What simple Android app are you missing?
October 2, 2018 9:12 PM Subscribe
For a friend. What would be a good first/second android app for a beginner to try and tackle?
The cashier at the gas station across the street is a young-ish (comparatively) person dipping feet into programming. He's found out that I'm a programmer of sorts and asks bunches of questions that I have no good answers for. (use your imagination).
He asked me for "a simple android app he could make that doesn't exist", and gave an example of an app that tracks your mileage and notifies you when it's time to change the oil in your car. I shot that down like nobody's business over "it's been done by car maintenance apps, or a calendar or etc." And just nailed him on how to determine if somebody is in their car and driving or not and just how do you know how many miles they've driven. I'm bad at this.
So... What's a simple android app that you are missing that might be a good project
(He's evidently already made an app to support chat on 'his website'). He seems against Cloud Computing sorts of things.
So, is there something simple-ish you want that isn't (like an egg timer or such) that I could tell this person to try and build?
You might get a free android app out of it... 'cause he doesn't know about ads an cloud yet. (muahahaha).
The cashier at the gas station across the street is a young-ish (comparatively) person dipping feet into programming. He's found out that I'm a programmer of sorts and asks bunches of questions that I have no good answers for. (use your imagination).
He asked me for "a simple android app he could make that doesn't exist", and gave an example of an app that tracks your mileage and notifies you when it's time to change the oil in your car. I shot that down like nobody's business over "it's been done by car maintenance apps, or a calendar or etc." And just nailed him on how to determine if somebody is in their car and driving or not and just how do you know how many miles they've driven. I'm bad at this.
So... What's a simple android app that you are missing that might be a good project
(He's evidently already made an app to support chat on 'his website'). He seems against Cloud Computing sorts of things.
So, is there something simple-ish you want that isn't (like an egg timer or such) that I could tell this person to try and build?
You might get a free android app out of it... 'cause he doesn't know about ads an cloud yet. (muahahaha).
I would like an app that you can tell it what you're going to do and it will remind you to do things associated with that thing.
For instance, I'm going to the grocery store: I hit the grocery store button and it shows me the list I've been adding to of what I need at the grocery store.
I'm going to bed, I click the bed icon and it reminds me to take the dog food out of the freezer.
I'm going to my friend's house and it reminds me to bring the books I borrowed from her.
I'm going to the bookstore and it shows me a list of books I've added to my want list.
I'm going to the doctor and I hit the button and it shows me questions I want to ask him.
Etc.
I currently use a gazillion apps (habit tracker, grocery list app, calendar, to-do list, note keeping app) to do the above, when, in my mind, they all require the same solution: remind me of something when I perform an action, usually irregularly.
posted by dobbs at 9:30 PM on October 2, 2018 [2 favorites]
For instance, I'm going to the grocery store: I hit the grocery store button and it shows me the list I've been adding to of what I need at the grocery store.
I'm going to bed, I click the bed icon and it reminds me to take the dog food out of the freezer.
I'm going to my friend's house and it reminds me to bring the books I borrowed from her.
I'm going to the bookstore and it shows me a list of books I've added to my want list.
I'm going to the doctor and I hit the button and it shows me questions I want to ask him.
Etc.
I currently use a gazillion apps (habit tracker, grocery list app, calendar, to-do list, note keeping app) to do the above, when, in my mind, they all require the same solution: remind me of something when I perform an action, usually irregularly.
posted by dobbs at 9:30 PM on October 2, 2018 [2 favorites]
Best answer: This might be a bit on the advanced side for a first project, but an app that can OCR a grocery store receipt and turn it into an inventory tracker that I can use to keep track of groceries I buy would be cool.
posted by Aleyn at 9:32 PM on October 2, 2018 [1 favorite]
posted by Aleyn at 9:32 PM on October 2, 2018 [1 favorite]
The cashier at the gas station across the street is a young-ish (comparatively) person dipping feet into programming.
Someone in this position is not in any kind of shape to actually support a production application that gets any kind of serious traction. I think if you really want to be supportive, the best thing you can recommend is that he build something he wants without even thinking about a market. The *next* app, maybe he can think about marketing it. Real useful apps have a completely different way of being constructed than things that are actually useful for learning purposes, among which that they start not with "what do you know about programming and what comes next to learn" but "what does the user need".
Variants on to-do lists are popular for this kind of thing for a reason. Don't make him think that he's going to actually take his first foray into producing something and turn it into a thing that's really ready for prime-time. It's just a route to really bad disappointment.
posted by Sequence at 9:34 PM on October 2, 2018 [11 favorites]
Someone in this position is not in any kind of shape to actually support a production application that gets any kind of serious traction. I think if you really want to be supportive, the best thing you can recommend is that he build something he wants without even thinking about a market. The *next* app, maybe he can think about marketing it. Real useful apps have a completely different way of being constructed than things that are actually useful for learning purposes, among which that they start not with "what do you know about programming and what comes next to learn" but "what does the user need".
Variants on to-do lists are popular for this kind of thing for a reason. Don't make him think that he's going to actually take his first foray into producing something and turn it into a thing that's really ready for prime-time. It's just a route to really bad disappointment.
posted by Sequence at 9:34 PM on October 2, 2018 [11 favorites]
Seconding starting with a clone of something already made that's useful. There's very, very little low hanging fruit left as far as useful small things that haven't been made, and Sequence is right that you need to get a few apps under your belt before you begin to think of making something for the public.
posted by Candleman at 10:13 PM on October 2, 2018 [5 favorites]
posted by Candleman at 10:13 PM on October 2, 2018 [5 favorites]
Best answer: If he plans to make it free, would he be interested in making it open source? One particular path to take in cloning an existing category of app would be to look through F-Droid, or some other alternative-to-Google repository that specializes in open source apps, and find an as-yet-unfilled niche.
posted by XMLicious at 10:42 PM on October 2, 2018 [3 favorites]
posted by XMLicious at 10:42 PM on October 2, 2018 [3 favorites]
If he has access to a smartwatch, there's more low-hanging fruit in that ecosystem that hasn't been made yet. And comparatively fewer apps, so he might be able to actually get some visibility for his.
Even just nice smartwatch faces that have some tiny piece of interactivity are great, and relatively easy to make. For example, I would like one where I can set an arbitrary goal, and have the watch face change colour over the day as I update my progress on that goal. You can get such watch faces for fitness that change colour, or have some other feature (stars or whatever) turn up based on your number of steps etc. But what if I want to record e.g. number of words written, or mls of water drunk, or something else that I define?
You'd just need to create an incrementer (like "+1" button, where the user defines what the increments are when they set their goal. E.g. for water it might be +1 glass. For writing it might be +100 words). The main interface (on phone or watch) would let the user define their goal, their increment, and then add or minus an increment to the total (or click "reset" to zero). The watchface would take the global variable that stores the total and use that to determine something like background colour or similar. You could make something like this in watchmaker pro, I think, without even having to deal with the Android Wear or IOS watch developer stuff.
A similar thing could work as a phone app and could probably be done with Tasker rather than jumping straight into full-on programming. In fact, in general, he might want to start with scripting things using Tasker. You can build some quite sophisticated apps with that.
posted by lollusc at 11:56 PM on October 2, 2018 [1 favorite]
Even just nice smartwatch faces that have some tiny piece of interactivity are great, and relatively easy to make. For example, I would like one where I can set an arbitrary goal, and have the watch face change colour over the day as I update my progress on that goal. You can get such watch faces for fitness that change colour, or have some other feature (stars or whatever) turn up based on your number of steps etc. But what if I want to record e.g. number of words written, or mls of water drunk, or something else that I define?
You'd just need to create an incrementer (like "+1" button, where the user defines what the increments are when they set their goal. E.g. for water it might be +1 glass. For writing it might be +100 words). The main interface (on phone or watch) would let the user define their goal, their increment, and then add or minus an increment to the total (or click "reset" to zero). The watchface would take the global variable that stores the total and use that to determine something like background colour or similar. You could make something like this in watchmaker pro, I think, without even having to deal with the Android Wear or IOS watch developer stuff.
A similar thing could work as a phone app and could probably be done with Tasker rather than jumping straight into full-on programming. In fact, in general, he might want to start with scripting things using Tasker. You can build some quite sophisticated apps with that.
posted by lollusc at 11:56 PM on October 2, 2018 [1 favorite]
An app that averages your steps for the past week/month/year. I have missed this functionality from the Apple app ever since I moved to Android and further advances in Google Fit have not gone in the right direction.
posted by robocop is bleeding at 3:23 AM on October 3, 2018
posted by robocop is bleeding at 3:23 AM on October 3, 2018
Best answer: Someone just asked a question that seems like a potentially good fit.
posted by hoyland at 4:14 AM on October 3, 2018 [1 favorite]
posted by hoyland at 4:14 AM on October 3, 2018 [1 favorite]
I've been told that Android homescreen widgets can't do animation. In the past though, I've had a widget that would display any webpage as a widget, and when I put in the url for an animated gif, it did indeed animate. And there are clock widgets out there that have displays for seconds alongside minutes and hours, so those sort of count as animated, just at one frame per second.
The gif I was using happened to be for a doppler weather radar station in my local area. The process wasn't reliable though. I suspect that the owner of the doppler radar may have decided that the unusual downloading pattern of that specific file without its associated webpage was some sort of attack and blocked it, even after i tried useragent changes and VPNs.
So, what I'd like to see, is a simple widget that, given a location (either from location services or user defined), will periodically pull weather radar images from a reliable and publicly available source (perhaps NOAA) at the top of every hour and display whatever the latest one is.
If the widget is tapped fetch any new images posted since the top of the hour, and then play an an animated loop of the last 10 images (maybe the number of images could be user defined too). One frame per second would be an alright speed for the loop, but anything up to 5 seconds per frame would be acceptable. After 120 seconds from the tap, revert back to the static display of the latest image.
If it's not too much of a complication, the ability to crop a specific area out of the source radar image so that the widget can display a zoomed in view of a particular city and not the whole region would be icing on the cake, but is not really a requirement.
posted by radwolf76 at 6:00 AM on October 3, 2018
The gif I was using happened to be for a doppler weather radar station in my local area. The process wasn't reliable though. I suspect that the owner of the doppler radar may have decided that the unusual downloading pattern of that specific file without its associated webpage was some sort of attack and blocked it, even after i tried useragent changes and VPNs.
So, what I'd like to see, is a simple widget that, given a location (either from location services or user defined), will periodically pull weather radar images from a reliable and publicly available source (perhaps NOAA) at the top of every hour and display whatever the latest one is.
If the widget is tapped fetch any new images posted since the top of the hour, and then play an an animated loop of the last 10 images (maybe the number of images could be user defined too). One frame per second would be an alright speed for the loop, but anything up to 5 seconds per frame would be acceptable. After 120 seconds from the tap, revert back to the static display of the latest image.
If it's not too much of a complication, the ability to crop a specific area out of the source radar image so that the widget can display a zoomed in view of a particular city and not the whole region would be icing on the cake, but is not really a requirement.
posted by radwolf76 at 6:00 AM on October 3, 2018
Best answer: Almost all possible apps have been made. If the goal is just to learn programming, it's probably not worth worrying about originality. In fact, it is better to do something unoriginal, because it will probably be bad; he should save the good ideas for when he has developed the skills to execute on them. An unoriginal idea that he would actually use himself is probably the best bet.
To actually make an app that doesn't already exist, and where the barrier to entry isn't too high, the trick is to find a weird niche, maybe one where people are currently using lots of Excel spreadsheets, and make a custom app for their workflow. Like weird minor sports where people want to keep track of scores or workouts or something. If he can figure out the workflow really well, it's possible to make something genuinely useful for that small group.
posted by vogon_poet at 12:19 PM on October 3, 2018 [1 favorite]
To actually make an app that doesn't already exist, and where the barrier to entry isn't too high, the trick is to find a weird niche, maybe one where people are currently using lots of Excel spreadsheets, and make a custom app for their workflow. Like weird minor sports where people want to keep track of scores or workouts or something. If he can figure out the workflow really well, it's possible to make something genuinely useful for that small group.
posted by vogon_poet at 12:19 PM on October 3, 2018 [1 favorite]
Response by poster: Sequence/Candleman: yeah, not that serious. Probably not looking to make an app that needs supporting or monetization or dedication. Simple, maybe useless, maybe fills a small niche.
XMLicious: I have no clue how he wrote the first app or even if he knows FOSS. I'm just not assuming the worst and going under the assumpton that like any ~20 something he's browsing the web and reading tutorials and maybe using VisualCode or whatever to create an app. No clue if he's sideloading or putting it in a store. But good point on F-Droid (and maybe XDG? forums or the like).
Garm's barcode and note and rating and just auto-remembering sounds close to on point. Depending on the barcode scanning bit. I'd use that.
hoyland's finger pointing to a schedule of bells is right up there in the not too hard, possibly useful, could be repurposed /extended into maybe useful in a general case. And also basically a calendar/reminder app. Play that clocktower tune with the bongs on the hour and the half, ring bells at times. Be smart enough to do what you want just by starting the app. A bit of config and using the Time/Date input widgets of the OS, keeping database and tracking time.
So keep them coming... just maybe a bit more along the lines of trivial, but a bit above making a button that makes fart noises.
I'm sort of a cell phone Luddite. I use it for Waze/Maps, maybe checking mail on a rare occasion, checking the temperature... but mostly to tether to a laptop. I'm mostly clueless on apps.
posted by zengargoyle at 2:46 PM on October 3, 2018
XMLicious: I have no clue how he wrote the first app or even if he knows FOSS. I'm just not assuming the worst and going under the assumpton that like any ~20 something he's browsing the web and reading tutorials and maybe using VisualCode or whatever to create an app. No clue if he's sideloading or putting it in a store. But good point on F-Droid (and maybe XDG? forums or the like).
Garm's barcode and note and rating and just auto-remembering sounds close to on point. Depending on the barcode scanning bit. I'd use that.
hoyland's finger pointing to a schedule of bells is right up there in the not too hard, possibly useful, could be repurposed /extended into maybe useful in a general case. And also basically a calendar/reminder app. Play that clocktower tune with the bongs on the hour and the half, ring bells at times. Be smart enough to do what you want just by starting the app. A bit of config and using the Time/Date input widgets of the OS, keeping database and tracking time.
So keep them coming... just maybe a bit more along the lines of trivial, but a bit above making a button that makes fart noises.
I'm sort of a cell phone Luddite. I use it for Waze/Maps, maybe checking mail on a rare occasion, checking the temperature... but mostly to tether to a laptop. I'm mostly clueless on apps.
posted by zengargoyle at 2:46 PM on October 3, 2018
Best answer: An alarm clock that can be tied to astronomical events by location. Sunrise? Moonset? Begin civil twilight? High noon? Peak meteor shower? Just a reminder each day of the natural world's rhythms.
posted by VelveteenBabbitt at 10:53 AM on October 4, 2018
posted by VelveteenBabbitt at 10:53 AM on October 4, 2018
Response by poster: Thanks all, pretty much wrapping it up because the email has been sent. I picked things that I imagine could be done without using Cloud or API keys and is really pretty simple but ticks many boxes of what could be done while leaving it at bog simple.
I like the teacher's bell but as a more developed app. Mad bonus if it's in C++ or Go or something that can be 32-bit Windows (for the teachers) and on up. Not actually attached to the phone on that one. (I use cron and at and heavens forbid systemd timers for that. But I still might put it on my phone if somebody else wrote it and the config was portable.
posted by zengargoyle at 8:29 PM on October 7, 2018
I like the teacher's bell but as a more developed app. Mad bonus if it's in C++ or Go or something that can be 32-bit Windows (for the teachers) and on up. Not actually attached to the phone on that one. (I use cron and at and heavens forbid systemd timers for that. But I still might put it on my phone if somebody else wrote it and the config was portable.
posted by zengargoyle at 8:29 PM on October 7, 2018
« Older What are these bits of metal coming out of my... | Parking in Boulder CO on Thirsday? Newer »
This thread is closed to new comments.
posted by Garm at 9:17 PM on October 2, 2018 [4 favorites]