Apple's MacOS Server has a 1.5/5 rating on the Mac App Store
apps.apple.com
apps.apple.com
And that's the key-word here: appliance. A NAS is a standalone box that receives automatic updates, is 100% remote-management enabled by default, can be easily reset to factory settings, and can't be messed up by employees who think it'd be neat to run local native apps on it, "since it's a computer just sitting there."
Plus, since a NAS isn't going to be doing any desktop-OS things, the software for it can be slimmed down enough to run on lower-specced hardware, making the appliance itself much cheaper than the sort of machine required to run Server.app smoothly.
No SMB office-manager who needs this kind of functionality these days would buy a Mac or PC, to set up server software on it and leave it sitting there headless in the office. They'd buy a NAS.
And, IMHO, that's why Apple haven't kept Server.app up to date. There's just no demand for it any more. Not because of the cloud (which is mostly a complementary use-case) but rather because this software-solution vertical has now evolved into a turn-key hardware-appliance solution vertical. And it's one with slim-enough profit margins, that Apple doesn't want to compete.
This seems to be a primary driver of the software-rot found all across Apple's ecosystem. iOS is generally pretty well-maintained; macOS less-so. Apps like Music are kept in better condition than, say, Podcasts. Etc. You can almost directly correlate how unpopular a piece (or feature) of its first-party software is with how buggy it is. Which I guess makes some amount of sense, but it really tarnishes Apple's image as a company selling high-end products.
Also Apple: We don't want to invest as much in parts of the ecosystem that have relatively low returns.
There's a lot to like about Apple, but their closed ecosystem and adversity to consumer repair rights ensure I will never buy one of their products.
In software, on the other hand, the latest macOS update introduced a bug in Notes where now if I hit the Tab key (entering an actual tab character), that whole line of text becomes invisible. Just invisible. I can highlight the text and see that it's still there, but without highlighting, it's invisible. It's very reproducible, this is not an obscure corner case. And this is at a time when Apple is attempting to ramp up its online services like Notes.
But putting that aside, I still see that as an execution flaw, or at least, it's a way in which design and execution are so intertwined that the distinction no longer makes sense. A computer with a non-replacable keyboard that breaks after a year is a badly-executed product. The way to execute it better would be a keyboard that doesn't break, whatever design decisions that entails.
So what you're saying is it was a design flaw from a design company.
1) Enable dark mode
2) Open a new note
3) Tab twice, then type text which will be invisible
It appears to be any line with two or more tabs of indentation. Not two or more relative to the previous line, but two or more period.
The biggest part of their business has shifted from computers to phones but I doubt they expect the iPhone to be their biggest product in the future forever, so I don’t think they would pigeonhole themselves when their name and logo is so ubiquitous.
Companies like Apple, Nike, Mercedes, Coca-Cola will never completely change the name (maybe some variations here and there, without touching the core name) simply because the brand is so recognizable. It would be throwing money and image down the drain and it would just cause confusion.
In Germany both are known. Daimler the business, Mercedes the car...
Kinda Apple and iPhone...then Mac (aka Mercedes Benz Trucks?!)
Another interesting case is when HP decided to spin off their PC business: they gave the HP brand to that new company.
I wonder what results in these dissimilar approaches? Are these cultural? are these youthful company versus decades-old bureaucracy?
Also, I wouldn't call Google a youthful company, even if Apple is 44 years old.
Synology actually encourage this. In particular, the Docker package is great and all sorts of uses have cropped up. There are some gnarly rough edges that I have encountered but that package and others make the unit very useful. I’m a home user so it’s a different use case, but my 918+ is probably my favourite machine ever.
The popular ones for macOS are Jamf and Mosyle. There are some other ones as well that are more cross-platform oriented, but both Jamf and Mosyle more or less cover the features that macOS Server provided for administration of a set of devices.
Tbh, I don't think Server was popular as a NAS service ever. The only thing I've ever used Server for was managing devices.
Their Synology Drive app is a drop-in Dropbox/GDrive replacement too. Works great on Mac at least. Haven't tried Windows yet.
We maintain an official ZeroTier package for it too, so it works anywhere.
We've seen this a number of times with macOS, and some other Apple adventures. Apple will put out something, then if a third-party offering comes along that is equal or better, it cancels the Apple product expecting everyone to switch to the third-party solution.
Off the top of my head, Aperture comes to mind. But I've had this thought before and came with several other examples at the time.
Apple also works the opposite way. If third-party offerings for a particular function are OK, but not great, it will make its own version. Stocks, and several other apps for iPad come to mind.
When the first happens, people complain that Apple is fickle and abandons its customers. When the second happens, people complain that Apple is stifling competition.
Apple is either more schizophrenic, or less predictable, than other big tech companies.
1) Way better battery life. Until the other browsers give a damn about battery life I won’t even touch them!
2) Cross-device syncing. My passwords and my tabs are always available on both my iPhone and my MacBook. Everything is synced through iCloud. I go back and forth so often every day that I could not see myself using a different browser on my Mac.
Really Server.app would run fine on the same slimmed down hardware, but Apple doesn't make that. The cheapest Mac is what, $800 for a Mini? $250 in PC hardware would more than suffice.
If you're on an even tighter budget you hardly need to buy new hardware at all, just set aside one of the old machines the next time you do a desktop hardware refresh and install any of a dozen free Linux or BSD NAS solutions on it.
Most NAS fail at one or more of these.
In addition 90% of them are very bad about security, and putting these kinds of devices on your company or office network transforms it from an internet access network into a garden for insecure appliances to be coddled, a stepping stone that must be vigorously firewalled, scanned, and monitored.
https://www.ixsystems.com/truenas
Also, it's Open Source.
That’s a strong statement. Care to articulate? Server.app used to have services for Calendar, Email, Contacts, Wikis, a profile manager (basically an MDM server), Open Directory, VPN, DNS, DHCP and, well, a web server. Would a Sinology really cover all these features?
Qnap isn't far behind neither: https://www.qnap.com/en/app_center/?top_place=1&qts=4.4.1
Image silent data corruption via bitrot can be frustrating. Without a regime to prevent it, it spreads into all backups. And typical workflows depend on backups for eventual migration to new storage which allows any corruption to be replicated via the backup strategy. You end up backing up the corruption, unwittingly.
Which was one of the big problems with the original G4 Xserve. They added it later in the G5 model, but I suspect much of the damage had been done. Most technical people, just ignored it after discovering that it didn't have ECC and the disks were low end ATA rather than SCSI. With the cluster/HPC version the stories changed to how unreliable the machines were.
At that point it seems only the diehards were willing to trust their data to apple server hardware and it died a couple years later.
How do you think I could discriminate between file corruption and the user making an edit to a file?
(Perhaps if the mtime and filesize doesn't change, but the SHA does?)
The location for metadata depends on the application. For DNG workflows, the metadata is a separate location in the DNG file, with separate metadata and data checksums. For other workflows, the metadata is in a sidecar (a per image file), or stored in a database managed by the application.
We are in between the legacy 'directory' based networks with authentication etc, and the more robust and expansive beyondcorp type setups. Do you really still need classical 'user' and 'group' membership things? Network accounts? Local web servers? Pretty much anything of significance is in a private or public cloud, if only because the fast pace and scaling. All that is left is basic file sharing that is sometimes done locally, and even that is becoming more and more stupid to implement in a workflow or business process. (regardless of your OS or vendor)
There are of course still some true file-based processes where Apple is still used a lot, but even that is no longer local-file-only; a lot of the still and motion processing is done either super small locally or simply distributed to dedicated systems like render farms. And for the 1-man-photo business all of the local stuff still works and you really wouldn't have gotten anything out of 'server' anyway.
The most problematic holdouts are the likes of Adobe who refuse to write applications that play nice with NAS-type filesharing so that just leaves you with file syncing instead.
What do you think is providing the authentication and user management in those private clouds? OpenLDAP, FreeIPA, ActiveDirectory, Keycloak?
And even when you move everything out of the office to a datacenter you still need all normal IT stuff that you needed in your little office rack just in the datacenter now.
The only part of the "old-guard" IT stack I see disappearing is the office file server.
SAML. Yes, the provider's own network probably hosts a directory service as the default SAML backend, but for integrating with enterprise clients who already have their own directories? SAML. Now, rather than making ACL assertions about local users, you have to make ACL assertions about what ACL assertions each foreign directory is allowed to make on behalf of its own "guest" users in your environment.
If you're, say, Box, you're implementing only the endpoint, and mostly don't need your own directory, because all your customers bring their own. If you're, say, IBM, you're implementing only the raw directory, but rely on cloud providers to actually interpret it, rather than hosting any of your own Intranet. And if you're Joe SMB, then you're implicitly using GSuite as your SAML-exposed directory through OpenID Connect, and you don't need an IT department for this at all.
(And even if you're an enterprise who also writes your own cloud services, then this is still a change for you, because your endpoints are probably now written "generically" with a SAML interface—treating your own directory as just one among many, talking to it through a public exposure—rather than interfacing directly to it as a privately-configured PAM binding on your servers et al. In other words, now your services [and the departments providing them] are decoupled from your directory, because they both just see each-other through the lens of SAML, and so don't know that your company's employees are any different than a client company's employees.)
For some categories of applications, it'll always be much more performant to have local servers, even if the network is 100Gbps symmetric fibre.
Everybody talks about capacity elasticity, but easy geographical distribution is the much cooler property of public clouds IMO. I could drive to a local colocation site and maintain my own boxes if it came to that, but having dozens of datacenters around the world would be physically impossible for me otherwise.
Of course your applications have to support that, but if they can be distributed across in-office servers then they already do.
We need a buzzword for these. "lolDAP?"
Did I mention no CALs?
The move to the cloud has largely made identity management and directory services more widespread, not less.
2) every business is going to have internal applications wether bespoke or something on top of ERP/CRM software so the same applies here.
3) when you use an app (either professionally or personally) you need to have an exit strategy, you’d think Oracle’s latest games would teach everyone that. All these SAAS apps have nothing, so self hosting with directory/file servers is the way to go.
3) MS and Apple OSes really need some of these things to function in a corporate environment. Unless you’ve got everyone on Linux or are doing BYOD then you’re going to need directory servers to coordinate IT.
We really had no choice but to migrate onto other products and these days even hard core Mac folks use AD for directory, Exchange for email and 3rd party products for device management.
> Device Enrollment: Enable FileVault support on first user login
> Send all traffic through VPN
Great work Apple!
> Release date: Apr 1st 2020
Oh dear, Nevermind.
No magic left in there at all.
Ended up dropping it and using the functionality of Airwatch.
If you pay for Apple TV+ / Apple Music / Apple News and then cancel, you can keep using it through the end of the period you paid for.
That mostly leaves enthusiast and small office scenarios, most of which can get by just fine with "non-server" macOS. After all, they've always literally been the same operating system. My Mac mini is a headless media and file server and has, in the past, been the printer server as well -- and all of that's enabled just by clicking boxes in the "Sharing" control panel. (Heck, I could even turn on "content caching" to have it keep a copy of all macOS/iOS software updates on it, which would be great for a small-to-medium office.)
However, it's easier for me to buy some canned solution; either pure software, an SaaS, or a hardware solution.
The only thing I know is that it had a really nice icon and box art.
Shared local storage, like for video production houses. At the time, it was ludicrously fast.
It also worked with Windows, Linux and Unix machines. Pretty much anything that supported Fibre Channel connections.
At the time is was useful, as the Mac desktop was my daily driver. Unfortunately current macOS versions are really sluggish, so its currently a home "server" running Fedora 31.
The 16GB RAM limit is becoming a bit of a problem, so I will likely replace it with an Intel NUC.
First, even though there were quite a few decent-sized Mac Mini deployments like you're describing, I'd be genuinely surprised if that accounted for a decent chunk of Mini sales overall. Some, sure! But most? I wouldn't expect it.
Second, does any of that require running Server? We have a database and application server in the office here, but it's not running Server: it's just plain macOS running daemon processes.
Why would people buy that over the iMac for daily use?
Also, I bought a Mini for my desktop because it's tiny, cheap, and powerful, and I can use my own screens.
Apple just doesn't seem to care to put in the effort. I don't think their neglect of MacOS Server has anything to do with trying to protect their cloud offerings. Even a wildly successful MacOS Server strategy wouldn't be that much revenue versus some of their other lines, but it could provide real value and help to certain kinds of businesses and users.
Macmini and OSX Server were products for niche markets that “second Apple” is completely disinterested in.
Think "apple id/icloud in a box".
That would be actual privacy.
I understand this isn't a remotely core part of their business, but I don't understand why you would want such a quarter-assed product out there with your name on it.
A small team, which Apple can afford, could keep this a good offering that helps out certain kinds of Mac users.