Some time ago I shared a simple way to create custom UIButtons in code. As I’m starting to polish NewsWave for Mac, I wanted the same ability to easily create custom buttons. So, I ported the function to macOS, feel free to leverage:
There’s many things that have changed since social distancing began about 15 days ago: meeting friends, going for morning coffees, weekend escapes and many other activities are now on hold. However, there are also some positive consequences: my wife and I have really gotten into board games (Forbidden Island, Exploding Kittens, Scrabble…), I’ve gotten a lot better at cooking and, probably the most relevant for this blog, I have significantly more time to code on weekends, so… NewsWave 2020.1 is now live!
NewsWave 2020.1 is a major update. In a nutshell it improves the onboarding experience and builds the foundation for the upcoming macOS version.
Here are the key changes in a bit more detail:
1. Onboarding:
This was the area the app that needed more work on and I’m very happy with the result. To recap, this is how it used to work:
“when you launch NewsWave RSS for the first time, you’re asked to create an account or restore an existing one before you can interact with the app. For privacy & simplicity reasons I do this with a free IAP. Behind the scenes, when a user unlocks a free in-app-purchase (IAP), Apple provides me with a uniqueUserId I use to identify the customer. This allows me to grant the customer a 30 day free trial, allow him to try sync and easily migrate him to a Premium account if he so chooses. “
The new model is much simpler, users can create an anonymous account with the click of a button. When a new user downloads NewsWave, a new device-specific token is created for him, no emails, no IAP, no complexity.
You may have spotted a significant downside vs. the old model. As the identifier is tied to the device, and because there’s no personal information shared, if the user deletes the app the device-token is gone too. What this means in practice is that if the user re-downloads the app, he would need to create a new account.
Wow! Big downside. Well, yes, and no.
Once the user becomes a subscriber, his device token is linked to the in-app-purchase (IAP) identifier. From then on he can restore his account from any device by tapping “Restore”. This is true even if the user is no longer subscribed.
Therefore, the only time this could lead to a potential challenge is for users trying the app. Given that it’s only 30 days, and that they would need to remove the app, I’m confident the upside is a lot more significant than the potential downside.
A final consequence of this changes is that the onboarding UX is significantly more streamlined. Here’s the new page:
I believe there will be significant impact on conversion-rate, which was the key challenge of the app in the past. I’ll keep you posted on how it changes.
2. Default Feeds:
In the old model, once a user had created an account, they were taken to the “Feeds Section” and guided to start adding feeds. This was not a great first experience, and if the user tapped away from the “Feeds Section”, he would not be able to interact with anything as there was no content to show… Not great.
In the new model, when a new user is created, he is automatically subscribed to a few starting feeds. Once the initial setup is complete, he is taken to the main screen -already populated with content- where he can start using the app. A huge improvement vs the past.
This change resulted in something else, I had to choose these starting feeds.
Default feeds should:
– Touch a variety of fields/subjects.
– Be free to access (or have generous free access policies).
– Provide a good RSS summary and image to go along with the posts.
– Post enough but not too much, ideally 5 to 20 times per day.
– Represent both mass appeal and niche feeds.
Here’s what I settled on:
1. Daring Fireball (Blog – Apple)
2. Six Colors (News – Products/Apple)
4. Ars Technica (Technology)
5. Vox (General News)
6. Mr. Money Mustache (Blog – Personal Finances)
7. The Intercept (General News & Reporting)
3. Synchronization & under the hood improvements:
It’s no secret I’m working on the macOS version of NewsWave.
To prepare for its upcoming release, I’ve invested a lot of time on refining synchronization between devices. All calls now use one streamlined framework, which makes testing a lot simpler and reduces the likelihood of bugs.
Finally and most importantly, today both iOS and the macOS versions of NewsWave share the same API back end code, this dramatically simplifies the effort needed to update and improve the apps.
4. Other improvements:
There’s been many other refinements and under the hood improvements, including:
– Improved wording on several menus and notifications.
– Improved saving & loading speeds for users with large amount of feeds.
– Removed redundant code (I love this part, love it!).
– Fixed a bug that, under rare circumstances, could show a black NavBar.
– Updated Feed Recommendations.
I’ll be monitoring the servers to make sure everything is going well. With this milestone out of the way, I’ll be fully focusing on NewsWave for Mac from now on. Hopefully will be able to ship sooner than the original 2020 plan 🙂
I believe this will be a great year for RSS, if you’re not yet a user give NewsWave a try! And if you don’t like it, there’s other great options out there.
If the last post was focused on looking back, this one is all about what’s to come.
2020 will be all about focus. Only if I meet all objectives for the existing apps will I think about new ideas. Without further ado, here is the 2020 roadmap:
As mentioned in a previous post, the NewsWave onboarding experience needs work. As it stands, it’s severely limiting the number of users trying the app.
In the current version, when you launch NewsWave RSS for the first time, you’re asked to create an account or restore an existing one before you can interact with the app. For privacy & simplicity reasons I do this with a free IAP. Behind the scenes, when a user unlocks a free in-app-purchase (IAP), Apple provides me with a uniqueUserId I use to identify the customer. This allows me to grant the customer a 30 day free trial, allow him to try sync and easily migrate him to a Premium account if he so chooses.
Here’s a high level chart I drew recently to refamiliarize myself with the current account creation steps
The problem with the approach is that many users are scared, with reason, of in-app-purchases. They have been burned so many times by shady companies with deceptive free content that now all developers face the consequences.
So what’s the plan? I already did some minor changes last year, but I’m working on a complete redesign.
Here’s the idea:
a) When the user opens the app for the first app he’ll be assigned a randomID automatically. With the exception of sync, the user will be able to try all premium features for 30 days.
There’s a clear con however, if the user deletes the app and is not a Premium user (if there’s no an IAP tied to the account) there’s no way to restore their data. I’ll make sure to reinforce that, probably with an in-app message, after the user has been active in NewsWave for 30 days.
b) When the user migrates to a premium account all information will be linked to an IAP, which will also enable sync.
c) If there’s a duplicate, for instance if the user already has an IAP but just installed the app on a new iPhone, the IAP account will prevail at restore time and the temporary one will be flagged as migrated.
The fact that the users will be able to use the app from the get go opens the question: “should link them to some feeds from the start?”. I’m not keen on it. Alternatively, I’m considering having a few NewsWave articles explain how the app works. Still not sure, will decide in the next few weeks.
Excelling continues to do moderately well since its release. Every year I update it to ensure it looks and feels currents with the new iOS versions.
Its codebase is old however, the codebase is also showing its age. It’s time for a more significant internal update that ensures it’ll continue to work for years to come.
That’s the plan for 2020! Hit me on twitter if you have any questions, I’ll be sure to keep you posted on progress.
As we start 2020 I wanted to take a moment to look back into 2019.
The launch of NewsWave this past May was the indisputable highlight. Releasing a new app after months of work is always a nerve wracking experience.
Fortunately the app was very positively received, since launch the app has been updated multiple times to improve and polish it. Even though there’s much else that can be done, more on that in the upcoming “2020 roadmap” post, I’m very happy with where the app is at today.
Excelling, the first app I ever released (in 2013 no less), continues to have its stable niche audience. As every year, I updated it with UI improvements and under the hood changes to ensure it remains functional as iOS continues to move forward.
Denarius, the finance app launched in 2017, continued to struggle to find an audience. When I built Denarius banks did not offer an easy way to manage your spending in Europe. This changed drastically since then though as most banks now offer a way to track where and how you’re spending your money. This, together with the fact that Denarius was designed around the European banking system, thus limiting its success elsewhere, made me realize it was time to sunset the app.
I’m very proud of Denarius, it’s hard to retire an app you’ve invested hundreds of hours into, but its time.
RiverNews is a stand alone RSS reader for the Mac built in 2018 as a POC inspired by a post from Brent Simmons. Even thought it had no marketing it quickly gained traction and has convinced me to use it as a foundation for NewsWave for Mac (more on this soon).
And that is all! 2019 was a great year, thank’s to all my users and fellow developers that have helped.
I’ll soon post about the 2020 roadmap, couldn’t be more excited about what’s to come for NewsWave and other apps.
Nothing like vacation to accelerate personal projects, very happy with how NewsWave for Mac is shaping up. No promises on when, but I’m planning to post more about its progress very soon.
One of the items I worked on today is how icons and images will appear in the app, the intent is to show as much content as possible while making them attractive. To to that, images and icons will be scaled down, using “Aspect Fill’, and their edges will be rounded. This is a very simple thing to do in iOS, but it needs a bit more work on macOS.
I wanted to go from this:
To this:
The cleanest way I’ve found is to subclass NSImageView, overriding the image variable, and simply use that new class for the images you want. Here’s the class code, works like a charm:
In NewsWave, when you are in the “Add Feed” view you have the option to search the RSS repository, add a RSS URL or choose from ‘recommended feeds’.
Recommended feeds may change over time so, instead of using static images, NewsWave looks at the feeds from each recommended category and creates a collage to represent the group. Here’s an example from NewsWave for iOS:
For the upcoming NewsWave for Mac version however, the code needed some tweaking. Below you’ll find both the mac and iOS versions -the function takes a rect (size of the desired outcome image) and a collection of images for the collage-.
In the version below it will output 2 rows and 3 columns but that can be easily tweaked by changing ’totalNumberOfColumns’ and ’totalNumberOfRows’.
When Catalyst was announced at WWDC I was very excited about the possibilities that it created for NewsWave. Even though NewsWave launched as an iOS-only app, bringing it to the Mac has always been a priority.
A couple of months ago I started woking on porting it. As it turns out, and as you probably figured out based on the title, there were complications…
I got the app working on the Mac with little work, almost none really. However, when I started to change it to make it more Mac-like is when I started to face difficulties. It became clear this would not be a relatively easy undertaking: missing APIs, non-Mac behaviors and bugs meant I would need to invest a lot of time to, at most, end up with a barely decent Mac port. If you look at the best Catalyst ports out there you’ll see what I mean, they’re not bad but they are certainly not great either…
When I look at how I spend my time (remember I’m doing this part time), I don’t believe using it to create something decent makes sense. Here’s the Catalyst port of NewsWave running on the Mac for your reference:
So, what am I doing in the end? Well, I’m building a native Mac app! I’m considering experimenting with SwiftUI but based on what fellow developers said, I’ll likely will stick to Storyboards for now.
It’ll take a bit more than a Catalyst port but certainly will be a significantly better app for the Mac and for my users.
About a week ago I released NewsWave 2019.8, the iOS13 feature update. The objectives were to simplify the onboarding experience, add Dark Mode support and strengthen the app. Here’s a summary of what has changed and the impact it has had so far:
a) Simplifying & clarifying the onboarding experience to avoid the poor conversion rate.
This was a big deal, as mentioned in my previous post, conversion rate -users downloading the app vs. users creating an account- was roughly 20%. This meant, that about 80% of users who downloaded NewsWave never ended up using it.
To solve this issue I believe I’ll need to completely redo the onboarding experience. However, I did a couple of changes hoping to mitigate the issue. First, I renamed ’START FREE TRIAL’ with ‘CREATE ACCOUNT’ to avoid having users incorrectly assume they were getting into an auto-renewable subscription. Second, I changed the description that appears immediately below to clarify any questions the user may have. It now reads: “Create your private account and try all features for free. After 30 days you can subscribe or continue using the app in free mode.”.
Even though it’s too early to tell, after two weeks I can confirm a very significant jump on conversion rate: from 20% to 50%.
b) Adding Dark Mode
This was quite fun, it involved creating new colors in the Assets for every group of non-standard text, button or background color that should change…
It also meant changing the way colors were called in code…
And finally, creating a new dark-version of all the icons and images… One of the benefits of not having a designer is that I can do this changes myself and relatively fast. Here’s the example of the dark gray vs. light gray.
(the dark appearance color of the icon is almost the same color as the background which is why you don’t see it)
c) Under the hood & other minor changes:
On the under the hood area this involved replacing iOS13 deprecated APIs with their newer counterparts and further strengthening of back-end processes to (hopefully) fix a bug that I can’t reproduce.
As for other changes worth highlighting: improvements to the Export OPML process, improvements to how onboarding and renew windows are drawn and tweaks to some NavigationBar titles. For instance, the main timeline is now called ‘Your Timeline’.
d) Bonus: Changing App Store Keywords & Screenshot:
Something I also tried to improve is discoverability. My ranking is not stellar, if I search for ‘RSS Reader’ NewsWave appears at the 120th position, and if I search for ‘Feed Reader’ the position is 107th… So, not good.
Following advice from ’the web’ I’ve changed the app name in the store to ‘NewsWave: RSS Feed Reader’ and the subtitle is ‘Feeds, Blogs, RSS & News’. After two weeks, not measurable improvement…
I’ll continue to investigate what I can do better, discoverability is by far the key challenge I -and probably many other developers- face. Ah, almost forgot, I also added new ‘Dark Mode’ screenshots to make the app more appealing.
As October gets underway and leaves are starting to fall in the Bay Area, with the recent launch of iOS13 and the upcoming macOS Catalina release, it’s back to iOS development time.
One of the first things I learned as a part time developer is that you have limited time, prioritizing is critical to get things done. Here are the items I’m focusing on in the next few months:
1. Excelling iOS13 compatibility update (done!):
Even though it’s my oldest and simplest app it’s also the most profitable. Excelling continues to be downloaded frequently and brings regular income every month -typically between $50 to $200-. It’s starting to show its age and it would benefit from an internal makeover but for now I wanted to ensure it worked well in iOS13; this meant adding support for dark mode and fixing bugs and deprecated APIs. After roughly 4 hours of work I submitted the app to Apple and it’s already in the App Store.
2. NewsWave iOS13 feature update (wip):
NewsWave continues to gain users at a steady peace and feedback is at large very positive. There’s however room for improvement, one of the areas that need attention the most is “Onboarding”: data shows out of every 10 people that download the app only 2 end up signing up. I’m guessing this is due to two factors:
a) people are cautious of the $0 In App Purchase, they may incorrectly assume it will start charging them if they don’t unsubscribe.
b) the onboarding message does not convey enough value.
Therefore for the iOS13 feature update, in addition to dark mode support and API improvements I’m tweaking the onboarding experience to mitigate these problems. In addition this update will include many other improvements across the app: syncing, OPML export improvements, etc.
3. NewsWave for iPad / macOS (not started):
Once I ship the iOS13 feature update I’ll start working on a complete overhaul of the iPad (and potentially macOS Catalina) app. Not much to share just yet but I’ll be posting about the progress here.
I’ve loved the vacation break but at the same time I’m so happy to be back to coding on the side. Until next week,
As I do every summer, this last few weeks I’ve been deep diving into a new subject: AI & Machine Learning.
Almost every day I pack my laptop, Aurélien Géron’s book -he has done a phenomenal job-, and walk to a coffee shop close by to study.
Yes, that’s an HP laptop right there! It’s great for ML on the go.
It’s been so much fun, not only because of how refreshing this change is: Python instead of Swift, Jupyter Notebooks instead of Xcode, cafeteria instead of home office… But also because of how interesting Machine Learning has turned out to be.
Not only deep diving into Machine Learning gives you many ideas on topics you could work on, but it also allows you to understand its pitfalls (hello YouTube video recommendations or Amazon suggestions). So far I’ve been focusing on everything but neural networks but that will change starting tomorrow. Can’t wait.
On other news, I likely won’t be able to post the following two weeks as I’ll be in Australia for holidays. I’ll pack my faithful HP Spectre and Aurélien’s book to keep at it during flights and during idle times though, so I’ll have plenty to report once back.
