What do you think about the fluidity of react-native apps ?
So far, I've seen apps made by Chinese e-commerce giants like Alibaba's Taobao and Jindong, their apps requires lots of resources and are all very slow even on last end Android devices.
I've also seen some startup using it for chat app and it's simply become unusable after a while because all the memory leaks.
Do you think these are just poorly made app or react-native is just not ready to replace native development ?
Which app would you recommend to check out in order to see how powerful this framework can be ?
The only case I've seen that can be tricky to keep performant at the moment is when you're doing CPU heavy tasks whilst listing out uncompressed images >2mb. Memory consumption of images this size do seem to be a tricky problem, though libs like react-native-fast-image aim to solve this sort of thing.
I'd say almost certainly they were developed badly, there's no reason now why you couldn't make apps indistinguishable from native. My top tip for this is to move away from JS based navigation libraries (e.g. react-navigation), that's where you see a lot of the performance gains.
Performance wise the Patient View app I made shows this off, I sync and encrypt thousands of records and chart them out whilst keeping everything 60fps.
I see, thanks for the answer and cheers for the 100 apps, that's a hell of a pace :)
What do you think about start app using expo? Is better start app with react-native Cli?
What package do you recommend for UX?
No, I would never use Expo for a serious project. I imagine what ends up happening in most projects is they reach a point where they have to ultimately eject the app (e.g. needing a native module) which sounds very painful.
The majority of the apps I've wrote have definitely reached that point, for the individual benefits expo provide (e.g. code push) I've considered trying out using just expo kit but I'd rather just uses a tool specifically for that purpose. Above all, I always like to reduce the level of magic in any application, mobile or not.
Regarding "What package do you recommend for UX", I guess it really depends what you mean. But hopefully this gives you everything you need to know:
I'm very grateful for your answer, this is very helpful !! 😊
I use always expo because I'm starting programming in RN, but I think I will start starting project with react-native-cli.
This is a really good question, in a meeting at the moment but will definitely respond to this!
I'm working on my second app now, but this will be the first one I release. My biggest fear going in to this is not building that app, but the red tape of getting it published. With that being said, my questions are.
Good question and congrats :).
How hard much harder is it, to publish both Android and IOS, vs just IOS?
To get the last question out of the way, I'd say if you face any red tape it's almost certainly going to be on iOS side. Also the time bottleneck is going to be iOS side too, publishing can take 1-2 days but sometimes more.
What should I be prepared for when publishing my app?
I can't give you a simple one sized fits all answer here, but I would say that as long as you're not doing something outrageous you'll probably be ok. Best thing I can do is list some of the common scenarios I've found that might catch you out:
With the above being said it literally depends sometimes on who views it.
What can I do to help avoid headaches during the process?
You rock and that was greatly informative. Thank you so much.
That's alright, enjoying all these questions!
Sounds fantastik! This is your glory day, horray!
Each 10th day – new app!
I would like to give you 100 hundred copy-pasted emoji beer.
What is percentage of bridging to native modules?
Low, less than 10% maybe, but I've worked on some weird projects. There's so many repos now on GitHub when comparing to 2015.
Here are the times I've had to delve into native modules.
Was anything except Keyboard Extension on iOS? For example connection ABAddressBook written on ObjectiveC.
Interesting your integration approach with CoreML for example.
All the above required me to write ObjC. The HUD app in particular required me to bridge a pretty beefy SDK that was pretty closed off to the public.
I haven't integrated with core ml, I do want to play around with using tensorflow and React Native at some point.
What is the major benefit over React (web)? Of course for apps where you need access to the users music for a music player it makes sense, but for festivals for example a web app app should be sufficient.
It ended up being an amazing app store SEO hack, a lot of the time we were the top search result for x festival - that's a lot harder to achieve on a website.
Also, quite simply, people take their phones to a festival not their laptop. This worked offline and tapped into things that only fit well on an app (e.g. find my tent, chat with image upload). It was definitely a lot easier to compete against the official app than website.
Ok, I'll keep in mind... Thanks...
No problem! :)
Hi I'm diving into an already existing react native project, which hold a ± of 100.000 users. For now it's quite static and I was asked to add more delight and interaction to the app. On my research for this I stumbled onto the Lottie library, integrated into react-native. I tested it on my virtual machine and it looks amazing. But I have some questions though:
Is it stable to use in real life products?
Are there do's and don'ts for the use of this library? (like 'Use it for buttons, but definitely don't use it for...)
When using this in live projects, did you often have performance issues?
If you find the time I'd love it if you could enlighten me on this. :D
Have a nice day!
What are the most 3 troublesome issues with React Native and how did you overcome?
1 - Upgrading projects from very early versions of React Native. This has been, at times, an absolute nightmare (mainly iOS)! Especially when your project relies on several libs that require native changes. In the end, it was often easiest to create a new react native project from scratch and manually link libraries one by one.
Thankfully, this is rarely an issue now as it has matured. Having said that, I think it's really important to stay on top of React Native versions if your project has longevity.
2 - I had to integrate a react native app to read sensors for a car HUD, no simple way of overcoming this I just had to learn Objective C to write a React Native Bridge. Getting comfortable doing this is really valuable if you run into projects that require specific native functionality.
3 - Once, I hit an error which I could only replicate at the point of uploading a react native app, it would error out with "You must supply a CFBundleIdentifier for this request". This had me banging my head against my desk for a few days. In the end, it was something simple like the way I was creating my release scheme in xcode.
How many of them do you actively maintain?
Can you give a few examples of some of them?
Only a handful, which is a bit annoying because my xcode organiser / itunes connect etc ends up looking like this :D. prntscr.com/klu0jm
A lot of them have been released for a temporary period, due to them being PoC's or only having a limited shelf life (e.g. my festival apps). Other's just get taken over by other people post MVP.
A few that I'm maintaining, I'll add to this list in the coming weeks too if there's appetite :)
How do you handle Keyboard hiding inputs ?
It's a bloody pain!
I think I've got a post somewhere in GitHub saying this is the biggest pain point at the moment with core React Native, and I posted that in 2017 github.com/facebook/react-native/i....
I just use a combination of the keyboard avoiding view and react-native-keyboard-aware-scroll-view, I've always been able to achieve what I want but it's always fiddly.
Also do you have any tutorials you could point out to building react native bridges and projects which android native requires libraries written in C
I don't, and to be honest I've always been out of my comfort zone doing it. Ironically, as a result, I'm more comfortable reading ObjC than Java as I've bridged more in iOS even though I spent the first 2 years professionally writing Java.
I'd love (if I had the time) to progress more in native development, I think it'd be the thing that'd help me become a better React Native developer. Realistically you can achieve absolutely anything so long as you have the Native Code to back it up.
How we can get a better debugging experience? When I was trying to develop some React Native apps and has an error, the stack trace were so confusing. There is a way to have a great debugging session with breakpoints, watches... and don't need to keep changing our code until it works?
I guess it depends where the error sits, some errors are hard to debug yes , either errors in the render markup itself or in native modules.
However I would say apart from that it's pretty much like debugging any web app in chrome with watches and other stuff you mention. I'd even go as far as to say developing with hot reloading is better than the react/webpack equivalent.
What's your experience (if any) been like with RN-web?
We're considering rewriting an existing hybrid app to RN w/web, uses camera but no other hardware. I read above that you wouldn't recommend create-react-natve-app? For first timers, would you recommend using that, then when comfortable rewriting instead of ejecting?
I haven't used it. I love the idea of sharing web code with react native, but I don't think that idea should go as far as to share dumb components so I probably wouldn't use it. I haven't gone through the process of ejecting an app but I've been told it's really painful, but it's probably not quite as bad as rewriting from scratch.
Sharing code with web was what got me into react native in the first place, I don't think that benefit is talked about enough.
Hi Kyle well done. so many projects. I'm worried about the performance for huge sets of data. How did you handle that? I would like to know your thoughts.
The performance of what with huge datasets though? I guess the main thing you'll want to be worried about when dealing with large amounts of data is how you're rendering it.
In which case, I'd say as long as you're virtualizing the rendered data (i.e. using FlatList or something similar) you'll be fine.
That being said, if you're worried, write a mock test against said data!
Hi Kyle, you were right. Just using some components and applying best practices in general we can have a good performance.
Thanks a lot.
Where do you think react-native is headed in the coming years? Do you n think it's evolving at the same pace as the native counterparts?
The biggest changes I've seen with React Native is the introduction of bigger native bridges. Proper native modules like react-native-navigation and react-native-charts-wrapper are basically making it so apps can be made to be completely indistinguishable from their native counterparts. So hopefully, assuming the community stays massive it means more of that.
I can see it easily becoming the de facto tech choice for startups. Having said that I think we'll also see React Native used more and more for parts of existing native apps that (e.g. some screens) as an attempt to gain cross platform sharing without rewriting whole apps.
Hi kyle, some help here dev.to/elaziziyoussouf/react-nativ...
Will take a look
What are most used react-native plugins in your apps?
What do you use for mobile backend ?
I've gone from having a big boilerplate with having everything I could want (e.g. camera support, contact lookup) to stripping back to what I'd use 90% of the time. It's easy enough to add modules when you want them now with react-native link.
This is what I'd consider core:
The rest is down to if I'd need it, charts, contacts etc generally all have libs nowadays.
By mobile backend do you mean what server? That totally depends on what the project is, maybe sometimes I (or maybe even sometimes the company I work for) wouldn't even have a choice on this.
If it were me developing on my own though it'd be node.
For someone with a good background in Android Native development, does it worth the hassle to learn React Native or just learn Native IOS development?
React Native's benefits come into play when you start sharing across platforms (including sharing react native codebases with web), and being able to develop on a codebase where web developers could easily get onboard. That being said, I guess it comes down to how you'd find the learning curve of each. If you're not savvy with JS and don't want to reach the point of having a shared codebase with Android, maybe go with Native.
A side note on this, I think there's a real untapped opportunity for native developers to work with React Native developers to build out the box native components that perhaps they use as a toolbelt to make cross platform apps way more efficiently but perhaps offer a bit more fidelity than most React Native Apps.
How do you cost the application?
We started out costing an iOS build and then adding like 20-30 % at the end for an Android build. However now because of the parity in React Native they can now be costed as 1 build. It's not hard anymore to keep things consistent between platforms IMO.
There might be flat costs for things like getting iTunes to connect and play store builds up and running and maybe doing a sweep through on Android and iOS to ensure consistency.
Apart from that, pretty much no different to how I cost for the web. This might be a bit out of scope for the AMA but here's the high level:
Depending on the project might start with very quick PoC to test things are feasible or further research into how the industry works and if we are likely to run into any issues that could affect cost.
Add high-level estimates for "epics" based with cost being time estimated, contingency assessment on some of the more unknowns which might add a 10-20%.
With the help of the client, break those stories down into more detail and organise into sprints of work, then work/charge on an agile basis from there.
That's a crazy work rate! Well done, man! Do you see yourself producing a similar amount over the next 3 years or will you be aiming to slow down at all? ✌🏻
There's always more side projects! :). The company I work for has definitely focused more towards making mobile apps so there will be plenty more to come!
Why? Are they all useful?
They range quite drastically.
A lot of them were festival apps, they totalled around 200k downloads, not the most useful app but had a large, temporary influx of users. Around 5 of them achieved top 10 in the iOS music category.
On the other hand, one of the more useful ones of mine is Patient View, this probably the proudest points as a developer as I could literally see it helping people with people's illnesses on their day to day life
I've developed quite a few apps for startups too, definitely helps out people with tight budgets who can't afford two native developers and a website developer.
How do you maintain reusable styling throughout the application? Do you think styled-components can solve this problem?
I was about to write a long post about how I'm unable to allow a part...
How well do you know your own code?
We’re a place where coders share, stay up-to-date and grow their careers.
We strive for transparency and don't collect excess data.