I finished up the latest release for both Android and iOS and QA accepted it as ready to release on Friday. I submitted it to the Android Market where it appeared nearly instantly. To my surprise it appeared in the Apple App Store on Sunday - I got an email from Apple on its updated status. I did not think they actually worked over the weekend. This is the second time it appeared two calendar days after submission.
Based on activity they just gave it the illegal method check once over and let it go to the store as it is an update to an existing application that requires a login to use. It is only useful to our clients. It is not like we are selling it and forcing you to buy a login outside of the App Store so I don't think Apple gives a rip about what it does, if it leaks memory or how clients use it. I don't they even care if it follows their UI guidelines. As far as I know it is not doing anything out of the norm. I triple make sure I am not using any private API calls, run it through the leak detector as I add features and really the app is used to access our servers and get to the data the client owns and to send updates to the data as needed. I am very happy it went through the process quickly.
Now that everything is out of beta we can turn on access at our server level and let the doctors have some real fun with the application. I am really curious to see how many installs we have for each platform and how much data traffic it generates. Doctors can take pictures of the medical paperwork on site so we should start to see images flying about soon enough. Well I won't see the images, I don't look at live medical records, but I can see file sizes and what not from the log files.
We will await feedback from the clients before we plan the next release. I have some items I would love to add to the program but if no user requests them it is better to work on the real needs and not my guesses. This means I am on break from the mobile side of things for a bit. Good thing to happen just before the holiday season.
I will be working on our front office appointment scheduling program. It makes heavy use of MigCalendar. I have messed around with that a bit but am a total novice. The original developer is not here so I will have to figure out most of it on my own. MigCalendar is a bit of an odd bird. Things are stored in statics that I don't feel belong there. The API is very powerful but not easy to understand. I started work on this a few months back and made a lot of notes so I am reviewing those and diving into the code. Various object constructors have numerous parameters which can get hard to follow.
Our current UI is lacking in usability. There is a tabbed interface but the tabs are really just controlling what is visible in the calendar. There are two buttons inside each tab that control the filtering operations. I am converting these into a combobox with 4 items mirroring the tab / button combinations listed above. This gives more vertical room in the UI and gets rid of a tab that is not really a tab and buttons that toggle but don't. It allows me to add more root + filter modes in the future and cleans up the code nomenclature as the buttons were namee after their functionality in the first tab which has no relationship to what they do in the second tab. It also leaves one instance of MigCalendar running instead of two. Yes, this whole application was in dire need of help.
Can't say I am looking forward to getting this all up and running again. One of those projects you just want to scrape and start from scratch. I feel like it should appear on "Extreme Code Makeover" and we should blow up a computer with C4 while some dude with OCD runs around screaming. Once I get over the initial learning curve it should not be too bad. Of course once it starts working like it should more requests will come in and I will be tagged as "the guy" that knows the appointment stuff. Since I am taking off the last week of the year I also get a chance to forget most of what I figure out this week. Nothing like starting over twice in a few weeks period.
No comments:
Post a Comment