Apple can approve it quickly

Apple has been working in record time lately. We submitted PI to Apple a week ago before the Christmas Holidays and today its approved( (25 minutes ago as of my writing this). Of course it takes time for this to propagate through all the Apple servers so I don't expect it to show up as an update for everyone for another 24-48 hours.

To see what's new in 1.20 - here is the list:
webis.net/whatsnew.php?xml=piip120

Now everyone is going to start bombarding me with questions on Outlook sync. The PI IP side of Outlook sync for public beta is ready. The DESKTOP side is not yet ready. Its working fairly well, but we have a couple bugs we're still trying to debug and because its the holidays, not much is getting done right this moment. We are actually all on vacation this week and I will be until January 7th.

So long and short of it - please be patient. Apple surprised us by approving 1.2 a bit earlier than expected (I wasn't expecting until AFTER New Years to be quite honest) and with me being on vacation and since I'm the one that does the publishing of stuff to the servers it means I'll post it after my vacation is over. Its not vaporware, its real and it works. It just needs a bit more polish for me to be comfortable for a public beta.

A month late, we finally submitted PI IP 1.2 to Apple tonight. This doesn't mean everything is done, though. We are still working on the desktop portion of Outlook sync and also the Mac sync. Version 1.2 will bring a public beta of Desktop Sync (WDS) which I frankly hope runs so short that it goes to full release sooner than later. Then we will soon after be releasing 1.21 to provide support on the Mac side and also make known improvements on the desktop sync support (mainly better threading). I won't talk about future releases past that right now, but I'm just happy to finally get 1.20 up to Apple.

So what took so long? So a quick Outlook story. We started design on Outlook sync as soon as PI 1.0 was released. First builds of the desktop side of Outlook sync were done by August, but that's just the desktop side. You still have to write the client side within PI and at that time I was so focused on 1.10 and getting push alarms, 3.0 support, performance improvements out, and all that. Now remember that PI IP is not a one man app - we have two of us working on it and a third on the desktop sync side. But I was the PI IP sync primary guy. So I simply couldn't get to the PI IP client for it during that time. I spent a day here or there on it to get the protocol implemented, but most of the time our desktop sync guy sat waiting for me until November when PI 1.11 was finished before I could actually get to work on it.

Finally I was able to focus on Desktop Sync and look at it from not just the Outlook side, but the PI side and in the few days I started working on it, I realized that my original spec was too simple. I needed it to do support more and do more. On the desktop side stuff like supporting multiple calendars properly, more UI elements, settings, and so forth was needed. On the iPhone side I wanted to go full bore with Tags, task fields, and a lot more. This now turned into three separate projects instead of just one major one and one minor one. It affected everything in a positive way because during that time Chris was able to focus on improving many other aspects of PI other than just Desktop/Sync related items. I believe we were able to resolve nearly 100 fixes and improvements! The last month of development was consumed by Desktop Outlook sync tests and fixes. We had to go through a lot because networking in today's environment on Windows isn't easy or simple. Plus I had made a few newbie mistakes with Outlook that I should've known better from my WM days, but had forgotten (like all day appts have an extra 24 hour duration attached to them when coming from Outlook which caused all day events to be 1 day longer and shorter coming and going respectively).

As of today WDS sync is working for the majority of beta testers. There were a lot of people who asked to beta test and I had to turn most of them down due to our limited number of ad-hoc slots. WDS is not perfect yet and I expect in the time that Apple has 1.2 we'll be making a few more improvements, and then in January when WDS is in public beta we'll make more and then release it as final in later January timeframe.

So while Desktop Sync started in March, it didn't *really* begin until 1.11 was released and it encompassed 3 major projects.

---

One side effect of all the sync changes we made with 1.2 is that we had to take a step back for some people in sync. Starting with 1.2 we no longer sync colors with Google. We don't really do it with Outlook either. Why? Its a very complicated answer that basically sums up to the fact that no color sync could work as well as anyone wanted it to, and therefore no color sync was preferable to poor color sync. See, every service has issues here. Google only supports 20 or so some odd specific colors that frankly nobody liked within PI. Outlook does not provide us with Calendar colors or even category colors, but instead only the ONE color of the background appt being shown and is not really associated with anything in particular. So when we get a color FROM Google to a brand new calendar, we use it. But because 1.2 can now pick any one of millions of colors for the calendar or tag, we do not sync it back up to Google as it would error out. With Outlook we try to match the color we get from the appt to our calendar and back, but we cannot do the same with tags. Tags created by sync are orange. I could've picked any color, but I picked orange.

So what does this mean to you? It means colors are now a PI thing mostly. You set them in PI in how you want them to look in PI and the same in Outlook and Google. Furthermore, each service uses the colors differently so in reality while in PI a calendar or tag may be green, in Outlook you may want white. Outlook uses the colors for the background, Google for the text, and PI - its up to you.