🤣🤣🤣🤣🤣

The founder of Marafiki watembezi is attempting to do a 13hr walk either on the 3rd of January or 4th Jan 2026. Mtakuja Twende ?? If you are interested please reach out to him pale IG and follow their account 🫰🏽🥳 Thanks:)

An update on the multiple collisions at the southern bypass yesterday caused by the steel roll that came off the flat bed. The other cars and truck have been cleared from the scene but the main truck who various axles are shuttered is still here with the roll

People who prioritize fitness carry a different kind of confidence.

last long run of 2025. special thanks to @SiroDevs !!

An hour ago we just had multiple accidents on the Southern Bypass near Dargoretti Rd Interchange, Near The Bull

good morning from HERE

@limentuga @SiroDevs Excessive weight is served at the table. Leave the hormones out of this.

I was able to cut initial setup time for the Android app from 125 seconds to 17 seconds and that's huge leap. Now maybe I will work on api calls to see if I could reduce the time. Batch processes == concurency == better ux

It’s hard to believe, but both photos show the same person: Wa Musau from my running club. He went from 135 to 70 kg through consistent running & mindful eating. He’s over 50—and he did it. Not with prayers, shortcuts or operations—just intention & consistency. Truly inspiring.

"There is more pressure in Africa than here, we've good players there.." that's the quote!

Full article here https://t.co/JFnvGKS8eA

The update is now live. New levels unlocked thanks to the wall of rejection that pushed me to this.

The entire initial setup, from fetch to persistent store, now took just 7 secs. I submitted the update, holding my breath. This time, the reviewers noticed nothing amiss. The app simply launched, prepared itself silently in the background, and was ready for use in under 10 secs.

A refactor later: 29 secs. Better, but I knew it wasn’t enough. That was still alot. The real bottleneck, I discovered, wasn’t the api requests—it was my mapping logic. I audited the code, stripping out redundant mappings and streamlining from DTO to Core Data. I ran the it again

Last Monday, a new approach crystallized: concurrent batch processing. I refactored the entire sync into smaller, parallel operations. My testing rig was my phone running the build and another phone with a stopwatch. The first run slashed the time from 120 secs to 40.

When I’m stuck like this and AI falls short, I turn to my secret weapon: GitHub Advanced Search. It almost always reveals a project doing exactly what I need. This time, it failed me. Every implementation I found felt alien to my architecture.

I knew Apple’s BackgroundTasks framework was the solution, but implementing it was another battle. The Swift SDK had evolved, clear examples were scarce, and one perfect article was locked behind a Medium paywall.

Defeated, I stepped away from Xcode for almost a week, focusing on other projects to clear my head. Then, the breakthrough idea struck: Background Tasks.

Suddenly in this review series Apple’s reviewers became clever. They flagged it immediately: I was “downloading extra resources without notifying the user.” My explanations—that this was essential offline data syncing—sparked a frustrating back-and-forth. The update was stalled.