What are the key takeaways from this Accidental Tech Podcast episode?

Key insights include: **App Store Review Bottleneck:** When App Store review times spike from the typical 1–2 days to 6–12+ days, the entire developer pipeline freezes. Developers cannot release approved companion apps, cannot push TestFlight builds for version 1.0.1, and cannot cancel pending submissions without restarting the clock. Pre-order dates become meaningless, editorial co-marketing windows close, and in-app purchase approvals tied to one platform block the other. Apple's self-appointed gatekeeper role means this is entirely a problem of their own making to solve.; **In-App Purchase Submission Architecture Flaw:** When launching a universal app with shared in-app purchases across iOS and Mac, developers must bundle the IAP review with only one platform's submission. Choosing iOS first — historically the faster reviewer — can backfire if Mac review completes in 6 days while iOS stalls at 12+. This architectural constraint means one delayed platform holds the entire product hostage, blocking revenue and development on both platforms simultaneously.; **DRAM Shortage Severity:** All memory production capacity from Samsung, SK Hynix, and Micron through the end of 2027 is reportedly sold out as of August 2026. Apple's attempt to use Chinese manufacturer CXMT as pricing leverage failed because CXMT's older DUV equipment raises production costs 30% above Korean competitors, and Huawei and Xiaomi have locked up most of CXMT's output under long-term contracts. Companies that secured supply early have pre-booked up to five years of capacity.

How long is this episode of Accidental Tech Podcast?

This episode is 128 minutes long. SignalCast provides an AI-generated summary so you can get the key insights in about 3 minutes.

Accidental Tech Podcast

704: Past the Ternus Horizon

August 13, 2026

128 min episode · 3 min read

Episode

128 min

Read time

3 min

Topics

Productivity, Health & Wellness, Fundraising & VC

AI-Generated Summary

Published Aug 14, 2026

Key Takeaways

✓App Store Review Bottleneck: When App Store review times spike from the typical 1–2 days to 6–12+ days, the entire developer pipeline freezes. Developers cannot release approved companion apps, cannot push TestFlight builds for version 1.0.1, and cannot cancel pending submissions without restarting the clock. Pre-order dates become meaningless, editorial co-marketing windows close, and in-app purchase approvals tied to one platform block the other. Apple's self-appointed gatekeeper role means this is entirely a problem of their own making to solve.
✓In-App Purchase Submission Architecture Flaw: When launching a universal app with shared in-app purchases across iOS and Mac, developers must bundle the IAP review with only one platform's submission. Choosing iOS first — historically the faster reviewer — can backfire if Mac review completes in 6 days while iOS stalls at 12+. This architectural constraint means one delayed platform holds the entire product hostage, blocking revenue and development on both platforms simultaneously.
✓DRAM Shortage Severity: All memory production capacity from Samsung, SK Hynix, and Micron through the end of 2027 is reportedly sold out as of August 2026. Apple's attempt to use Chinese manufacturer CXMT as pricing leverage failed because CXMT's older DUV equipment raises production costs 30% above Korean competitors, and Huawei and Xiaomi have locked up most of CXMT's output under long-term contracts. Companies that secured supply early have pre-booked up to five years of capacity.
✓iPhone 17 Production Risk: TSMC has completed approximately $1 billion worth of A20 Pro processors on its N2 process node, but packaging cannot proceed without corresponding LPDDR5X DRAM. This bottleneck threatens to compress final assembly schedules at Foxconn and BYD in China. The practical implication for consumers: iPhone 17 supply at launch could be constrained not by chip manufacturing capability but purely by memory availability, potentially affecting both pricing and initial unit availability through late 2026.
✓Cloud Storage Is Not a Backup: Services like GitHub, Apple Photos, Google Photos, and OneDrive do not constitute backups unless they support file versioning. Without versioning, a deleted file disappears from both local and cloud storage simultaneously. The more realistic failure mode is not data center loss but account bans, hacks, or Kafkaesque automated enforcement actions that lock users out permanently with no human escalation path. Backblaze at roughly $8 per month provides genuine versioned backup with actual restoration capability.

What It Covers

ATP episode 704 covers App Store review times spiking to 12+ days in August 2026, the cascading developer workflow failures this causes, Apple Watch lineup stagnation over 12 years, a global DRAM shortage blocking iPhone 17 production with roughly $1 billion in completed TSMC processors sitting unpackaged, and the impending November 5 Verizon CableCARD deactivation deadline.

Key Questions Answered

•App Store Review Bottleneck: When App Store review times spike from the typical 1–2 days to 6–12+ days, the entire developer pipeline freezes. Developers cannot release approved companion apps, cannot push TestFlight builds for version 1.0.1, and cannot cancel pending submissions without restarting the clock. Pre-order dates become meaningless, editorial co-marketing windows close, and in-app purchase approvals tied to one platform block the other. Apple's self-appointed gatekeeper role means this is entirely a problem of their own making to solve.
•In-App Purchase Submission Architecture Flaw: When launching a universal app with shared in-app purchases across iOS and Mac, developers must bundle the IAP review with only one platform's submission. Choosing iOS first — historically the faster reviewer — can backfire if Mac review completes in 6 days while iOS stalls at 12+. This architectural constraint means one delayed platform holds the entire product hostage, blocking revenue and development on both platforms simultaneously.
•DRAM Shortage Severity: All memory production capacity from Samsung, SK Hynix, and Micron through the end of 2027 is reportedly sold out as of August 2026. Apple's attempt to use Chinese manufacturer CXMT as pricing leverage failed because CXMT's older DUV equipment raises production costs 30% above Korean competitors, and Huawei and Xiaomi have locked up most of CXMT's output under long-term contracts. Companies that secured supply early have pre-booked up to five years of capacity.
•iPhone 17 Production Risk: TSMC has completed approximately $1 billion worth of A20 Pro processors on its N2 process node, but packaging cannot proceed without corresponding LPDDR5X DRAM. This bottleneck threatens to compress final assembly schedules at Foxconn and BYD in China. The practical implication for consumers: iPhone 17 supply at launch could be constrained not by chip manufacturing capability but purely by memory availability, potentially affecting both pricing and initial unit availability through late 2026.
•Cloud Storage Is Not a Backup: Services like GitHub, Apple Photos, Google Photos, and OneDrive do not constitute backups unless they support file versioning. Without versioning, a deleted file disappears from both local and cloud storage simultaneously. The more realistic failure mode is not data center loss but account bans, hacks, or Kafkaesque automated enforcement actions that lock users out permanently with no human escalation path. Backblaze at roughly $8 per month provides genuine versioned backup with actual restoration capability.
•Apple Watch Market Stagnation: The Apple Watch design has remained fundamentally unchanged since its 2014 introduction, making it visually indistinguishable from a distance across Series 0 through Series 11. Analytics from Overcast show a steep drop-off at each engagement layer: fewer watch users install the app, fewer of those use it, and almost none download files for local playback. The dominant use cases are notifications and fitness tracking — needs that do not require a full watchOS computer — suggesting a screenless or simplified device could address a large underserved market.
•Dealer Service Revenue Drives Out-of-State Sales Refusals: BMW dealers decline out-of-state sales primarily because buyers will not return for service, which is where dealers generate ongoing profit margin. This calculus is especially counterproductive for battery electric vehicles, which require significantly less routine maintenance than combustion vehicles. Third-party car buying services like Delivered (approximately $1,000 fee) and CarEdge, both founded by former dealers, source and negotiate vehicles on behalf of buyers, handling the entire process so the buyer only needs to sign at delivery.

Notable Moment

During a discussion of memory shortages, it emerged that TSMC is currently holding roughly $1 billion in completed A20 Pro chips that cannot move forward in production because the required DRAM has not arrived for packaging. The processors are finished and sitting idle — a striking illustration of how a single component shortage can stall an entire supply chain at enormous scale.

Know someone who'd find this useful?

Episode Transcript

I have a happy thing and a annoying thing. Okay. I'm gonna start with the annoying thing. Yes, please. I have been waiting twelve days for launching my new app. I've been waiting for app review. Two turtle doves. And a partridge in a pear tree. I wanted to launch it today. I wanted today's show to be all about my new app and promoting it. And I even set a preorder date from today thinking, like, obviously, in twelve days because I submitted on August 1, Say it's August 12. Like, obviously, twelve days is enough to get at least one loop around from App Review. And if they happen to reject it the first time around after, you know, two to seven days, I will be able obviously, I'll be able to just push that date back and, you know, do the whole thing again. And the Mac app was approved after six days, which is a lot, and the iOS app is just sitting there waiting. I filed a ticket today. No response. I don't know. App review is really having problems. Only in the last month or so, it seems like review times have spiked up to, like, six days for most apps. I can't imagine why. But okay. So everyone's like, yeah. It's Vibe Coding. Yeah. But Vibe Coding didn't just start last month. I know. But it's like, when do people, you know, when do the people start doing it, playing with it, figuring they're gonna put it up on the store? Like, it's kind of like, you know, a thing that builds like a snowball. And as as as you frequently point out, when people are stuck in traffic, you are traffic. You're submitting a Vibe coded app. You're one of the traffic people. Oh, it's partly it's partly AI coded. It's I didn't Vibe code the whole thing. But, in fact, I've wrote most of the code in it. But anyway, it's Apple's job to get the review times under control regardless of the spikes in traffic. They just wanna be stingy about it, and it gets harder when there's burst. But it's like, look, you chose to be the, as the EU would say, gatekeeper, so, figure out how to get it done. That's the thing. Like, you know, I I feel like in times like this, I think Apple often sees itself as the victim of external forces. That that is certainly the impression they give off. And when you talk to people inside, they they definitely have a victim complex about a lot of things. And and app review, it seems like they've always kind of seen themselves as, like well, when when times explode, it's obviously because the external world is forcing upon them some factor. Oh, all these people are submitting updates all at once, or, oh, AI coding is making things or spam attempts or whatever. But, like, they created this whole system. They put themselves …

Get the full transcript (27,616 words) + summary by email — free

One-time email with the complete transcript and AI summary of this episode. No account needed.

One email, no spam. We’ll also show you what SignalCast does.

Browse all Accidental Tech Podcast transcripts →