High Sierra broke AFP searching on our server | Hacker News Reader

High Sierra broke AFP searching on our server

idoru.be

114 pointsidorube106 comments

Firstly, Apple has never been a respected actor in the server market. If you're investing heavily into Apple servers, you're investing in a niche. Presume accordingly. Secondly, converting/reformatting 300 Tb of disk is a lot of work, but it's probably less work than contacting Apple, and whining about how Apple is going down the drain in a public blog post.

I've never been a big Apple fan in terms of the company, but the iPhone is an excellent product. The iPod was an excellent product. I have never heard anyone say that in terms of server infrastructure, Apple makes "excellent products". The only lesson here is: don't drink the kool aid, and investigate every use case thoroughly without making emotional assumptions.

edit: Despite knowing almost nothing about AFP, I found articles on Google saying that Apple shifted from AFP file sharing to SMB2 in an article dated 2013 -- that's five years ago! Apple themselves state that AFP is deprecated. If you elect to run your own servers and support your own services completely, these are news you should be reading.

I have never heard anyone say that in terms of server infrastructure, Apple makes "excellent products".

Once upon a time they did. The Xserver and Xserver RAID where really amazing products when they where first released. However they couldn't really gain traction and quickly gave up on that market.

Certainly the hardware was great and the Xserve RAID was a surprisingly good value for its price.

given that BSD is having trouble of keeping enough contributors to keep up with security issues its not an idle claim that Apple is going down the drain...its in fact reality but its like boiling a frog in boiling water its tarting out slow

“Apple shifted from AFP file sharing to SMB2”

* SMB2 with undocumented Apple-specific extensions (https://www.mankier.com/8/vfs_fruit)

> edit: Despite knowing almost nothing about AFP, I found articles on Google saying that Apple shifted from AFP file sharing to SMB2 in an article dated 2013 -- that's five years ago! Apple themselves state that AFP is deprecated

I too found third-party articles, but the only official notice I could find from Apple was in the recent APFS FAQs [1]. Either I'm searching for the wrong terms, or Apple could have communicated the issue better.

[1] https://developer.apple.com/library/content/documentation/Fi...

It was worded at the time (2013) in a way that I can see people not familiar with typical soft-balled IT marketing speak could have misinterpreted.

"SMB2 is the new default protocol for sharing files in OS X Mavericks. SMB2 is superfast, increases security, and improves Windows compatibility. ... SMB2 is automatically used to share files between two Mac computers running OS X Mavericks, or when a Windows client running Vista, Windows 7, or Windows 8 connects to your Mac. OS X Mavericks maintains support for AFP and SMB network file-sharing protocols, automatically selecting the appropriate protocol as needed. ...

The Apple Filing Protocol (AFP) is the traditional network file service used on the Mac. Built-in AFP support provides connectivity with older Mac computers and Time Machine–based backup systems."

https://images.apple.com/media/us/osx/2013/docs/OSX_Maverick...

> ... converting/reformatting 300 Tb of disk is a lot of work, but it's probably less work than contacting Apple ...

He didn't know what the problem was, so he contacted Apple support. Unless you are suggesting he reformat 300tb everytime there's an issue, I don't see how he could have avoided contacting Apple in his situation.

> ... and whining about how Apple is going down the drain in a public blog post.

I don't see why the author can't voice his opinion on his personal blog. You've voicing your appreciation of specific Apple products in a public forum.

> Apple themselves state that AFP is deprecated. Deprecating isn't the same as removing. Unless Apple has previously said they are removing (or will remove) AFP support in High Sierra release notes, or otherwise announce it somewhere, I don't see how it's the user's fault at all.

1. I'm suggesting that if you are hosting and supporting your own physical infrastructure and file hosting solution, your first interest should be solving the problem, and as far as I can see Apple gave this outdated setup the best solution available. You can either whine about it or accept the fact that you did not keep your house in order and do whatever it takes to solve it for your users and customers.

2. Of course anyone can voice any opinion about Apple in any forum they want, but I'm not the one with a current ongoing issue that I know a solution to but am choosing not to implement to the benefit of shaming Apple in public instead.

3. You are correct that deprecating isn't removing, but when you're in a niche market (Apple servers), using any setup that includes deprecated protocols or components is a bad idea, and you should know this and plan for it if you elect to roll your own.

The person who wrote this post seems to have a very entitled sense of what he as a customer deserves in terms of continued software support from Apple, and very little sense of what in turn his customers and/or colleagues are entitled to and should expect from him/those that are responsible for keeping their business critical solution working.

The real problem isn't the server. SMB client implementation in MacOS sucks rocks, period. He could use any high end top of the line cutting edge server hardware, and he would have the same problem: MacOS doesn't know how to move data quickly through the network unless you're using AFP. Or eventually, NFS, but then the Finder behaves erratically.

Easy solution: Don‘t use Macs as servers, they are completely the wrong tool for the job.

> Nowadays, I purchase iMac18,2’s to realize it has no thunderbolt2

Another tip: Read up on the hardware you are buying. I can‘t believe this guy... buys hardware seemingly without even looking at the specs and then complains.

Apple sold these Macs as servers ... I even had one that was a 1U rack-mount unit. Businesses (unless you're a hosting service) generally depreciate server hardware on at least a 5-year schedule so it's no surprise to me that these are still in the field. It would be very interesting to know exactly how many are still out there running!

You are absolutely right. Apple quit the Server buisness when they killed the Xserve, and relying on Macs as servers is really a bad idea given their support interval for macOS. Seems the author is very angry that his next server cannot be a fancy Apple device.

I'm easing into it :-)

I'd suggest looking at a replacement solution around an open filesystem. Then, even if the world moves on to other technologies, you still have the source you can update or even hire someone to maintain.

>fancy

You don't need this part. Per the article, it is a brand the author knows and has used reliably in the past, and has some knowledge of - you don't need to belittle his choices as fanboyism.

I build Linux servers, and the problem lies with the MacOS SMB client. It sucks so bad you gasp for air when in the same room, seriously. See my other comment: same hardware, running MacOS: SMB 150 MB/s, AFP 1GB/s (yeah, that's more than 6 times faster). Same hardware running Windows on SMB: 900 MB/s. All from the same server and same disk share, of course.

Apple scheduled AFP for deprecation with OS X 10.9 in 2013 - 4.5 years ago. There has been plenty of time between then and now to prepare for this.

Dear author,

Do you have an unexpired warranty? A support contract guaranteeing that certain features will continue to work for the length of the contract? If so, please have your lawyers get in contact with our legal team.

Signed, Fake Tim Cook

P.S. Next time, either make sure your asses are covered, or use FOSS instead, so that in the event that public maintenance is no longer provided, you still have the option of forking the codebase and hiring an engineer to do whatever maintenance you need, as opposed to using our closed, proprietary product, where you are now SOL.

First of all, let me tell you that I feel your pain. Having something that behind the curtains 'just worked' in your toolbox stop working and being forced to spend time on it while you could be doing productive work just sucks.

We all do it. We tend to focus on the troublemakers, dousing their frequent little fires, and just forget the 'good' guys that are chugging along day after day without a hitch. Is this the hardware equivalent of 'technical debt'?

That lovely machine of yours is 10 years old. In our industry that does most definitely make it qualify as an antique (unfortunately not with the 'antique' valuations, going by the ~400$ these go for on eBay). Over those ten years, you've gotten good mileage out of it, but it looks like those maintenance free years are now taking their toll. Time to bite the bullet and find a new solution. even if they would revert on AFP, it would just be a temporary stay of execution. And I wouldn't look at Apple for this tbh. Servers and workstations have long Ceased to be 'mainstream' products for them.

Apple doesn't take the professionals seriously. Yes, they supposedly shifted from AFP to SMB in 2013 but ever since, SMB performance of Mac OS hovered between laughable and abysmal. And it's not so much the SMB server, but the SMB client implementation that sucks.

On gigabit ethernet, AFP as well as NFS on Mac reach easily 100 MB/s, while SMB hardly passes 50 MB/s. On 10GigE it's even worse: AFP, 1GB/s, SMB 150 MB/s. Testing on a Hackintosh, the same hardware that hardly passes 150 MB/s in SMB reaches 900 MB/s running Windows 10.

SMB on MacOS is a bad joke for everyone needing to move big amounts of data. NFS works OK, but alas, the Finder has (many) bugs and some things don't work well (refreshing, icons, etc). AFP is still by far the best solution.

I've found that tuning some of the SMB options ("man nsmb.conf") has dramatically increased my recent Macs' SMB network throughput and the responsiveness of finder in huge/deep shared directory hierarchies (due to caching).

I don't know if any of those apply to using OSX as an SMB drive host, but they might.

None of this should be taken as disagreement with your post; just ideas for improvements if you find yourself thus frustrated again.

Add to this the macOS Server components will be significantly deprecated soon:

https://support.apple.com/en-us/HT208312

Oh God. Just... wtf. Apple really is trying to get out of anything that's not iOS?! macOS Server is the equivalent of MS's Small Business Server package, and now they're, essentially, removing core features out of it.

My first thought was along the lines of "what's left in macOS Server to make it a server?".

I think the Update Cache and the MDM will still be there. Other than that I can't think of anything (major) that's not on that list. :(

Of course, both of those things are pretty big components of iOS management. Which reinforces the idea that Apple doesn't really care about anything except iPhones.

> Apple really is trying to get out of anything that's not iOS?!

Yes, this has been obvious for years now

MacOS had a Wiki service?!?

Apple has been removing features from OS X for a while now, and I’m sure it’s going to continue with their focus on iOS. Eventually we will all be running iOS on macs.

P.S: I’ve always stayed one version behind as I can get stuff done rather than spend time and energy fixing whatever Apple broke in the new version.

Anecdotally, I've been migrating macOS client networks away from macOS Server for a few years. The preferred solution is an HPE MicroServer with Linux (typically Ubuntu LTS) with `netatalk` and various trimmings (`avahi` etc). Bonus points are garnered for a Time Machine option, too.

The icing on the cake is the clients can also choose their "Mac" server icon for Finder:

http://simonwheatley.co.uk/2008/04/avahi-finder-icons/

This is also supported on FreeNAS and works like a charm.

I have a NAS with both SMB and AFP (netatalk-based) support, and despite Apple deprecating AFP, it's is still faster and more stable than Apple's SMB support. I guess the Apple devs who made that stuff in the 90's knew what they were doing...

> it's is still faster and more stable than Apple's SMB support.

Anything is faster and more stable than Apple's SMB support. MacOS is horribly slow and buggy with almost every modern NAS. Fixes include enabling SMB 1.0 (WTF?).

Cross post from elsewhere on this thread:

Apple's SMB support does indeed suck, but if you need to use macs as SMB clients, tuning the SMB client config ("man nsmb.conf") can dramatically improve performance! It's night and day for me when I force a newer protocol, increase the cache size, turn off signing/verification (if acceptable given my network conditions), and increase the async read/write counts.

Unfortunately, Gen8 microserver requires proprietary driver for it's RAID controller (for SATA-2!) and fan control; and Gen10 doesn't support hotplug for drives.

I wonder why there still isn't a company making highest quality products like Apple once did, while not being a total dick about customer support?

The machine seemed to work fine, despite its age. Why this smugness "Why don't you upgrade to/buy a newer inferior more expensive product which doesn't suit your needs at all?"?

Is a product obsolete as soon there is a newer product? Where is the line between products which provide infrastructure services and interchangeable consumer products?

I agree, as a video archivist, we support a lot of technologies that are much older than most computer systems. e.g. uMatic, half inch tape, BetaSP, Film etc. Albeit daunting, there is a slower decline in support for these machines.

Someone else pointed out that Apple announced the deprecation of AFP 4 years ago - it’s a bit unfair to say Apple caught you off guard here.

As someone else pointed out, the deprecation of AFP was announced over 4 years ago, 8 years into this servers lifespan (it’s a 2006 model, right?). This was a problem their network administrator should have been aware of for 4 years, that’s probably enough time to figure out a solution!

Those aren't hardware-focused companies, but I'd say Adobe and Autodesk do quite well on maintaining high quality products (and, like Apple does, being pricey)

> Is a product obsolete as soon there is a newer product?

If it's an Apple product, yes. Apple only makes money when they sell hardware. Several years ago they made a conscious business decision to quit supporting old hardware the way they once had because it didn't make money. Now the only purpose of Apple software is to push more hardware sales, which is why the MacOS and IOS upgrade cycles are so frequent, why there has been a continual decline in Apple software quality, and why you see things like the IOS battery scandal that slows down older devices.

Clearly your server is obsolete ... and this describes why my last ever Apple purchase was in 2012.

I bought an iPad 1 ... it's still a wonderful piece of hardware and I (try to) use it for web browsing nearly every day. Unfortunately, Apple realized that it was under-powered just after the iOS 5 release and it hasn't gotten updates since early 2014. It works fine except that it also doesn't have a browser that will run current protocols and standards. Sigh!

Planned obsolescence is cancer. It's not just Apple that does it though.

Planned obsolescence is literally the opposite of cancer.

It's an expression of speech for distaste of the malignant practice of making something redundant even if there was no need to do that;

Much like even if a human soul would rather not perish but yet there was a class of malignant diseases that will make it.

> Clearly your server is obsolete

The server works as well as it ever did. The problem is entirely in the software used on the client machines. That's not what "obsolete" means.

Sorry ... I should know by now that sarcasm doesn't work well on the Internet. But the truth is that the server is de-facto obsolete if there are no clients that can talk to it. Hopefully my second paragraph is clear that I don't appreciate planned obsolescence.

Your server ran 2018 software back in 2012?! That's amazing!

It is. With a design focused company, everything you do on the system is thought through and designed — your copying of a file is part of a user story that somebody wrote.

Unlike Microsoft, Apple tends to focus on one or two ways to solve a problem. Generally speaking, if you’re having trouble doing something on a Mac (as a user) and it seems very difficult, you’re off the path and would be better to start over and rethink what you are doing.

When you’re doing something like relying on a deprecated protocol that Apple barely gave a shit about, you’re on thin ice. At best there are a couple of engineers and a strategic customer who care about the feature and it will sort of keep working. If not, you are out of luck.

FWIW, you can't share APFS formatted volumes via AFP at all. AFP is probably not something you should depend on going forward.

What I find odd is that no analysts are asking Tim about the direction Apple is taking wrt. to build quality and developer support.

To me, Apple seems to have chosen a direction that will push developers over to other platforms. This will not happen overnight, but I suspect that in 2-3 years time the guys who have to develop on Macs will be the ones groaning over their OS - not the Windows people.

The combo of too expensive hardware and low build quality on OSX + bad DX is Apples biggest threat at the moment.

Success hides all problems. Didn't Apple just end their largest quarter to date? What would analysts be upset about? What would Apple think their problem is when they're making more money than ever.

Perspective is important here. I agree generally with this whole "Macs going downhill" theme, but this isnt new, and the market doesnt really care. I meant the infamous "Apple has lost the functional high ground" article was from beginning of 2015, after months of this sentiment going around last time https://marco.org/2015/01/04/apple-lost-functional-high-grou...

I think the same thing. Abandoning servers as it's not their main product is fine, but what about all those developers who want to build a basic CI system? For Android it's simple - a bunch of cheap Dell servers with Linux or something. But iOS? Most teams are using Mac Minis as servers, but they haven't been updated in years and are dual-core only, so they're not great options.

Past a certain number of OSX instances needed for testing, a lot of shops virtualize. The licensing can be a hassle, though. I suspect a lot of places that do this wouldn't pass an audit.

It will push OSX developers to other platforms. OSX likely serves the purpose (for Apple) of a Mac hardware value-add plus a revenue funnel into iPhones/iOS. At worst, OSX/Macs are basically a glorified iOS device dock. Even if the worst isn't true, though, there's very little incentive for Apple to spend resources on enticing developers onto OSX; they don't make much money/prestige off of it, and I'd be willing to bet that OSX app-store-distributed payware apps' revenue to Apple is a tiny, tiny fraction of what iOS apps get them.

TL;DR they need to attract developers onto iOS, not OSX, and developers are already plenty attracted to iOS because money.

How long does it take to reformat 300TB of disk? I guess the most time it takes is the restore from backup, because that seems to be the safest and easiest way to do this. But I might be wrong here.

I can see the problem here, and it's annoying that Apple doesn't handle this better, and doesn't give you any insights whether they will solve this or not. But I'm pretty sure that reformatting solves this problem for him a lot faster than waiting for a bugfix.

To move 300TB (say, restoring a backup) at an optimistic 300MB/s would take just under two weeks.

Can you not use OSX to perform the conversion for you in place?[1] There should also be a `diskutil` command that does it as well. Then you get snapshots and better volume management.

As always do your research first, this conversion is a one-way thing.

[1] https://datarecovery.wondershare.com/apfs/how-to-convert-hfs...

My information is that reformatting HFS+ to APFS is only a matter of seconds or minutes because what is mostly replaced is metadata (not the real data). So going from Sierra to High Sierra should be no problem with reformatting, but going back is a big problem then ;)

"And I have one particular server that we have loved for a long time."

Yes, for over 10 years now! (I looked up the manufacturing date from the serial number, week 29 of 2007)

I feel for your acute pain, but I think you got your moneys worth and then some out of that system, there should be plenty of saved budget to replace that with a new storage server. One that has support from the vendor.

:-) I know, but isn't it sad. I'm forced bu Apple to switch to another server OS if I want to keep using the 300TB SAS storage.

*by

10 years old? Seriously, if the server is that important to your business, you can afford to replace it after 10 years!

Hi All,

I've read all your comments and agree that a migration towards newer hardware and OSS Server software is needed. I've taking some first steps away from OSX Server and have come to the following results regarding the issue's I had. See below a comparison between old OSX server and new Ubuntu test server with regards to High Sierra:

AFP Finder search on OSX 10.6.8 Server from High Sierra Client (17D47): No result (most often) -or- Incomplete results (certain mounted shares are not searched)

AFP Finder search on Ubuntu 16.04.3 Server from High Sierra Client: Fast result Opening file works

SMB Finder search on OSX 10.6.8 Server from High Sierra Client: results show after 7 seconds Opening file fails "The alias "<filename>” can’t be opened because the original item can’t be found.

SMB Finder search on Ubuntu 16.04.3 Server from High Sierra Client: Fast result, Opening file fails "The alias "<filename>” can’t be opened because the original item can’t be found.

AFP Finder search on OSX 10.6.8 Server from Sierra Client: Fast result Opening file works

AFP Finder search on Ubuntu 16.04.3 Server from Sierra Client: Fast result Opening file works

SMB Finder search on OSX 10.6.8 Server from Sierra Client: Fast result Opening file works

SMB Finder search on Ubuntu 16.04.3 Server from Sierra Client: Fast result Opening file works

Some speed indications: SMB file copy from Ubuntu 16.04.3 Server to Sierra Client: 20.91MB/s 21923002.62 bytes/sec AFP file copy from Ubuntu 16.04.3 Server to Sierra Client: 49.19MB/s 49459712.12 bytes/sec SMB file copy from OSX 10.6.8 Server to Sierra Client: 48.98MB/s 51399308.67 bytes/sec AFP file copy from OSX 10.6.8 Server to Sierra Client: 64.58MB/s 67710528.40 bytes/sec

Sorry to hear this. But it sounds like careless system management. I work at a lot of video shops and none of them are even on Sierra, let alone High Sierra. Update one laptop and test. Apple has put AFP on the backseat for several years now (since they switched the default in Mavericks), in favour of SMB2.

I admit, I usually let my users upgrade their clients as soon as they wish. This is the first time it broke something. :-(

I run a nearly identical server at home (the Mac Pro that is, not the massive RAID).

I made the decision to install FreeBSD when I set it up and haven't regretted that for a minute. The hardware is actually lovely, right down to the literal nuts and bolts.

Just stay away from the Apple software ecosystem and you'll be fine.

I'm so happy I'm still on Sierra. They were so busy on the new filesystem, that they broke so much other stuff. But hey, at least if you don't like an OSX version, now you can just wait a year and you'll get a new one! Which may or may not have more bugs.

Quite tangental to the article, but:

This is why you shouldn't name protocols/standards after products!

I've worked with a few know-nothing users trying to slog through mac file sharing for the first time and got confused when I told them that the better/faster/better-supported sharing options, between macs, were the ones without "Apple" in the name.

"But it's named after Apple; it's going to be the most official/best integrated thing if I'm on a Mac!" is a common barely-technical-user refrain.

The same thing applies to the "Apple Partition Map" bootsector option when formatting disks.

Give things a descriptive, concise, memorable name that has nothing to do with a brand. If you deprecate them, you'll be glad you did.

/pedantry

Going to throw my unpopular opinion out there...

High Sierra has been nothing but a smooth update for me. APFS has corrected several external disk issues I've had. All in all, it has been a solid, if small, update, similar to the other releases like Snow Leopard and Mountain Lion.

I'm quite interested in migrating to APFS, I just can't find any hardware that can connect to Serial Attached Storage.

I'm assuming you didn't read the article since it is about the Apple Filing Protocol (AFP) and not APFS.

It's not that your opinion is unpopular, but that it's completely irrelevant.

idorube

You can take the SAS cards out of your Mac Pro and put them into a Thunderbolt expansion chassis, and keep using the same disk array. No need to reformat or convert to APFS.

And you can export your disk array via SMB2 or NFS to the other Macs.

It's a pity Apple don't make a decent Mac Mini anymore which could serve as a decent server. Maybe they'll release a new one some day.

Meh, if your servers are valued that much, you should know better.

I even know that Apple is not investing in servers anymore and hasn't in years. AFP was also abandoned.

They should've planned for this and switched to some more solid server configuration like Linux.

My advice: Build your own workaround / software solution for this problem with a future opt out option to replace the server with Linux and Samba (yes I know Samba is a pain but it is the standard nowadays).

A HW question: how are all these cards attached to the system? IIRC the cheesegrater Mac Pros only have 4 PCI slots.

SAS can be daisy chained.

Article says "bunch of 10G ethernet cards" and "some SAS cards", though.

What's a computer?

Only thing that catches my eye is, only 16GB ram for 300TB storage?

Relatively low number of clients with relatively sporadic access to archives which are rarely accessed?

Doesn't surprise me.

I don't find that surprising at all: these are very large files, and they aren't probably not scrubbing around in them back and forth but streaming them off the disk in a single pass; so you could have 128GB of RAM and you are unlikely to get much caching on the files themselves... the only thing that matters is the directory index structures, and I wouldn't be shocked if they needed more than 16GB for that (because these are large files).

I was with C-COR/ARRIS when we bought N-Cube for their VOD streaming technology (cable television). You'd be very surprised at the difference between the amount of content stored and the 5 minute buffered content currently being streamed. I'm not surprised at the amount of memory OR that the server isn't sweating under the workload.

I guess now that clients can't attach, it's technically in the warm-down phase of its workout.

I noticed 100MB/s.

You don't need 10GbE and Cat6 for that. 1GbE and Cat5/5+ are fine. The SOHO NAS from Synology or QNAP with ARM CPUs, 1-2 GB RAM and 1x 1GbE ports are achieving such performance.

That's what caught my eye too. 100MB/s is nothing. The author claims that he's getting half that with SMB, which means something is horrifically broken with his system.

Anything worth using as a file server should have no trouble doing 100MB/sec with pretty much any protocol.

No, something is horrifically broken in Apple SMB client implementation. That's probably why they never actually phased out AFP, because it's the only way to move data fast on MacOS.

Paradoxically, the speeds that I'm getting from my rMBP13 for SMB transfers are comparable over Ethernet (both Apple Thunderbolt adapter and rangom assortment of TB2 docks) to Linux and Windows machines, but over Wi-Fi, rMBP is much faster than any Linux or Windows machine I have. The Mac machine is 802.11ac 3x3 MIMO by Broadcom, while the others are only 2x2 by Intel, but I'm not sure that the speed difference could be explained just by this factor.

Yes, there were small bugs in the past, like waiting for some timeout when browsing shares on the server when the auth is done via Kerberos, but they were ultimately fixed.

This will hit everyone using non-iOS Apple products either sooner or later:

You're no longer Apple's priority. You haven't been for some time. MacOS isn't getting major updates, the development team has been largely disbanded: MacOS is a niche platform that only exists to develop iOS apps. One day it won't even do that.

Apple already markets and wants iOS to replace your laptops, and doesn't significantly care about the server or workstation markets.

If you rely on MacOS you need to think about this, and the longer you delay it the more it's going to hurt.

The iMac Pro is the strongest evidence that this isn’t true.

It might be nice if Apple did something more to convince us they care though. Make a strong statement? A commitment to quality?

Is it though?

It seems that the Pro is designed to appeal only to the faithful and have a very limited lifespan.

Putting everything in an integrated system with limited to no expansion possibilities rules out a lot of potential users (anyone with common sense) and means that if one component fails or becomes obsolete the whole thing is useless. If there were awards for excellence in planned obsolescence those systems would be a shoe in.

It means none of those things and it is totally baffling to me that you refuse to accept that other people may have different views on these systems' utility to them and that they aren't idiots without "common sense" for that.

A single computer release doesn't make up for:

- removing features and essentially deprecating OS X Server: http://www.zdnet.com/article/the-death-of-macos-server/

- letting the Mac Pro languish for _four years_ without a single update or mention until recently

- the Touch Bar, which clearly shows they put some work into the technical aspects, given how it integrates with included Apple programs and thorough developer support. But they seemingly never talked to any actual users or developers in usability testing, given how well it's caught on.

- ignoring the keyboard quality issues in their "Pro" laptop line

- replacing included programs with crappy, feature-poor replacements (e.g. iPhoto to Photos, Disk Utility, etc)

The evidence seems to indicate that Apple doesn't care about the Mac and hasn't since 2011 or so. If they've started caring again, good, but it's going to take quite a few sustained Mac releases over the next couple of years for me to take them seriously again.

Not sure about that. The trash can mac pro seemed like that at the time too. Let’s see when they’ll abandon this toy.

My MacBook Pro 2016 keyboard finally failed beyond “usable” level and Apple has been keeping it in the shop, refusing a warranty repair. Another MBP we have started having issues with the keyboard.

I believe they are going downhill fast on the Mac front, both hardware and software.

What? In what way does releasing a new iMac reform the MacOS team, or undo Apple's iPad Pro messaging?

Did you even read the comment you're responding to?

Are you also going to buy the new Blackberry smartphone?

   > MacOS isn't getting major updates

There is new version each year. And if you are talking about features, what feature would you consider "major update"? New filesystem maybe?

  > the development team has been largely disbanded

Source?

  > only exists to develop iOS apps

Source?

  > Apple markets and wants iOS to replace your laptops

Source?

But I agree that they probably do not care about server markets.

I don't have the time to locate all those source reference requests, but I know I've seen stories related to each point here on HN recently.

Seems fairly common knowledge that Apple have disbanded macOS team and is focusing on iOS based future. Am I mistaken?

>New filesystem maybe?

You mean the one that shipped on iOS first and benefits iOS the most? To me, that shows how far the Mac has fallen in Apple's priority list.

> There is new version each year. And if you are talking about features, what feature would you consider "major update"? New filesystem maybe?

As a Mac user, it seems like the last 4+ years of MacOS updates have added nothing but greater iOS integration, bugs, and misfeatures.

> > the development team has been largely disbanded

> Source?

Bloomberg.

Apple re-organized its software engineering department so there's no longer a dedicated Mac operating system team. There is now just one team, and most of the engineers are iOS first, giving the people working on the iPhone and iPad more power.

https://www.bloomberg.com/news/articles/2016-12-20/how-apple...

> > Apple markets and wants iOS to replace your laptops

> Source?

apple.com, major print media, and all Mac Stores, which have ads talking about why iPad Pro are laptop replacements. Here's one: https://www.youtube.com/watch?v=sQB2NjhJHvY

It sounds like you have very little involvement with Apple products.

  >> the development team has been largely disbanded

  > Source?

https://www.theverge.com/2016/12/21/14037686/apple-macbook-m...

TL;DR :

> In another sign that the company has prioritized the iPhone, Apple re-organized its software engineering department so there's no longer a dedicated Mac operating system team. There is now just one team, and most of the engineers are iOS first, giving the people working on the iPhone and iPad more power.

If you’re going to call BS, please just call BS. Everyone knows what “source?” means.

I think most people who follow Apple already know this and those who don't could obtain sources easily - when sources are provided they're simply downvoted.

The whole thread has developed into Apple fans downvoting everyone who criticises the new approach to MacOS.

Reply on news.ycombinator.com