We speak with Ian Alexander, founder and lead developer at MoodSpace, an app making mental health exercises accessible to everyone. MoodSpace was released in 2019, and has over 150,000 downloads. After trying several databases, Ian settled on ObjectBox because of its high performance and ease of use.
Alyssa: Hi Ian, thank you so much for joining me and for using ObjectBox. Let’s start with the basics about MoodSpace and your role there.
Ian: Hi, Alyssa, thanks for having me. I’m the software developer, founder, and runner of the company – a jack of all trades. MoodSpace is an app that teaches concepts from mental health. There is a massive problem with accessibility to mental health. In the UK, for example, you have something like 1 in 4 people that have some sort of mental health problem, but only 1 in 113 go through therapy and complete it. So our essential goal is to take concepts from therapy and bring them closer to people, teaching them techniques that they can do on an ongoing basis. There’s no end date like in therapy, no waiting list, and it’s a lot easier to use it in places where you wouldn’t necessarily have access to a therapist. In the western world it is much easier to access therapy, still difficult in some ways, but much of the world doesn’t have that benefit. So that’s the goal we’re trying to reach. We started last year, we released the MoodSpace MVP in September, and now we’re going through the next stage of trying to raise our next round of funding – it’s quite exciting.
A: That’s great, congratulations! Can you tell me a bit more about your team?
I: We’re based in the UK, and in terms of the technical side, it’s just me. We also have various other roles: designer, copywriter, and another co-founder who handles much of the business side. But in terms of technical, it is just me for now. Hopefully after we get funding, we’ll be able to expand the technical team..
A: What’s your background, what did you do before MoodSpace?
I: Actually, I was originally a chemical engineer – I worked in oil & gas for a couple of years, but then I taught myself to develop and for the last 5-6 years, I’ve worked across a lot of startups, for example the dating app Once, when they were just starting up, also ITV, and then started on MoodSpace last year. Moodspace has actually existed for quite some time, it started as a hobby project of mine about 5 years ago.
A: There are other mental health apps out there, what makes MoodSpace special?
I: If you look at apps in the space, they’re generally fairly small and limited – they’ll have maybe 4 or 5 exercises. Versus the realm of therapy, which has literally thousands of exercises. Any app that exists at the moment takes just a fraction of a percent of therapy and tries to teach it. Our USP is that we are a very technically minded team, and with new technologies which have come out along with our internal processes, we can make a much bigger app, building something far bigger than what currently exists, much cheaper and far faster. Our USP is strangely, less about the app, and more about our process and the technology that we use to make the app. The tech we will be using is Kotlin Multiplatform, which is a cross platform framework which lets us maintain a single codebase which enables us to build fully native apps with full access to native APIs.
A: It sounds like the app is quite comprehensive – who is your target audience?
I: At the moment, we haven’t hit the product-market-fit stage. We’re still figuring out who the typical user is. We find that the unique art style of our app has helped our growth so far, we often find a lot of people sharing screenshots of the app on social media. So we seem to have hit a niche, but we’re still figuring out what that niche is!
A: Do you have any interaction with your users?
I: Mid-last year, I put a survey in the app, so after using it for a certain time users get the survey. There are some questions like who you are, why you are using it, and they gave us way more knowledge about who is using it and what they use it for, which was very helpful. Apart from that though, it is very difficult to know.
A: Yes, it can be very challenging, we’re familiar with that struggle at ObjectBox as well. Switching gears a bit – I’d love to hear a bit more about how and why you ObjectBox.
I: As I mentioned, MoodSpace is about 5 years old, so it’s gone through several databases. One of them was really time consuming to make – it wasn’t ORM based, so you had to write a lot of stuff yourself. Then the next was an ORM called Sugar, but it stopped being developed – it was a side project by someone, so maybe I shouldn’t have used it in the first place (laughs).
I: So then we switched to ObjectBox, and actually the reason we switched was essentially to skip asynchronous code – I’ve always been a frontend developer and what I’ve come across is that asynchronous code makes things very complex, and it means app development takes much longer. Because we had a lot of time constraints and we wanted to develop as much as quickly as possible, I actually wanted to completely skip asynchronous code – which I wouldn’t recommend – but essentially ObjectBox let us do that because it’s very fast. You’d have to have a ton of data in the app, before it would visibly slow it down – and I did a lot of testing around that and it would have needed several years of data before noticeably slowing down the app. So, that was our original reason, perhaps a bit of a strange reason. And we’ve since changed the app so it’s asynchronous, so it won’t slow down any longer, no matter how much data you add in the app. Overall, I like ObjectBox a lot – it’s just simple, very easy to use.
A: What features in your app use the database?
I: Actually everything is in the app, as we don’t have a backend. So we need it to store all the data in the app.
A: Okay, sure. Keeping everything in the app is also practical from a data privacy standpoint. How did you actually find ObjectBox?
I: It was someone I used to work with at Once – they used greenDAO and mentioned that ObjectBox (by the same people) was coming out. I looked into it a little bit and wanted to use it for a while, but it wasn’t I started developing MoodSpace again that I had a chance to.
ObjectBox is very fast, it would have needed several years of data before noticeably slowing down the app. Overall, I like ObjectBox a lot – it’s just simple, very easy to use.
A: Are there any other developer tools that you’re excited about and would want to share with the community?
I: Yes, Kotlin Multiplatform. Having been an Android developer, having used Kotlin for quite some time and having tried cross-platform tools before, I think Kotlin Multiplatform will change the way you make cross platform apps, because it lets you share so much of the code base without sacrificing the native experience. It has the potential of leading to massive cost savings in app development. Maybe in the next year or two I can see it having a huge impact on frontend development across mobile, web, and desktop.
A: What are your big picture goals for MoodSpace? Upcoming milestones? Does ObjectBox help with those at all?
I: Actually, it potentially will, with regards to ObjectBox Sync, which is part of my plan for that app. The app right now is only available on Android, and providing we get our next round of funding, we are going to be adding iOS – where we’ll need some sort of backend. We want to avoid, again, spending much money, and one of ObjectBox Sync, Realm Cloud or Firestore can help us do that – obviously as ObjectBox Sync is nearly ready, we’d want to use that. The main point around that is cost saving because it solves a lot of problems that otherwise we would have to solve ourselves – things like offline access and syncing with an API, that would otherwise be very time intensive.
A: Ian, thank you for your time and sharing more about MoodSpace and working with ObjectBox. We wish you the best of luck with your fundraising round!
Benchmarking the performance of databases is a science in and of itself and it’s hard to get reliable and comparable results. Therefore, we decided to note down some standard patterns and pitfalls when doing database benchmarks we learned about over the years. We are including specific notes and things to consider when benchmarking ObjectBox.
Designing a benchmarking test
Phrase your research question
When you want to benchmark databases, you will usually have a specific use case in mind and want to answer questions regarding that case. Therefore, before writing your first line of code, have a look at what it is you want to benchmark, why, and what statement you want to be able to make at the end. Just writing down a simple question like “Is X faster than Y doing A?” can help you verify that the finished benchmark actually measures what you want it to.
Whatever you do, document each step diligently. Benchmarks that are not properly documented are challenging to reproduce, and thus of limited worth. So, the mantra in benchmarking really is: Document, document, document. That way it is also easier to go back later and make adjustments if needed.
Select the sample
If you have a clear use case and goal in mind, this probably determines the type of databases you are going to consider for your benchmark. Generally speaking, benchmarks should compare databases of similar type and not mix approaches that are too different. A database might be designed to run on a cluster of servers and not, like ObjectBox, on a constrained device (e.g. a smartphone, Raspberry Pi or IoT device). Or data may be stored as documents (e.g. NoSQL) and not in a relational table-like structure (e.g. SQL); or on disk and not in-memory. Consistency guarantees can also make a difference (ACID-compliance vs. not transactionally safe). Document why you chose these databases for comparison.
Become an expert
Once you have decided on what databases and features to compare, familiarize yourself with the APIs of the products, e.g. by reading the documentation or looking at code examples. Making wrong assumptions about how a feature works can skew your results and make a product appear much faster or slower than it actually is. If the code is open source, it can also help to dive into the code to see how things actually work. For example: an insert function runs asynchronously so the function call returns immediately, but the actual insert still executes in the background. To correctly measure the actual time it takes to complete the insert, you need to do some additional work.
Things to watch out for when coding benchmarks
Let’s move on to some specific coding tips. A benchmark is only as good as its time measurement. Check what APIs your test platform offers to get the time; they might be affected by how and from which thread they are called. Make sure, you find an accurate way to measure the time that does not skew up your results. For example, on Android there are numerous possibilities beyond using currentTimeMillis().
If a measured code block executes so fast it is near the available precision of your clock (e.g. time is in milliseconds, but code takes microseconds), consider running it multiple times and measure the total time of all runs.
Next, before starting each measurement, make sure to free up memory and clear references no longer in use by the previous measurement or setup code. If the environment your benchmark is written for uses a garbage collector, check if it can be triggered manually to free memory (e.g. System.gc() in Java). Otherwise each consecutive measurement might be skewed due to less and less memory being available or a garbage collector halting execution to free memory. In your benchmarking results, you should look out for strange results like a continual decrease in performance from run to run.
Furthermore, take into account that some runtime environments, for example the Java Virtual Machine, do just-in-time compilation. This can cause a delay the first time code is executed, but provide better performance on subsequent executions. The effect of this on the final result can however be minimized by proper testing procedure, e.g. by running a code block multiple times instead of once and measuring the total execution time.
Then something obvious, but easily overlooked, is to ensure that between the start and end of a measurement only the functionality that you actually want to compare is executed. So avoid or turn off logging (a seemingly innocent string concatenation can skew results) and construct test data outside of a measured code block.
To be absolutely sure that your code is doing what you intended it to, use a profiler to inspect resource usage during a trial run. IDEs like Android Studio and Xcode come with an embedded profiler, and there are also several standalone profilers to choose from.
Optimizing benchmarks for meaningful results – with ObjectBox examples
Before benchmarking the chosen databases, make sure you understand their differences and default settings to adjust all settings to be comparable. In the following section, we will go through the most important differences and settings you need to look out for based on ObjectBox as an example.
Transactions, Durability, Consistency, ACID – how to make your benchmarks comparable
First, be aware of the impact the use of transactions or lack thereof can have. For databases, committing a transaction is an expensive operation as it requires waiting until the disk has safely stored the data. If possible, group multiple operations into a single transaction. For example, in the ObjectBox code snippet below, there is a notable speed-up when wrapping multiple box operations into a transaction block.
// Would be slower if run outside of a transaction.
Speaking of transactions, also check if using bulk operations is possible. These also use transactions to speed up execution. E.g. instead of performing a put on each entity in a list, put the whole list.
// Total cost: allUsers.size transactions.
box.put(user);// Implicit transaction
// Total cost: 1 transaction.
box.put(allUsers);// Implicit transaction.
The ObjectBox transactions docs provide more details and are available for Java, Swift or Go – though the basic principle is the same across languages.
Second, and closely related to transactions, are durability guarantees when writing data. This is about the “D” in the popular ACID acronym (Atomic, Consistent, Isolated and Durable). ObjectBox transactions and standard (non-async) operations are fully ACID-compliant.
Thus, pay close attention to what durability modes other tested databases guarantee, or respectively, which durability mode you want to measure. Most NoSQL databases don’t give hard durability guarantees. Some provide an extra command or special mode to enforce durability. Therefore, if your use case needs to ensure data is actually stored safely after a write operation, you would need to enable this durability for other databases when comparing to ObjectBox.
On the other hand, if you are interested in scenarios that emphasize performance over durability, you should look into the OjectBox async APIs. Those don’t come with durability guarantees unless you define “checkpoints” in your code to wait for async operations.
Indexing – how to make your database queries efficient
Third, when measuring query operations, see if you can use indexes, another typical database optimization. If the database has an index on a property that is used in a query condition, it can find matches much faster.
// Adding an index in a fictitious User class:
// Makes this query faster:
An index makes queries “scalable” – the more objects are stored, the more an index makes sense. Without an index, a database has to do a “full scan” over all potential results.
Number of objects and the disk bottleneck – how to measure the database and not the disk
And lastly, keep an eye on the number of objects you operate on. For example, if you put a single object, something like 99.99% percent will be spent on disk I/O. Thus, if you test this on several databases, the chance of getting about the same results on all databases is quite high. The limiting factor is always the disk. So if you want to measure the efficiency of converting objects into their persisted counterparts instead, you should look at much higher object counts to factor the disk out of the equation. Depending on the disk and device speed, a bulk “put” of 10K, 100K or 1 million objects will make more sense to measure in this context.
Multi threaded tests – how to set writers and readers
ObjectBox is build upon a multiversion concurrency control foundation, and thus is ready for multi threaded access. Each thread will have a consistent transactional view of the data. ObjectBox differentiates between “readers” (a thread currently in a read transaction) and “writers” (a thread currently in a write transaction). Readers never block; no matter what goes on in other threads. However, a writer will block other writers when using standard transactions. Writers are sequential; only a single writer can be run at any time. Thus, if your load is write-heavy in multiple threads, you may want to look at the asynchronous APIs of ObjectBox. These handle write operations very efficiently; no matter how many threads are involved.
Extending the database size limit – size matters
By default, ObjectBox limits the database size to 1 GB to avoid filling up the disk by accident (e.g. your code has a bug in a data insertion loop). So if you use large data sets to benchmark, we recommended increasing the maximum database size when building the box store.
.maxSizeInKByte(10*1024*1024/* 10 GB */)
Preparing the benchmarks: devices and platforms
Once your benchmark code is ready, it’s time to set up the test device(s). The first step is to ensure other apps or processes are not doing (too much) work while your measurements run. Ideally, you would use a clean device with no apps installed or services configured. This is especially true for mobile devices: once you connect them and they start charging, the operating system might wake numerous apps to perform background work or network updates. You can somewhat avoid this by switching on airplane mode. And to be safe, wait a few seconds after connecting the device.
Also ensure, this is again important for mobile devices, that your device does not enter a low power mode which reduces performance. For example on Android, keeping the screen on typically prevents that. When running on a laptop, check whether the power supply is plugged in and which power mode your operating system is set to. You may explicitly want to test in a certain power state or with the default behavior.
And just to mention it, make sure to turn off any profiling or monitoring tools that you used during building your benchmark. They can significantly skew your results.
When running on multiple devices, pay attention to the differences in hardware, like available memory and processor model, but also in software, like the operating system version. Different hardware and software might have different optimizations that can skew your results. Again, do not forget to document your device and software setup, so results can be properly interpreted and reproduced.
Running the database benchmark and collecting results
When it is time to run the benchmark it’s good to run it not once, not twice, but many times. This minimizes the impact of the various side effects we discussed above. Output the results of each run into a comma separated (CSV) or tabulator separated (TSV) format, that can easily be imported in a spreadsheet application for analysis. You can look at how the ObjectBox benchmark does it.
Once you have collected some data, verify its quality. You might have already spotted outliers while perusing the results. Alternatively, calculate the average of all measurements and see if it deviates a lot from their median value. If you are familiar with it, look for a high variance value instead. Too many outliers or a high variance might hint at side effects of your benchmark code or device setup you have not considered. Better double check to be sure.
How to publish meaningful database performance benchmarks
The final step is to share your results with the world (or just your team). Make sure it’s clear what exactly you have measured and how you have arrived at those results. So include which device and software was used, how they were configured, what you did before running the benchmark and how the benchmark was run.
If possible, share the generated raw data so others can verify that your calculations, and remember to lead to the published results. Even better, publish the source code as well, so others can run it on other devices or help you spot and fix issues.
Last not least: Be careful to draw conclusions. Rather let the data speak for itself. Respond to questions and feedback. Be honest, if you learn your benchmarks may be skewed and update them. In the end, everyone wins by getting more meaningful results.
Vivien: Hi Arne, great to talk with you today. Let’s get started by learning more about you and billiger.de.
Arne: Hi Vivien. I’ve been doing software development for more than 10 years, and API design for the last 5 years. I am currently responsible for mobile development for billiger.de, the most widely used and award-winning price comparison portal in Germany. We’re especially proud of our data security, which was just recently awarded too.
The company behind billiger.de is solute GmbH, which is based in Karlsruhe. They also have a few other brands: shopping.de, an online shopping platform for men and women, and friends communication, an online marketing agency. At billiger.de we’re about 150 employees.
Some of our stats:
active daily users on billiger.de
prices in the database
comparing 1M products
So clearly, the database and its performance on the server side is very important. Companies update their prices all the time, and on top there are all kinds of vouchers that can be applied. All of these are changing frequently – and you never know who updates their prices when. So, you can see the challenges – from a technical standpoint but also for consumers. It is hard to get the best price.
V: Tell me more about the billiger.de app – why did you decide to go for a native app?
A: Well, to be honest there was an existing native app when I came into the company. But aside from that, it’s essential for UX. We also need some offline capability for features like the notepad function or when users are in the store without an Internet connection and scan barcodes. Once they are online again, the query goes to the cloud – and the user gets his result.
V: So are most of your users on the app? Or rather web?
A: We definitely still have more web users, but user numbers are shifting to mobile more and more. Also, our web users are often one time users only. Our loyalty rate is much higher with app users, so we are trying to increase app installs. We’re seeing that – even on the web – the majority of users are coming from mobile devices. Therefore, we relaunched the website a couple of years ago to be responsive and mobile optimized. So we are focusing more and more on mobile, on both the website and through the app.
V: Why did you need to implement a local database? How is it implemented in your solution?
A: We need data persistence mainly for certain features. We’re still using SQLite, but it’s too much boilerplate code and too little fun. We have been using an ORM on top of SQLite until recently, but it didn’t work well in combination with Proguard on some Android versions anymore. So it resulted in lost data. We’re currently using ObjectBox in the billiger.de Pro version and in a fun new project called PricePretzel, which gives users the best price actively and tracks savings. In these projects, ObjectBox has proven its worth, so we want to migrate the billiger.de app too.
V: Yes, SQLite with an ORM can get very messy. So, why did you choose ObjectBox as the alternative?
A: I looked at several SQLite alternatives and ObjectBox looked interesting. The main decision factors were: ease of integration, stability, and performance. But ease of use and integration were really the most important factors. Stability and enough performance were rather basic necessities. We found ObjectBox really easy to use – we did the migration and everything and because ObjectBox handles that automatically, it was really simple.
We found ObjectBox really easy to use – we did the migration and everything and because ObjectBox handles that automatically, it was really simple.
V: So did performance matter to you at all?
A: For our needs, performance was secondary. Obviously the performance needs to be good enough, but we do not have super high requirements regarding performance.
A: Synchronization obviously is a super interesting feature and we are keeping an eye on it once it is publicly released. From the setup we have, we would need to do it with a connector to our existing database. Currently the web data and app data are separated and we are working on integrating them. So, this needs synchronization.
V: Which other tools do you use in your solution/are you excited about?
A: Retrofit from Square, a networking library, we recommend it to everyone and it works super well with ObjectBox. Both libraries work well together with our business objects. Retrofit fetches the fresh data from our servers and deserializes it into our business objects, which are then persisted with ObjectBox without any additional boilerplate code.
V: billiger.de has over 500.000 downloads and about 4 stars on average – how many daily users does the billiger.de app have? Do you have peak times?
A: Obviously holidays like Christmas and Easter are busier. During the day, early evenings get the most traffic – about 1000-2000 daily active users in the billiger.de app, 200 in our Pro-app, and iOS is similar. As I shared before, we get about 300k daily users on the website.
V: Thanks for sharing, and for talking with me today. Any last words?
A: Thank you for having me! I am looking forward to do more with ObjectBox and am very excited about what comes next!
Sideloading can cause crashes when used with Android App Bundle. Google is pushing Android developers to publish their apps to Google Play using the new Android App Bundle format. While it comes with benefits, it can also cause your app to crash, if users sideload your app. We’ll explore why this happens and how to fix it.
Android App Bundle (AAB) enables smaller app downloads and updates, increasing the chances that more people will install your app. Developers benefit as well by only having to build a single artifact instead of multiple APKs tailored to various types of devices.
App Bundle and Sideloading
The Play Store will take care of installing the right set of split APKs from the App Bundle for each user’s device configuration. But what happens, if users bypass Google Play and sideload the app? Sideloading has been popular (APKMirror anyone?) to get early access to new versions of an app or to not waste expensive data volume on app downloads (relevant e.g. in various parts of Africa and India).
To explain how sideloading can break your app, let’s first have a look at what Google Play would actually install onto a device when you publish an App Bundle. For this we use bundletool, which Android Studio and Google Play also use, to convert an App Bundle to a set of APKs for a specific device. We’ll build an App Bundle for our ObjectBox Kotlin example app and start an emulator running Android Pie (you can connect a real device as well). Then we can run the command:
If we extract the created connected.apks file (e.g. using unzip or 7-Zip) we find three split APKs. A master APK containing the app’s manifest and all of its code. An x86 APK containing the ObjectBox native library. And a xxhdpi APK with some resources specific for the screen density of the emulator device. Depending on your app (if it uses dynamic feature modules or includes translations) and connected device there might be more or less split APKs. But let’s stick with these three.
Out of the three only the master APK can be installed on its own. The others will fail to install. You can try this yourself by dropping the APKs onto an emulator.
Crashing with LinkageError
Now how does this affect sideloading? You can probably already see the issue: due to lack of awareness that you are using the cool new App Bundle format, users only share the master APK. And as mentioned it installs just fine. However, at runtime your app might access a native library like ObjectBox or some density-specific resources. As these are not part of the master APK your app will crash. Bummer.
Et voilà, instead of your app crashing users will see a dialog asking them to reinstall the app from the Google Play Store (or “an official store” if the library can’t detect it).
How it works
So how does Play Core know if parts of your app are missing? The library is obfuscated – as most Play libraries are. However, decompiling (just open the class files in Android Studio) allows some basic analysis.
At first, it verifies that the device is API level 21 or higher. Older versions of Android do not support split APKs, so no need to check on these devices.
Then it makes sure that the currently installed version of your app actually requires APK splits. After all, you might still distribute your app as a full blown APK to some devices. This check just looks for a special meta-data manifest tag. This tag is added by bundletool (read Play Store) when converting an App Bundle to a set of split APKs (drop the master APK onto Android Studio and check AndroidManifest.xml).
Now that it is certain your app requires multiple APKs for installation it asks PackageManager for the PackageInfo of your app. Starting with API level 21 this has a splitNames property which has the names of any installed split APKs for your app package. If it is empty one or more APKs were not installed and the user is warned to reinstall the app. Straightforward.
Curiously it also warns you, if there is only one entry with an empty name. Please let us know in the comments why you think that is.
And that is it. Make sure to add the Play Core detection if your app is using ObjectBox and App Bundle to avoid sideloading crashes and keep your users happy.