Yes. Each time it will report that it is finished, but when I look at the device’s Calendar, the event is there, but its alerts are not, or are incorrect. Sometimes, just adding an irrelevant change to the desktop Calendar event, will finally get the event to sync accurately.
1 Like
I wonder if macOS Sequoia developer beta 1 fixed this. I doubt it. If I had another Mac and iPhone to test with, I would. Anyone here want to try it? If it is still there, then please report it ASAP!
I didn’t think I could download it, but it looks like I can, so I’ll let you know.
Just a reminder that, once you get Calendar working as intended, there is an archive function available in macOS under File/Export.
Same for the Contacts app.
2 Likes
It’s still there… sigh…
1 Like
Will_B, please report this bug again. If others can do the same, please do so. The earlier and more people, the better chances Apple will fix it.
Done.
However, I have reported this since the first public beta of Ventura. I verified it with every public beta and release up until the last couple of Sonoma betas and releases. Between installing and testing, each release required a couple hours of testing and trying to guess at what was the core problem, then writing up a Feedback report (with logs). The behavior varied in bizarre ways. There was a run of releases that actually crashed SyncServer. I was actually encourage by that. Repeatable crashes tend to get attention, but no…
In late January-February, I spent three full afternoons with an Apple support person. He insisted that if he was to help me, I had to install Sonoma on my internal drive. After every test, reinstalls were necessary. Ultimately my internal drive was so frelled (and one of two local backups corrupted), that attempted restoration resulted in the embarrassingly stressed-out public and private discussion I had at the time with Mr. Bombich and his good folks. Data was lost.
The Apple tech said engineering wanted more logs. I guess they really didn’t pay much attention to all those Feedback reports.
I’m done. A DayTimer seems an easier solution than coding my own Calendar/syncing app. I don’t have enough pockets for a DayTimer and an iPhone. I don’t think I would miss my iPhone much…
1 Like
Hey there Will,
Sorry for all the grief you have been through, with respect to Apple Calendar (and possibly Apple Contacts, as well). It sounds pretty bad to me.
I can say that, I have switched to BusyCal, and it is much better than Apple’s Calendar ever was. I still haven’t committed a to server/sync scenario, and only (currently) use BusyCal on one of my Apple devices (a Mac-Mini).
In the future, I may incorporate BusyCal in a server/sync situation.
1 Like
I’ve been reading on Eclectic Light about Linux VM’s on macOS. I wonder about the feasibility of running a CalDAV server under a Linux VM?
I haven’t found any current CalDAV package that looked like it could be installed reliably under macOS.
I see now that MacPorts has a CalDAV package (finally). Has anyone had any experience with it?
Absolutely. Use UTM to run a Linux VM at boot (auto-login required), use the macOS hypervisor and memory balloon device, and you’ve got yourself a nice and lightweight environment for running any Linux software you want, including a vast array of useful and up-to-date server software. My preference is for Cyrus, for IMAP, CalDAV and CardDAV, but other, lighter options are available. Do it. macOS can only get you so far by itself, because it’s not a platform of choice for servers (though it should be!). I wish Apple would change this.
Which package do you mean? A quick search turns up the old calendar-contacts-server package, which is just Apple’s server, now very definitely long in the tooth, reliant on Python 2.7, PostgreSQL and an entire mail server stack. I’d not advise going with that one, no.
Yes, looking closer, MacPorts is just building the old Apple code. And if I’m reading the page right, it hasn’t worked for anyone in awhile…
“any Linux you want” is too many to choose from these days… ;~}
Last time I actually installed it was on our sons’ first edition iMac.
Well, so I did it once… I can do it again.
1 Like
Debian Stable (currently Bookworm). That’s what I’m using for servers. Nice package manager and (optional and very safe) automatic upgrades. Grab the ISO for Arm64, do a minimal text-based install, and install and use just the software you want with apt. Being in a VM gives you a tremendous advantage because hardware compatibility is the single biggest reason Linux installations fail or go wrong, and in a VM you don’t get that.
Do you have to log in?
VirtualBox documents how you can have it start VMs at boot up independently of any particular user’s login. Instructions are provided for Linux, Solaris, macOS and Windows.
See also: VirtualBox manual: Section 9.21: Starting Virtual Machines During System Boot.
The VirtualBox mechanism for macOS uses a launch daemon. VirtualBox includes the following (org.virtualbox.vboxautostart.plist):
<?xml version="1.0" encoding="UTF-8"?>
<!DOCTYPE plist PUBLIC "-//Apple//DTD PLIST 1.0//EN" "http://www.apple.com/DTDs/PropertyList-1.0.dtd">
<plist version="1.0">
<dict>
<key>Disabled</key>
<true/>
<key>Label</key>
<string>org.virtualbox.vboxautostart</string>
<key>ProgramArguments</key>
<array>
<string>/Applications/VirtualBox.app/Contents/MacOS/VBoxAutostartDarwin.sh</string>
<string>/etc/vbox/autostart.cfg</string>
</array>
<key>RunAtLoad</key>
<true/>
<key>LaunchOnlyOnce</key>
<true/>
</dict>
</plist>
To use it, you would copy this file to /Library/LaunchDaemons and change the value of the Disabled key from true to false. At system startup, it runs a (provided) shell script that starts the app and any VMs defined in the /etc/vbox/autostart.cfg file.
I skimmed the UTM documentation and did some web searching and it appears that this capability is not bundled with UTM. It seems to me that you should be able to do this, but it may require a bit of work to figure out how to make it work reliably.
1 Like
Thanks for that. I’m not sure what state VBox is in these days for Arm64, though unfortunately the (QT) accessibility story is rather sad. However I would honestly have preferred to simply start and monitor a process under launchd that exercised virtualization.framework directly, something like “vmcli”, rather than going through a UI and all that gubbins. Sadly despite looking I have yet to find an official or unofficial means of doing this that is of the same quality as UTM, and I don’t think there is a way to do this without a front-end at all. So we’re left with either rolling your own qemu invocation, which is certainly possible but which would not give you the full feature-set of virtualization.framework, or using UTM to configure the VM and then executing the embedded qemu binary externally and not having control over it in the UI due to the way the GUI interacts with qemu over a socket. I think it might be easier to just build and use qemu in headless mode and have launchd run that process under your user account. I’d love to be wrong about this but I don’t think there’s really a solution that provides the best of all worlds here. Good on VBox for baking this in as a feature.
Where do Howard Oakley’s “Viable” & “LiViable” fit in with VirtualBox and UTM?
I suggest you forget VirtualBox for M-series Macs at this time. The only versions that are anywhere near close to “working” are developer previews that are reported to only run Intel operating systems and don’t even do that very well. I’m reading about performance and stability issues.
You might want to look at VMware Fusion on Apple Silicon. It runs many Arm64 Linux distros very well, and has a CLI command that can be used to start VMs programmatically and run them “headlessly” without having to open the Fusion GUI. And for personal use, the price is right (free). Perhaps these capabilities could be sufficiently scripted to do what you want.
1 Like
Howard’s Viable and Liviable are wrappers around Apple’s high level Virtualization framework. UTM has integrated the Virtualization framework on Apple silicon — it’s an option for Linux ARM VMs and is what’s used to enable macOS virtualization.
I’m not sure if either of Howard’s utilities are scriptable or can run headless VMs.
I agree, Fusion is awesome, however there is no support for memory ballooning. Depending on your workflow this may or may not be a big deal, however especially on RAM-constrained Silicon Macs and for primarily server roles, I’d say it’s worth having.
Well, I wish I had that week back. Two VM apps, and lost count of CalDAV packages, and nothing to show for it. I can’t remember a Linux “package” that didn’t miss at least one dependency (that depended on something, that depended…). Linux remains a pathetic excuse for a GUI OS. If ALL the (who knows how many flavors) gave up their egos and worked together, maybe “Linux” would actually be a “user’s” OS. Not holding my breath.
By the end, I was editing Python code install errors, WTF?