Apple Core Rot: iOS 14 and macOS Safari Update Bring Treasure Trove of New Bugs

I updated to iOS 14, sucker that I am, because I hope to be traveling in a week or two, and better to take the hit now than while on the road.

There are all sorts of drawing bugs in iOS in which things go blank (nothing there), such as in Messages. Crap gets turned on I don't want too. I am sure I’ll discover many more “pleasures”.

Interestingly, I see the same drawing problems in Apple Safari in the URL bar on macOS—just a lock icon (for https) and... nothing else. Fortunately, this is infrequent. At least one major web sites is still unusable (hangs). Thank you... Apple.

• Some websites that used to work fine, e.g., Chase.com are drawn with a blank area, a refresh is required.
• Save Image As... is not working at all. I get a rainbow beachball, then nothing. WTF?!!!!!! I am guessing that some incompetent engineer forgot to reconcile download restrictions on a web site with saving an image from that web site—idiots.

Blank URL bar in Apple Safari in macOS Catalina

The accelerating unprofessionalism of the software development process at Apple is hurting everyone. Rush a release out, and let the suckers (er... customers) report the worst bugs, then fix half of them over 6 months, repeat ad nauseum.

Staale A writes:

After havning installed the last Safari update 14.0 my iMac (OS 10.15.6) has more or less stopped working.

Startup used to take 60-90 secs with all my “relevant” apps, now we are talking 10-15 min and every new app that I open takes 1-2 min to really open, like for Firefox, which used to be more like +/- 5-10 sec.

MPG: always wonderful to have Apple clusterf*ck your machine and make it unusable.

James G writes:

I too made the big mistake of upgrading toiOS 14 on my iPad and iPhone, and Safari in macOS.

Excel will now not let me save anything, stating that I don't have the correct permissions. If i go to terminal to repair all my permissions, it starts the process but I get the crawling slashes and the process wont complete. I can close Terminal and return to using my computer but i have no idea if the process repaired anything. Excel still wont save.

PDFs embedded in popup windows now show up blank.

Tabs in Safari now just show the website icon, which is totally useless if all you have is a generic icon, or maybe six Amazon tabs open. How the heck are you supposed to tell one tab from the next? The first time I opened Safari after the upgrade, the visual was like “what the fuck, man?” Fortunately, you can apparently undo this by going to Safari/Preferences/Tabs and unchecking the last box to deselect “Show icons only.”

Who the heck asked  for these useless features? Did some dweeb 20-something software engineer think this would be cool? Probably the same idiot who thought microfiche fonts in blue on a pale gray background with no contrast would look cool.

First rule in creating a software update is dont break what you already have. And if you’re going to introduce new features, keep the old preferences the same but publish a list of the new features and how to implement them if you want them.

I am still trying to deactivate the multi-window features in my iPad from the last software “upgrade.”

I swear Apple is now creating software for tik-tok stars and teenyboppers. Certainly not for people who use their macs for productive work.

MPG: I was going to advise James G to try the Security & Preferences control panel, but the crucial Security & Preferences control panel on my Mac Pro is now broken by Apple. Between this problem and hard crashes, Apple has managed to turn my $19K Mac Pro into the most unstable Mac I have EVER had.

Feckless Apple schmucks are running the show. The sheer jaw-dropping incompetence is unprecedented, but starting back in 2013 I saw the Apple Core Rot trend. It will never get better until the leadership makes it a priority, and that blame rests squarely at the top.

70% of music is non-playable "song is not currently available in your country or region"

Protecting Yourself in the Age of the Internet is Darn Hard: 3 Credit Cards Compromised in 2 Weeks after Mortgage Refinance Application

I recently applied for a mortgage refi with no points and very low fees.

I’ve worked with this local outfit several times before, so I know them to be solid. But there is a long chain of custody for my information between them and the processing end. All it takes is one crook somewhere in that chain.

In the past ~2 weeks, all of our credit cards have been compromised: business debit card, then personal ATM/debit card, then personal credit card, about 7 days apart for each of them. Because of the timing and the way they were tested in each case, it was clearly designed to fly under our radar. My conclusion is that somehow my mortgage refinancing information was compromised, and I have to assume—in its entirety. I've taken mitigation measures.

How credit cards are “tested”

In each case, a small charge was made to test the card first, subtle enough that most people are unlikely to notice—a few dollars at a frequented store or place. For example, a $6 charge at a local CalTrain station, a small charge at gas station, etc. In the last compromise, it was even more secretive—a small charge, then radio silence, presumably for a bigger heist later.

It sucks being held hostage: I cannot lock my credit because the loan has to close/fund first.

And the refinancing process these days is taking 2 months to complete due to enormous demand overloading the system. There is no way to speed it up.

Be aware of credit card or debit card activity

Most banks have some kind of alert service for both bank accounts and credit card.

Set it up so that you are both emailed and texted whenever any transaction occurs above a very small amount, e.g., 10 cents. You know when you charge something, so the idea is that you’ll be made aware when someone else is using your card via text message and email.

Also, scan your credit card charges and bank activity at least once a week. In this latest credit card compromise, the charge were made to look innocuous and five days had gone by before I saw it—I could easily have overlooked it.

Apple Core Rot: Syncing of Random Music Tracks to iPhone, iPad

Apple Core Rot is just getting suckier and suckier.

Besides erasing 200GB of music after an iOS update, the totally broken syncing function randomly syncs tracks that have not changed in 5 years (literally!). Every few days when I sync my iPhone 7 Plus, I see a few dozen tracks get synced-up, few if any I’ve even listened to for a long time.

Does Apple bother testing anything any more?

Bloomberg: Apple Gives Users More Time to Buy AppleCare After Sales Slow

This is how it used to be, before Apple’s bean counters decided to implement their latest anti-customer policy—very unfair and a very ugly policy that affected me and my kids with respect to our computers. This, after being an Apple customer since 1984. Shame on you, Apple.

But the best way to judge mistakes is seeing whether or how they are corrected/

Seems like Apple finally came to its senses, but it remains outrageous to pay a super premium price for a pathetic 1-year warranty, even for high-end professional products.

OWC, for example, offers 3-year and 5-year warranties on their excellent products for Mac and PC, as do companies like NEC on the wide gamut displays.

Apple Gives Users More Time to Buy AppleCare After Sales Slow

August 17, 2020

Apple Inc. on Monday told retail and customer-support employees that the company is expanding the time period when customers can subscribe to its AppleCare+ service.

Consumers currently have a chance to sign up to the warranty-and-support program within 60 days of buying an Apple product. This subscription window is increasing to up to a year now in the U.S. and Canada.

...The updated program requires a user to have their device screened by an Apple retail employee to make sure it is not damaged before they can buy the plan.

Apple Core Rot: Apple Messages.app Stops Working after a Few Images

Apple Core Rot is just getting suckier and suckier.

How hard is it to send a message? Pretty darn hard when Messages.app just goes AWOL.

I've tried multiple times, and each time it is the same: 3 or 4 messages send (images), then all subsequent ones fail. A

When this happens, there is nil outgoing network activity and the Messages.app progress bar just halts in place. Otherwise, the Messages.app behave normally, and text messages send OK. Worse, there is intense incoming (download) network activity (up to 14MB/sec)—WTF?

This is on wired ethernet and everything else is working. Recipients (4) all online with good cell reception (at home, I was sending some pictures to family).

Messages.app fails on all sends of images, no network activity

Messages.app fails on all sends of images after 2-4 images have been sent

Brian T writes:

I had the same problem. For me it had something to do with external drives. When I send from the MacBook  internal drive it works fine. Who knows on that one!!! Mac OS is getting to be a buggy POS.

MPG: I tried again today. It doesn't just fail on images, but fails on even simple things.

Jeffrey J writes:

I’m having the same issue with messages on Catalina.  Seems a common problem that has been reported for over a year without Apple fixing it.

A work around that works for me and some others is to copy images to a folder on the desktop and send from there (which is a major pain).  Certainly qualifies as core rot.

MPG: good to know this is not a unique problem on my end.

Kyle M writes:

Hello - i really enjoy your columns - especially the Apple CORE ROT ones. We are on the exact same page with that one. I have been coding on Apple products since the Apple // all the way up to today’s systems.

The last half decade as you know has seen the shine really come off the Apple. I wanted to share an example of that with you that I thought you would get a kick out of.

Apple's 2 factor ID thing is just another poorly done product. Anytime something needs the 2 factor sign in the code is sent to the device at the furthest end of the house, almost never the phone or iPad that’s next to me. The other day I needed to sign into the iTunesConnect website to check on some apps. And right away the iMac browser fires off the 2 factor ID thing and sure enough, an iPad half way across the house chimes. But what was really funny is they sent the code also to the iMac I was using to log into their website with - great security there guys. Maybe a little less work on emojis and a little more work on their code would be a good idea.

MPG: Yep, I’ve seen the same sign-in idiocy. Along with the forced use of a poor password (hard to type on iOS with mixed case/digits/chars, and cannot see the whole password to see if a mistake was made), the whole security approach is a mess.

I don’t see how this anti-customer sloppiness is anything but bad for Apple in the long run.

Apple Core Rot: Erases my 200GB of Music on iPad and iPhone Every Time I Update iOS.

Apple Core Rot is just getting suckier and suckier.

Every time I do an iOS update (e.g., 13.6 ===> 13.7), the geniuses at Apple have designed it to erase all 200GB of my music (electronic version of CDs I own).

Do you know how long it take to re-sync 200GB? F****g hours!!!

Apparently the Apple engineers do not know and/or don’t give a damn just how much maintenance they impose on users, with endless iOS updates full of bug fixes because they shipped garbage to start with, and then time-wasting syncing hassles like this.

Shame on you Tim Cook, shame on you Apple.

Audible.com: False Advertising Claiming that Audio can be Listened-To Offline

For years, I avoided signing up at Audible.com, because I was concerned that my time spent without internet access would deprive me of the ability to listen. And since that is 95% of the time I do listent to audiobooks, that was meant nearly all the time.

Then I signed up last year and things seemed to go well. But now either a bug has been introduced, or Audible.com is engaging in blatantly false advertising.

On my recent trip, I made sure to download 50 or so audiobooks to my iPhone 7 Plus (plenty of space on the phone), so that I could pick and choose as I saw fit. And yes, they were downloaded and show as such in the Audible app.

Multiple days when I had no cell phone service, I was greeted with a refusal to play *any* of those audiobooks with a “connect to the internet to continue listening” message for every book I tried (a dozen or more). In effect, my usage of what I paid for was rendered useless, because 90% of the time on my trip I had no cell phone service and the app would not play anything.

These jokers at Audible.com explicitly state that you can listen offline, but that was not possible over the course of ten days—not once.

The support at Audible.com first suggested that I delete all books that won’t play, and download them again—failing to understand that this eats up huge amounts of cell phone bandwidth, and that is impossible with no internet. Besides, I download all my Audible content on WiFi before I leave for my trips. The last thing I need is to chew up gigabytes of cell phone bandwidth doing it all over again—and no one should have to do it over again—downloaded is downloaded.

I then moved onto a supervisor, who is taking all the technical details and filing a bug report. Dunno what will come of it, and I am sure hoping my next ~3 week trip I won’t see this problem again. But at least I have to give Audible.com credit for taking it seriously.

Audible.com: refuses to play downloaded audiobooks when not on the internet

Scott MF writes:

Reading the audible books website carefully it seems that the off-line claim is for books you’ve purchased. 

All subscription services have a common problem with time-rental material. If your on-line it easy check the subscription status. Off-line - you need to have a window of time where the subscription is valid without checking it otherwise someone with a valid subscription can’t use the service. There have been more battles between the operations groups and the sales/accounting groups than I can remember over how long the time should be.  Sales always gets stuck on a possible revenue loss from someone who cancels his subscription but still has access because they stay off-line or some such nonsense. Always. Ridiculous.  Invariably, when executive management decides that they want users to be able to ride out Internet outages without service loss sales insists on some ridiculously short time interval - less than a day, sometimes less than an hour.  Politically, sales controls IT & OPS so they get what they want.   Until they lose customers, then sometimes they get more reasonable about the time. 

This is a generality about a situation thats been a long term frustration for me.  Amazon is bad but some of the boutique software vendors are just plain absurd.  Time to cut the rant off....

MPG: correct that all my audiobooks are purchases.

Even when I had internet for hours some days, only those books I played briefly would ever play again. With 138 titles and growing, it is not possible to “touch” all 138 books (start playing each of them), nor is it desirable meaning they are started over and then show up as in progress—very annoying. Yet I want to be able to pick and choose what to listen to among my library—that’s the whole idea of owning books.

Audible.com: annoying message occurring many times a day which is impossible to comply with

OWC Thunderbay FLEX 8: Discussing Various Configuration Possibilities

MPG tested the OWC Thunderbay Flex 8 with four 2TB SSDs and four Toshiba 4TB MG04ACA hard drives. Configurations up to 128TB (8 X 16TB) possible.

OWC ThunderBay FLEX 8, cover off

Looking for an all-in-one Thunderbolt 3 peripheral that covers just about all the bases? Look no further than the OWC Thunderbay FLEX 8!

Added to my review is a page discussing configuration as-shipped, and additional possibilities:

OWC Thunderbay FLEX 8: Configurations

OWC Thunderbay FLEX 8: Designing an OWC Thunderbay FLEX 8 Configuration for Still Photography or Videography

For example, it is possible to have up to 48TB of SSD storage using a 16TB PCIe SSD, and four 4TB SSD blades and four 4TB SATA SSDs.

Yesterday’s review pages:

• OWC Thunderbay FLEX 8: Overview
• OWC Thunderbay FLEX 8: HDD Speed Across Volume Capacity, RAID-0 vs RAID-5
• OWC Thunderbay FLEX 8: HDD Speed vs Transfer Size
• OWC Thunderbay FLEX 8: SSD Speed Across Volume Capacity
• OWC Thunderbay FLEX 8: SSD Speed vs Transfer Size
• OWC Thunderbay FLEX 8: SD Camera Card Slot Speed

CLICK TO VIEW: Example OWC Thundebay FLEX 8 Configurations

OWC ThunderBay FLEX 8

OWC Thunderbay FLEX 8 REVIEWED: all-in-one 8-Bay Thunderbolt 3 Enclosure for Hard Drives and SSD + PCIe Slot + DisplayPort 1.4 + USB Hub and SD and CFExpress Slots

MPG tested the OWC Thunderbay Flex 8 with four 2TB SSDs and four Toshiba 4TB MG04ACA hard drives. Configurations up to 128TB (8 X 16TB) possible.

Looking for an all-in-one Thunderbolt 3 peripheral that covers just about all the bases? Look no further than the OWC Thunderbay FLEX 8!

CLICK TO VIEW: Example OWC Thundebay FLEX 8 Configurations

• OWC Thunderbay FLEX 8: Overview
• OWC Thunderbay FLEX 8: HDD Speed Across Volume Capacity, RAID-0 vs RAID-5
• OWC Thunderbay FLEX 8: HDD Speed vs Transfer Size
• OWC Thunderbay FLEX 8: SSD Speed Across Volume Capacity
• OWC Thunderbay FLEX 8: SSD Speed vs Transfer Size
• OWC Thunderbay FLEX 8: SD Camera Card Slot Speed

OWC ThunderBay FLEX 8

Apple 2020 iMac 5K: Save up to 77.7% with OWC Memory

Get 2020 iMac 5K at B&H Photo.

More about the 2020 iMac 5K and see yesterday’s overview of the 2020 iMac 5K.

Apple wants a +$2600 for 128GB memory in the 2020 iMac 5K. This profit margin would make a Wall Street weasel blush.

Or you can get the 128GB at OWC for $600, a savings of $2000.

16GB (8GB x 2) OWC2666DDR4S16P only $69.99 $130 savings over factory
32GB (8GB x 4) OWC2666DDR4S32S only $139.79 $450 savings over factory
32GB (16GB x 2) OWC2666DDR4S32P only $134.99 $450 savings over factory
64GB (16GB x 4) OWC2666DDR4S64S only $268.99 $700 savings over factory
64GB (32GB x2) OWC2666DDR4S64P only $309.99 $680 savings over factory
96GB (32GB x 2 + 16GB x 2) OWC2666DDR4S96S only $449.99 N/A factory purchase
128GB (32GB x 4) OWC2666DR4S128S only $599.99 $2,000 savings over factory

Memory for 2020 iMac 5K

Memory wish list • Deals on OWC

$599 OWC 128.0GB (4x 32GB) 2666MHz DDR4 PC4-21300 SO-DIMM 260 Pin Mem… in All Other Categories

$450 OWC 96.0GB (2x 32GB + 2x 16GB) 2666MHz DDR4 PC4-21300 SO-DIMM 26… in All Other Categories

$308 SAVE $292 = 48.0% OWC 64.0GB (2x 32GB) 2666MHz DDR4 SO-DIMM PC4-21300 SO-DIMM 260… in All Other Categories

$269 SAVE $131 = 32.0% OWC 64.0GB (4x 16GB) 2666MHz DDR4 SO-DIMM PC4-21300 SO-DIMM 260… in All Other Categories

$135 SAVE $65 = 32.0% OWC 32.0GB (2x 16GB) 2666MHz DDR4 SO-DIMM PC4-21300 SO-DIMM 260… in All Other Categories

Apple Core Rot: Apple Notarizes Adware (Malware), Compromises User Systems

Sloppy engineering, irresponsible and negligent quality control practices, and disdain for the needs of working pros are now the hallmark of Apple software

Centralized trust is an inherently bad security design*, but that is how Apple now operates. A single mistake has worldwide implications for hundreds of millions of users. And the notarization process is guaranteed to be fallible.

"The vast majority of threats for macOS in 2019 were in the AdWare category." -Kaspersky

You cannot trust Apple to get system software right, but what can you do when Apple notarizes malware?

* Security that relies on one central authority is kaput should that central authority make a mistake or be compromised. In this case, that authority is Apple. The PGP model of distributed and graduated trust never caught on, but it is far superior and also allows users to build networks of trust—far more resilient.

Apple Approved Malware malicious code ...now notarized!?

We can confirm the payloads are indeed notarized via the spctl command (note the “source=Notarized Developer ID”):

As far as I know, this is a first: malicious code gaining Apple’s notarization “stamp of approval”.

What does this mean?

- These malicious payloads were submitted to Apple, prior to distribution.
- Apple scanned and apparently detecting no malice, (inadvertently) notarized them.
- Now notarized, these malicious payloads are allowed to run—even on macOS Big Sur.
- Again, due to their notarization status, users will (quite likely), fully trust these malicious samples.

To Apple’s credit, once I reported the notarized payloads, they were quick to revoked their certificates (and thus rescind their notarization status): Thus, these malicious payloads will now, no longer run on macOS. Hooray!

...This occurred on Friday, Aug. 28th.

Interestingly, as of Sunday (Aug 30th) the adware campaign was still live and serving up new payloads. Unfortunately these new payloads are (still) notarized:

$ spctl -a -vvv -t install /Volumes/Installer/Installer.app
/Volumes/Installer/Installer.app: accepted
source=Notarized Developer ID
origin=Developer ID Application: Aimee Shorter (73KF97486K)

Which means even on Big Sur, they will (still) be allowed to run: Big Sur, prompts, but allows.

If we extract the code-signing time stamp, we can see this (new) payload was signed on Friday PM (Aug 28, 2020 at 1:04:04 PM HST) ...likely after Apple’s initial “response”?

Both the old and “new” payload(s) appears to be nearly identical, containing “OSX.Shlayer” packaged with the “Bundlore” adware.

However the attackers’ ability to agilely continue their attack (with other notarized payloads) is noteworthy. Clearly in the never ending cat & mouse game between the attackers and Apple, the attackers are currently (still) winning.

...

Photoshop CC 2020 Rainbow-Beachballs and Goes Zombie-Process, macOS Catalina Goes Non-Functional Requiring Hard Reboot

Numerous bugs remain in macOS Crapalina. Nearly all of them will lurk around forever now that Catalina is a maintenance release. Soon we will savor an odiferous dumpster of new “features” shot-through with yet more bugs: macOS Big Sneer.

A fixed calendar date drives the release of macOS; quality control has zero to do with it. Year after year, a botched macOS Frankenstein drops like a cowpie. Frequent panicked updates follow, to fix what should have never been shipped to customers, and these updates are themselves poorly tested, causing yet more problems and bricking some machines.

Sloppy engineering, irresponsible and negligent quality control practices, and disdain for the needs of working pros are now the hallmark of Apple software.

Advice: professionals with working systems should stick with what works, never updating to a major new release without some essential benefit— for the past 3 releases, my work at least has only been degraded, with not a single benefit.

Apple or Adobe at fault? Demolishing productivity with failing software

No hardware changes, only latest macOS update and the latest Photoshop update about two weeks ago or whenever it was. I cannot know which to blame, and perhaps it’s only a macOS bug for which PS gets the blame, BUT these lockups happen ONLY when running Photoshop.

Ever since the latest Photoshop and macOS updates, my 2019 Mac Pro system has become unstable. EVERY DAY I have to hard-reboot it when Photoshop rainbow-beachballs.

Update Sept 20: the problem remains, but fortunately it is down to every 3-4 days.

It goes like this:

1. Out of the blue, Photoshop CC 2020 (latest) rainbow-beachballs. Its windows and palettes now cover all other apps and that alone makes the machine unusable. But it’s worse.
2. Force-killing Photoshop results in Photoshop reappearing as some kind of zombie process. Kill it again, and it’s still there. I am not sure if it is a new launch or Apple’s buggy software failing to kill it. I cannot be sure, because around this point, it become impossible to launch Activity Monitor to check and/or Terminal.
3. For a short while, other apps work if already launched, but after a minute or so, everything locks up and becomes unusable; attempting to launch any app hangs and it never happens.

It acts like the I/O system is locked up, which implicates macOS (a normal app should not be able to lock up the system). So that suggests that Photoshop might not be at fault. And yet this problem ONLY occurs when running Photoshop.

The only solution is to pull the power plug and hard reboot, or to do "sudo reboot -q" (if Terminal is already open or can be opened).

Sloppy engineering, irresponsible and negligent quality control practices, and disdain for the needs of working pros are now the hallmark of Apple software

For the past 3 macOS releases, my workflow as a professional has only been degraded, with not a single benefit. This decline has been going on since 2013, since I first coined the term Apple Core Rot.

Advice: professionals with working systems should stick with what works, never updating to a major new release without some essential benefit.

But lately, even minor updates are suspect, and caution is advised for any working pro who cannot afford downtime or glitches in workflow.

Apple software development practices are negligent and irresponsible

Numerous bugs remain in macOS Crapalina. Nearly all of them will lurk around forever now that Catalina is a maintenance release. Soon we will savor an odiferous dumpster of new “features” shot-through with yet more bugs: macOS Big Sneer.

A fixed calendar date drives the release of macOS; quality control has nothing to do with it. Year after year, a botched macOS Frankenstein drops like a methane-expelled cowpie. Frequent panicked updates follow, to fix what should have never been shipped to customers. Yet these updates are themselves poorly tested, causing yet more problems and bricking some machines.

Sloppy engineering, irresponsible and negligent quality control practices, and disdain for the needs of working pros are now the hallmark of Apple software.

Apple Core Rot: recent Apple macOS Catalina Update Crashes Mac Pro every other night with kernel panic: “thunderbolt power on failed”

Sloppy engineering, irresponsible and negligent quality control practices, and disdain for the needs of working pros are now the hallmark of Apple software

2019 Mac Pro kernel panic: “thunderbolt power on failed”

It’s not just this problem, it’s rainbow-beachball hangs that require me to hard-power-off the Mac Pro. <insert profanity to taste>

Along comes the recent Apple security update for macOS 10.15.6 (19G2021).

My Mac Pro now crashes with a kernel panic every other night. All Macs ought to run for months without issue (my older Macs running 24 X7 on older versions of macOS do!). Never before the update has this kernel panic been an issue, and the system configuration is the same: Apple botched things yet again.

panic(cpu 0 caller 0xffffff7f94f3b3c5): "XHC4(MacPro7,1):thunderbolt power on failed 0xffffffff\n"
@/AppleInternal/BuildRoot/Library/Caches/com.apple.xbs/Sources/IOPCIFamily/IOPCIFamily-370.141.1/IOPCIBridge.cpp:1398
Backtrace (CPU 0), Frame : Return Address
0xffffffd5d6cdba60 : 0xffffff801431a65d 
0xffffffd5d6cdbab0 : 0xffffff8014454a75 
0xffffffd5d6cdbaf0 : 0xffffff80144465fe 
0xffffffd5d6cdbb40 : 0xffffff80142c0a40 
0xffffffd5d6cdbb60 : 0xffffff8014319d27 
0xffffffd5d6cdbc60 : 0xffffff801431a117 
0xffffffd5d6cdbcb0 : 0xffffff8014ac1abc 
0xffffffd5d6cdbd20 : 0xffffff7f94f3b3c5 
0xffffffd5d6cdbd40 : 0xffffff7f94f22fab 
0xffffffd5d6cdbda0 : 0xffffff7f94f234ea 
0xffffffd5d6cdbdc0 : 0xffffff7f94f216d2 
0xffffffd5d6cdbe10 : 0xffffff7f94f2c023 
0xffffffd5d6cdbe30 : 0xffffff8014a137e4 
0xffffffd5d6cdbea0 : 0xffffff8014a135ea 
0xffffffd5d6cdbec0 : 0xffffff801435c605 
0xffffffd5d6cdbf40 : 0xffffff801435c131 
0xffffffd5d6cdbfa0 : 0xffffff80142c013e 
Kernel Extensions in backtrace:
com.apple.iokit.IOPCIFamily(2.9)[DF219CC1-366A-31FC-B8ED-17C584BA2549]@0xffffff7f94f1a000->0xffffff7f94f52fff
BSD process name corresponding to current thread: kernel_task
Mac OS version:
19G2021
Kernel version:
Darwin Kernel Version 19.6.0: Thu Jun 18 20:49:00 PDT 2020; root:xnu-6153.141.1~1/RELEASE_X86_64
Kernel UUID: 1D3A0F3D-D908-397B-BD16-8F48C0823A2E
Kernel slide:     0x0000000014000000
Kernel text base: 0xffffff8014200000
__HIB  text base: 0xffffff8014100000
System model name: MacPro7,1 (Mac-27AD2F918AE68F61)
System shutdown begun: NO
System uptime in nanoseconds: 373527826978039
last loaded kext at 216962633982576: >usb.!UHostBillboardDevice	1.0 (addr 0xffffff7f9594c000, size 16384)
loaded kexts:

Reader Comments

Fazal M writes:

Since the Mojave 2020-002 security update my Mac Mini started crashing every 2-3 days, with a crash dump having to do with the T2 coprocessor. Apple’s official answer: “upgrade to Catalina”, which is a non-starter. The only fix is to install Big Sur beta on another drive, which upgrades the T2 BridgeOS firmware to something less crash-prone.

Apple’s quality control has gone way beyond unacceptable, I don’t understand why you reward them by purchasing an expensive Mac Pro. My plan is to move over to Linux and keep the Mojave machine for legacy applications.

MPG: there are major factors involved for me—the pain points and cost would be great to make the switch, and I’m not persuaded that the other options are even viable, let alone free of other pain points. But yes, I despise rewarding these aholes with my business.

Shameer M writes:

I know you’ve posted extensively on Apple Core Rot over the years.  Thought you might like this one...

https://eclecticlight.co/2020/08/30/last-week-on-my-mac-is-macos-becoming-unmaintainable/

MPG: yep, with Apple it is now “it just doesn’t work”. But that started way back in 2013. What Apple is doing to professionals is contemptible.

Don’t get me started on my 2019 Mac mini, which is the biggest POS I have ever owned—it can take 10 minutes to get the display to sync, and the f'ing thing boots into password recovery mode (making ssh or Remote Access useless) without you knowing it if you can't see the screen and enter the password which turns the air blue with profanity.

Reader Question on diglloydTools IntegrityChecker: “verify terabytes of content back and forth between the cloud and other local RAID drives I have”.

IntegrityChecker Java (part of diglloydTools) runs on any platform with a JVM.

See also Data Integrity Over Time, and with OS Changes and Data Loss Prevented: IntegrityChecker Saves my Bacon by Detecting Corrupted Files after a Clone and Detecting Corruption / Validating Data Integrity Over Time and Across Drives and Backup/Restore.

Sobe S writes:

Here is what I am trying to do and I want to make sure your program can offer that.

I use Transmit 5 to download and upload lots of content sometimes terabytes of content back and forth between the cloud and other local RAID drives I have.

What I am looking for is a program I can run that can verify the quality of that transfer. I want to make sure my transfers have good data and be able to verify the files are not corrupt once stored somewhere. Can your program provide that?

MPG: IntegrityChecker verifies bit-for-bit data integrity anywhere—originals, copies or backups or clones—and on any operating system and file system, on read/write or read-only media. It’s a simple two-step process:

1. icj update the desired folder(s) or volume(s) in order to create the hash files used for verification.
2. icj verify the data (originals, backups or clones, copies, etc).

A single command can verify a specific folder or folders or an entire volume or folder hierarchy. Excepting quite slower computers, the speed of verification depends mainly how fast the data can be read.

Buy diglloydTools with IntegrityChecker Java

Additional features include the ability to instantly find duplicate files, quickly see status including missing files or folders, lighting-fast folder comparison based on hashes, and more.

Requirements

The enabling requirements to do this are:

• The folder(s) to be verified must accessible through a file system, e.g., a mounted volume on macOS, a drive letter ("D:" or similar on Windows), a mounted file system on Linux/UNIX.
• The folders to be verified must have hash information available. That means either a per-folder hash file or a hash hierarchy file. These hash files are created in the original folders using icj update, which does the initial hashing and then creates the hash files.

Addressing the question above

The uploaded data must be accessible on a mounted file system as per requirements above.

1. icj update the folder(s) involved on the local RAID volumes.
2. Sync or upload the data. The syncing obviously must include the .icj and/or .icjh hash files.
3. Verify the uploaded files using icj verify.

Some users feel ill-

More posts on IntegrityChecker.

2020 iMac 5K Display Stream Compression (DSC) and Its Role in Making Dual 6K Display Support Possible

The dual-6K display support of the 2020 iMac 5K puzzled me because of the huge bandwidth required. Bottom line is that Display Stream Compression (DSC) offers “visually lossless” bandwidth savings by compressing video data by up to 3:1. This is what accounts for the dual-6K display support.

Note that DSC is still poorly supported and even high-end video cards in the 2019 Mac Pro lack support for DSC.

See also Connect multiple displays to your 2020 iMac 5K.

Whether DSC is really visually lossless in the sense of professional photography and video work seems dubious*, so until and unless I can A/B compare, I will remain a skeptic that dual-6K support is suitable for professional use, even setting aside the bandwidth hit.

I didn’t really understand Display Stream Compression (DSC), so I did a little research.

* Tests show I have exceptional color discrimination, way out on the curve versus most people’s vision.

Wikipedia: Display Stream Compression

Emphasis added to excerpts below.

Display Stream Compression (DSC) is a VESA-developed low-latency compression algorithm to overcome the limitations posed by sending high-resolution video over physical media of limited bandwidth. It is a visually lossless low-latency algorithm based on delta PCM coding and YCoCg-R color space; it allows increased resolutions and color depths and reduced power consumption.[79][80]

DSC has been tested to meet the requirements of ISO/IEC 29170-2 Evaluation procedure for nearly lossless coding using various test patterns, noise, subpixel-rendered text (ClearType), UI captures, and photo and video images.[80]

... DSC supports up to 3:1 compression ratio with constant or variable bit rate...

DSC compression works on a horizontal line of pixels encoded using groups of three consecutive pixels for native 4:4:4 and simple 4:2:2 formats, or six pixels (three compressed containers) for native 4:2:2 and 4:2:0 formats...

Bit rate control algorithm tracks color flatness and buffer fullness to adjust the quantization bit depth for a pixel group in a way that minimizes compression artifacts while staying within the bitrate limits...

Note the “minimizes compression artifacts” part—great technology, but is it really suitable for professional work in color grading or fine-art photography?

There is additional information at “Tom’s Hardware: DisplayPort vs. HDMI: Which Is Better For Gaming?” that both clarifies and confuses: the table of data rates there does not mention the dedicated ~8 Gbps DisplayPort bandwidth that is available, which is over and above the Thunderbolt 3 data bandwidth. See Understanding Thunderbolt 3 Bandwidth and the Intel white paper referenced there for details.

Usable bandwidth using dual 6K displays vs data

Apple is lax at documenting 2020 iMac 5K technical details like the version of DisplayPort (“DisplayPort” with no version number), but it surely is Display Port 1.4 with 8b/10b encoding.

Assume a baseline dedicated 8 Gbps DisplayPort bandwidth along with a usable 25.92 Gbps for any general data (regular data, or video overflow data).

Dual 6K displays require bandwidth of 70.3 Gbps for 30-bit RGB color. Assuming 3:1 compression, the bandwidth requirement drops to ~23.4 Gbps—well within the bandwidth of a single Thunderbolt 3 bus but it does limit maximum performance for devices like fast SSDs.

Breaking it down, ~8 Gbps can be utilized from the dedicated DisplayPort bandwidth, and (23.4-8) = 15.4 Gbps of data bandwidth would also be consumed, leaving (32.4-15.4) = 17 Gbps write bandwidth for general purpose data. But with 8b/10/b encoding that really means (17*8/10) = 13.6 Gbps ~= 1.7 GB/sec usable bandwidth.

In other words, a fast SSD like a OWC Thunderblade would have a full gigabyte per second shaved off its performance (for writes).