Common Voice Android (unofficial app)

Hi everyone. I read some threads where you explain the reason why you don’t want to create an official android app for Common Voice.
I’m part of Mozilla Italia and many users ask if exist an android app to contribute.
For this reason, I decided to develop an android app (unofficial) for Common Voice project. It’s open-source and it’s on Github (link at the end of the thread).
I developed this app with the support of Italian community of Mozilla: Mozilla Italia, and especially thanks to Damiano Gualandri and Simone Massaro, who tested it and helped me. The app is developed in Kotlin, via Android Studio.

The idea of the app is to improve the experience also on mobile, because the website is very slow (on mobile) and there are too muc-h effects. In the app, instead, you can record/validate just with few clicks.

The app, now, is a “Alpha version” and the “Speak” section doesn’t send the audio to the server (I’ve some problems, maybe someone else can help me to understand how to send an audio file to cv server, as post request)
Now it shows the UI in English and in Italian (based on device language), but you can set the language for Record/Validate (in Settings) -> now supported English, Italian, French, German and Spanish.

You can find 3 big sections: Home, Dashboard and Settings.
In Home you can go to Speak ans Listen sections. You can also do the log-in/sign-up, and see information of your profile.
In Dashboard you can find Statistics (everyone’s statistics and, if you are logged-in, yours) and also the Voices online.
In Settings you can change your native language (which is used to show sentences and clips), link to go to Github repo, Paypal donation and credits.

I hope you’ll like this app, these are some useful links:

To send a feedback or report a bug, please open an issue on github instead of review negatively the app.

3 Likes

I don’t want to dismiss your work, but I’m really not sure this is the best use of your time. Please keep in mind the Common Voice team is very short in headcount.

Better fix the mobile experience directly. Honestly, I know a lot of people contributing with both desktop and mobile, and nobody ever reported any issue like you mention of the website being slow.

Unsure what you mean there.

That’s already the goal of the website. And, again, all the feedback I got was that this was the case.

There’s no support for an API exposed by consumer like your app. So it is :

  • not suprising
  • fragile and prone to breakage in the future

I’m really not sure that putting additionnal pressure on the Common Voice small team this way is effective: this is going to create confusion for everybody.

Hey there.
I just want to share my opinion on it.
Basically, in my honest opinion, the website it is a little bit clumsy and heavy and personally I think the app is better than the webapp because it is a dedicated thing, and I do not need to type the URL everytime in the bar in a browser.
Furthermore, it helps and remember me focusing and to partecipate on common voice because I see the icon.
The website is full of graphical effects (not needed) that personally I think they are useless.
In the end: the application is more oriented to help and focus Common Voice.

1 Like

I’m so sorry you don’t like the idea.
Also on Discourse some people wanted to an app (e.g. Android app?, Native app on Android and iphone, …).
I know Mozilla couldn’t spend time on it, because cv website works on android, but a native app could give a better experience.

Believe me that I want only improve the project.

In the website, now, there are so many animations and it’s very slow (especially on “older” devices).

I don’t want to put additional pressure to CV team, in fact I developed the app and I didn’t ask for that :smile:

Can you be more explicit ? It’s really not heavy here, and not on high-end hardware.

I really don’t see what “graphical effects” you keep referring to.

Nothing would stop you from adding a link on your homescreen to the website, if it’s just what it takes.

You make it sound like “works on android” is a not really important by-product, while it’s an important goal.

What does the app does that the website cannot do ?

I don’t see all those animations. Why don’t you file issues on the website to improve the experience on those “older” devices ? What devices are we talking about ?

But then you are asking for non-supported use-case of the website, effectively putting pressure.

I’m really sad to see contributor efforts and time being thrown away like that, when there is so much to improve already on the website.

I don’t have a prehistorical phone, however when I am not at home the website is just slow. It lags when I send recordings (while I validate sentences).
The graphical effects are how the menu opens and things like that. I simply don’t like them.

Of course, I can add a link but this would open my browser. And to spread better the idea of Common Voice I think an app in 2020 is a mandatory. Person is more inclined to download and use a light app than opening web browser put the address and so on. It’s a quicker way to access Common Voice.

Well, that’s still not very clear on your target device.

Ok, that’s still very vague, it could be from a number of reasons. When you say it’s slow when validating sentences, that could just be at network level.

Maybe those are utterly slow on your device for other reasons ? What browser are you testing in ? Maybe it could be improved by disabling the animations in some cases ?

I’m not a big fan of animations, but I don’t really care about those.

That’s a very strong opinion, that I don’t share at all. The web has everything needed to provide first-class experience here: getUserMedia, Notifications, etc.

Again, that’s what homescreen-shortcuts are designed for.

1 Like

I don’t share your opinion, contribute from an app or from the website should be the same, but the first one is thought exactly for android.

Personally I have some problems with mobile version (I’ve already said): too animations, speak section usually doesn’t work, listen section reproduce audio wrong (not by main output, such as music, but by “voice-call” output), and many other issue. I tried to report an issue on github, but it’s not solved yet, so i didn’t reported others bugs.

I hope other people would have other opinion about this idea of app.

Likely because the team is small and has a lot to do.

You should.

Can you elaborate ? That does not feel right.

I don’t get your point here. Are you talking about the Android-level audio channel used by the browser ? As much as I recall, website uses a <media> element, this should be routed through Media and not Phone.

Which ones ?

Exactly, the CV is small. I appreciate everything that Mozilla does, and in fact I became a Mozilla Italia volunteer, to help.

The speak section doesn’t work on mobile: I press on icon, it looks record audio, but the recording is empty.

The listen output is wrong: It’s reproduced by channel “call-speaker-capsule” (i don’t know what is the specific word) insteal of “main-speaker”

In the home page when I tap on “Speak” (or Listen) it shows more details, as I “hover” instead of “click”

The spinner is out of the container (in Dashboard)

In the same section (Dashboard - Top contributors) when i tap on the “eye” icon:

An example of bug in Speak section

(where is the icon ?)

An other bug

“Tap then …” it should show an icon after Tap and before then

Other bug in Dashboard

And there also other bugs

I’m sorry, but it’s not enough to diagnose anything.

Ok, but what differences does it makes in the end ?

I’m not sure to understand how this is a bug. You click, you get to speak directly, right ?

That looks like bogus layout because of the resolution and locale-specific. What’s your browser and device, again ?

Same.

I don’t know.

I know there have been issues in the past weeks, server-side, because of much increased trafic after promotion campaign. That could just be it, some overload server-side and resources having a hard time getting downloaded.

I don’t see anything unfixable. I also see a lot of reports that could be locale or device-specific, and hence harder to reproduce to fix or not top-priority.

Any help is welcome, but I am really not convinced that building an Android app is the best way.

That looks like bogus layout because of the resolution and locale-specific. What’s your browser and device, again ?

I use Firefox (of course!). It’s happen on Firefox Stable and Firefox Preview too.
My device is LG G6.

I’m not sure to understand how this is a bug. You click, you get to speak directly, right ?

If I press on “speak” icon (on mobile it’s small) it goes to speak section, otherwise, it shows those details.

Ok, but what differences does it makes in the end ?

The audio quality is not the same, and also the volume. It’s absolutely awful.

But I appreciate, again, the work done by CV team. And I hope the project will be fine, and in my opinion (the contrary of yours) an Android app could improve the project.

I’ll update it, so CV team won’t have other problems to maintain the app.

So it’s far from being a slow device, and has a decent screen.

“otherwise” in my testing is “I long-press on Speak”, and long is > 3 secs. Also, the icon is not “small”. How much small is it for you ?

Is there a bug file for this one ?

The team has had an iOS app because of Safari limitations in the past. It is trying to move everything to use only the web version, so I really doubt.

Your first post already shows that it is not the case, you need help for posting content, since there is no app-facing API that is being offered.

Of all the issues you mentionned, I still don’t see anything unfixable. Again, you might already have filed the issues and it went unfixed for now so it feels frustrating, because there is so much work already, and not enough people.

@Sav22999 Let me say it otherwise: the project could really benefit from external contributions and there are tons of things to improve. So seeing people actually investing time is really cheering and a good thing. Seeing that this valuable time is invested in a solution that brings much more problem than it solves makes me really sad.

Also, have you reproduced that on Nightly or using Chrome ? I’ve not noticed any issue on Nightly, it could also be Firefox-stable issues.

I tried on Chrome and there works “Speak”, but doesn’t work “Listen” (bug above reported).

I guess we have 2 different opinion :smile: in my opinion an app, as I said, is important.
E.g. There is a mobile app of every social network, but they have also a mobile version if you go to the website … why? because an app is more efficient of a website (this is my opinion, course!)

I’m trying to do the post request of the audio, i’m testing on test web site (azillom) so the main server doesn’t have problem. I’ve just invited others to contribute :smile: because this is an open source project (on github) so everyone can contribute.

My smartphone, in fact, goes very well, but on cv website it’s not really fast (the website, not the device).

I thought this app would be a great possibility for CV, I’m so sorry you don’t think so.

And by bug, you mean the behavior that on “hover” we display some informations ? I’m sorry, I still don’t understand (even re-reading your message above) why that qualifies as a bug, it does looks like intendend and sane behavior.

So far, all your arguments more points to visibility on your system, which is solved with a homescreen shortcuts, or issues that seems to not be reproductible trivially. There could be some legit cases where an app would be more justified, i.e. missing Web APIs for a feature. But that’s not the case for Common Voice.

I can find you several examples of website that are more efficients than apps.

That’s not really valid, test infra can have outages as well, It’s even more likely that it’s more often not working.

Yes, that’s what I was verifying. And this is obviously not the experience I can reproduce on other devices, including old Sony Xperia Z3 Compact.

I’m just very sad that your solution for fixing the website is to make an app: again, Common Voice team is a small one, and they already have a lot to deal with. Pushing and sharing an app, even with the best intentions, is just going to add more work for them, and risk degrading the experience overall.

Can I ask, without being judgmental, why you did not contribute by fixing the website instead ?

So far I don’t see anything really justifiying. CV could benefit, obviously, from some contributions to improve the website, and you are obviously free to pursue your own codebase.

I’d hate to see you wasting efforts and time on something that won’t really help the project and that will require extra maintenance from you.

1 Like

@Sav22999 Checking on the Github repo, there are PRs including one adding a web-based Android app: https://github.com/mozilla/voice-web/pulls?q=is%3Apr+is%3Aopen+label%3A"help+wanted" wouldn’t it be a good solution, solving your need for Android-level integration, and still relying on the website to ensure benefitting the project the best way ?

(extra bonus if you implement using GeckoView :D)

1 Like

:+1: i’ll see.
I can give you some examples of app that are not necessary, but they improve tje experience: Outlook, Gmail (email client), also Facebook, Twitter, ect.
Anyway, I prefer a native app, so I think an app is useful and not the useless :smile: