We have been in charge of maintaining one legacy Android app for our customer. It is an app, which is used by end-customers in production, but it does not have any active development going on because it’s been ready for years now. If it would be up to us, then we would not touch that app and would let it live its life happily ever after.
Of course, there is no happily ever after when closed application stores are involved, so everything went south from here. It amazes me that a lot people only seem to be waking up now to the realities so many of us warned about when closed application stores took over from freely distributable applications over a decade ago. What do you get for that 30% cut of your revenue? Delays, nonsense rejections, no people to contact, and so much corporate bureaucracy it would turn Ayn Rand socialist.
This is the reality of doing business with monopolists.
Wow, well said Thom.
The problems described are all Play Store issues, none of them are really Android issues as such. The solution is easy: don’t use the store, host the software on your own website. That way you don’t have to pay Google and don’t have to follow their arbitrary and ever-changing rules which are purely there to benefit themselves rather than users and developers.
Requiring developers to target API 33, when they may not have access to such an environment to test in, is irresponsible and leads to exactly the sorts of problems described in the article.
Minuous,
I completely agree, however for 99+% of applications, they don’t
This is a huge problem for those of us trying to use android alternatives like lineage minus google dependencies.
They have critical mass and we’re forced into the apple/google mobile duopoly for banking, work, store apps, theme park apps, etc. Too many of them presume the duopoly is all there is and unfortunately marginalized alternatives don’t matter to them.
Reading the story, there seems to be several major issues. And fixing at least one of them would have prevented the end result.
1. Even though there is a Sep 1st deadline, the old API is already deprecated, and will not work.
2. There is no rollback to a known stable release in the market. If you make a mistake, you are stuck with it (along with the users). This should at least be a developer option, better yet, users should be able to roll back a broken update if they wish.
3. There are no automated rollbacks in the marketplace analytics. If the app is consistently crashing at login, this logs should at least trigger an (optional) rollback.
4. Worst of all, there is no escalation path to talk to a human. There should always be some way to do so. At least by paying for a service ticket. (Big companies like RedHat for example will see premium service subscriptions).
Again, the situation of pushing a rushed release that broke the users’s apps, without a fast way to recover is the product of all these coming together.
(need the edit button back)
Here here. Having an edit window is nice, we don’t want to loose it!
At least by paying for a service ticket.
That sounds a very nice plan to discover what lies to the south of the South Pole…
Every application sadly needs maintenance to keep working in production. This is the reality of life.
100% true. The app store policies are enforcing good developing hygiene. Its frustrating and annoying as someone in charge of a mobile app. The real fail here is that they pushed a major change without testing it adequately. That’s on them. I understand I’m not perfect, and I myself was in a similar situation with Apple when I clicked a checkbox I shouldn’t have, then had to scramble to create a new version and get it released. But you have to understand the risk of your deployment. Deploying outside of the app store also sucks. I have managed apps that way too. Downloads get saved and cached on corp servers, so you have years of support dealing with an issue you solved years ago because they keep installing the wrong version ;/
I have empathy for all those who’ve gone through the situation, but that’s the environment we have and you have to take the bad with the good that the app stores provide.
Also, this is just modern development, unless you are doing a really niche project, you are dependent on some third party to help you when things go wrong f0r what ever reason. And their motivations for helping you fix your issue are not as strong as yours are. Many of them you directly pay significant sums of money on a period basis, but still that doesn’t grant you the same control as if they were a part of your company under your management. You can’t make them do anything on your time schedule.
I have installed some apps outside of the PlayStore and other stores that autoupdate themselves. This should solve the wrong version problem.
Was just highlighting what it was like before app stores. Yeah, you can self update, that requires planning it from the beginning, coding in the app and backend, managing the infrastructure to do that, making sure its available, etc etc. its non zero cost. Then there are reputational things to consider as well. Some companies or individuals rightly, or wrongly won’t install apps not in the app store. I can’t say I blame them, its a bit sketchy.
I personally wouldn’t want my company’s corporate apps in an app store.
Not cooperate apps, apps for b2b. Think Atlassian like products.
Sketchy how?