So they want to migrate data from 27,000 apps into one new platform?
By September?
Anyone got a lifetime's supply of popcorn?
Indonesian president Joko Widodo on Monday ordered government officials to stop developing new applications. According to the president, Indonesia's central and regional governments together operate a fleet of 27,000 apps, many of which overlap or aren't integrated. New platforms often get developed every time a new minister, …
1. Government app sprawl is related to government bureaucracy sprawl. You gotta fix the latter before you can fix the former. Good luck with that.
2. How were they planning to cleanse and canonicalize their citizen database? As-is, you've got multiple databases with entries for Joey Citizen, Joe E. Citizen, Josey Sitizen, etc., all referring to the same person.
3. Setting a software development schedule by fiat produces bad results.
>Setting a software development schedule by fiat produces bad results.
We all know that because we develop software.
Indonesia's president is most likely not a software developer. So the problem's simple -- you make an order and it happens. Work of a moment.
(Seriously, though, there should be a better understanding of what "an application" is. I'd guess that a lot of these applications are things like spreadsheet scripts. But then counting without classifying is typical disconnected management style.)
From 27000 to one? There is only one way such endeavour will end: https://xkcd.com/927/.
《That's still fewer than 2 apps per island. How many software developers do they have on each island?》
I would wonder how much of this sprawl to provide "local" employment (to the governor's nephew) and less cynically to address purely local problems or a local solution to common problems not being addressed more widely.
Knocking up a quick and dirty app using a high level framework to implement a local government program that will only last a year or so may well be much more cost efficient than waiting for the central government IT to get round to implementing the function in their one "ring-of-power" app which will typically be delivered long after it is no longer needed and millions overbudget.
Shadow IT on steroids ;) - With the coming advent of AI/LLM and low/no code user generated applications, shadow IT seems to have faded as a clear and present danger from security theatre's repertoire.
While the President's is probably an ambit claim, the development of a central capability to deliver government service would be wise to prioritize those that have the greatest benefit, shortest delivery time, largest user base and least cost.
A relatively simple national electronic identity and licensing system might be progressively rolled out with immediate benefit to users with later services leveraged off the identity management foundations.
I would assume any apps would be initially for a android mobile/cellphones as these are far less expensive than apple's devices and likely more common by many factors.
I don't know how quickly phones turn over in these nations but if signicantly slower than in EU/US etc addressing the range of android versions might be a challenge. The Covid-19 check-in app in Q/AU didn't work on my (older)Samsung phone because it was one point version (x.1 v x.2) too low and no upgrades (The app just opened a url, posting the date/time, user's location number, numerical identity - not rocket science.)
I wouldn't mind being pleasantly surprise by a public sector software success from this nation. India seems to have made significant progress in some of these areas which given India's much larger scale and unique challenges I would been rather sceptical of any success.
Compare that to the French government open source repositories which grew from 9k in 2021 to over 20k in 2024. Considering France probably have fewer than 2k islands worldwide I would say the Indonesians still have room to grow if they want to be as inefficient as the French. Assuming one country's inefficiency can be measured from its island count and software modules it maintains.
CII: country inefficiency index = software module count / island count
France CII: 10
Indonesia CII: 1,58
I imagine you might incorporated all 27,000 apps into a single multi-call executable like a (very) busybox ;)
《CII: country inefficiency index = software module count / island count
France CII: 10
Indonesia CII: 1,58》
How many islands compose the UK? Leaving aside the Channel Islands, Isle of Man and other Crown Dependencies and overseas territories. (Rockall 0.5 ;)
From what I have read in these columns of the misadventures in this arena of the various levels of government and associated institutions, I suspect the UK would be a serious contender in the CII stakes.