BC on ARM CPU (Apple M1) Macs

[ohtheguy]

Beyond Compare 4.3.7 (current version) is only available as an Intel CPU Mac application. However, it will run on ARM CPU (Apple M1) Macs using Rosetta 2 translation.

A native ARM CPU version of Beyond Compare is in development but doesn't have a scheduled release date.

Edit: A native ARM (M1/M2) Mac version is planned for Beyond Compare 5, release date not set.

The majority of Mac updates and releases, including Ventura OS, have been made.
Could you please provide us with a release date for BC5? That's because we, the users, are not in favor of running non-native apps on our machines due to battery and energy over-consumption.

[Aaron]

Hello,

It's something we are actively working on, but are holding off on announcing a timetable until we have something closer to a deliverable. Because of a combination of our upgrade policy and licensing, we've had experiences in the past where announcements have led to purchasing decisions for the future version rather than what was currently available. I understand the frustration of not having a roadmap, but we want to be sure any purchases now are for the current release and feature set. I would suggest subscribing to the News and Announcements sub-forum for email notifications to get a ping when we have something concrete to share.

[tcorbin]

Hello,

It's something we are actively working on, but are holding off on announcing a timetable until we have something closer to a deliverable. Because of a combination of our upgrade policy and licensing, we've had experiences in the past where announcements have led to purchasing decisions for the future version rather than what was currently available. I understand the frustration of not having a roadmap, but we want to be sure any purchases now are for the current release and feature set. I would suggest subscribing to the News and Announcements sub-forum for email notifications to get a ping when we have something concrete to share.

Aaron, this is an understandable approach for those contemplating new purchases, but what about for folks who have an unusable BC because of the massive memory leak leading to a system freeze issue on M1? Syncing small data sets doesn't help, as the system still completely freezes after ~10 minutes, even with doing multiple small syncs. I had to hard restart five times in just one day this week and still didn't get everything synced. Should we just seek out another solution and abandon BC, since Scooter doesn't want to announce a BC5 release date?

[Aaron]

Hello,

We're generally not seeing that level of instability. Are you performing many small but concurrent actions? My suggestion isn't to reduce your workflow until it is a single, tiny operation, but to start smaller and single task to test the stability, then slowly ratchet it back up to find the largest single task you can perform successfully and repeatedly. I think with some troubleshooting, we could get it to be a bit more stable and useful for you. I would not recommend multiple concurrent tasks, given that's usually the easiest way to trigger the issue.

As for existing customers, we certainly understand the frustration and really want to get something out. Since a lot of it is interconnected, we have a few things we really need to hammer out before we're able to announce an alpha or beta, but will as soon as we're able. We are planning on pushing out a technical test before a full release roll-out.

[tcorbin]

We're generally not seeing that level of instability. Are you performing many small but concurrent actions?

Aaron, as an example, two days ago, I was syncing video files (.mov). Even if I did them in batches of 7-10 files (each being 1-5GB in size), I could only get through about 3-4 batches before BC would beach ball and the entire system become unresponsive. This happens in about 10-15 minutes of use. Hard power off is the only solution. This behavior is consistent on a 2021 14" MacBook Pro M1 Max. I have ~300 files to sync, I can't do them one at a time — even small batches is a big productivity loss.

As for existing customers, we certainly understand the frustration and really want to get something out. Since a lot of it is interconnected, we have a few things we really need to hammer out before we're able to announce an alpha or beta, but will as soon as we're able. We are planning on pushing out a technical test before a full release roll-out.

While I know you all are working hard on this and it won't change anyone's mind, I think it is fair to point out that Apple silicon has been out for a LONG while now — two years. It's obviously frustrating to have paid for a product that is completely useless and way behind what would be considered a typical lag in development time for new architecture.

Thanks for your candor, it's appreciated. I wish there was some clear workaround for now, but I'm not sure what it is, if anything.

[Aaron]

Hmm, that's the second report we've had with unexpectedly bad stability on the M1 Max chip with Rosetta applications. I'm searching a bit into it now and one of the first hits I found was about some migrations or updates causing issues and suggested to try reinstalling rosetta2 with:
/usr/sbin/softwareupdate --install-rosetta --agree-to-license

If you give that a try, then try BC4, then throw in a reboot and retry BC4, do either of those trials exhibit better stability?

[kfriis]

Hello,

We're generally not seeing that level of instability. Are you performing many small but concurrent actions? My suggestion isn't to reduce your workflow until it is a single, tiny operation, but to start smaller and single task to test the stability, then slowly ratchet it back up to find the largest single task you can perform successfully and repeatedly. I think with some troubleshooting, we could get it to be a bit more stable and useful for you. I would not recommend multiple concurrent tasks, given that's usually the easiest way to trigger the issue.

As for existing customers, we certainly understand the frustration and really want to get something out. Since a lot of it is interconnected, we have a few things we really need to hammer out before we're able to announce an alpha or beta, but will as soon as we're able. We are planning on pushing out a technical test before a full release roll-out.

Maybe, that is because - for instance - I cannot be bothered to send endless crash reports on a well known issue, when a new version 4 update is released. It works on Intel Macs, but mine are on the way out (one is from 2015, the other from 2018) and day-to-day work happens on M1 (2021) or M1 Pro (2021), where BC 4 never survives for long without crashing the systems completely (panic followed by reboot).

You are of course free, to decide to ignore your customers reasonable requests for a tentative timeline. It's not that they haven't shown extraordinary patience in this matter.

I have invested in other solutions, so version 4 will only continue as long, as my Intel gear sees any use (the 2015 model is only existing as a spare in case of trouble on an Intel platform).

I'll repeat tcorbin's question:

Should we just seek out another solution and abandon BC, since Scooter doesn't want to announce a BC5 release date?

[Aaron]

Hello,

Sorry, this thread has gotten a bit difficult to parse and some of my generic comments are targetted replies to specific User posts within the thread.

To clarify, we are mostly on M-series hardware in-house, but still use older Intel hardware to test all of the currently supported macOS releases.
The comment you are quoting is specific to the M1 Max chip, which I had recently noticed a pattern of users reporting greater than expected stability issues on that specific hardware that we don't have on our in-office M-series hardware.

Your concerns are valid. We are *really* eager to have some users help us test a native arm build of BC5. BC4 is performing pretty simple transfer commands, but Apple's translation for those commands to ARM support is likely very complex and removing Rosetta as a factor would greatly help us troubleshoot BC bugs. We're in the early stages of sending out private invites for a Technical Test; we're not ready to label it as an Alpha yet and why we've been hesitent on releasing a timetable. I've been gathering almost every email and forum post to form a list. I also try to mention subscribing to the News and Announcements sub-forum for email notifications. As we're able to open it up to more users, it's one of our early resources for expanding our tester base. We're aiming to have a publicly available Technical Test or Alpha early next year.

We agree that the delay between the M-series launch and BC's native support for it has been too long. BC5 has taken longer to finish than we expected, and macOS has required more development resources than expected. We are working to reduce similar delays in the future.

It's also an important part of my job to help people get up and running with what we have currently available as soon as possible. While we've witnessed a level of expected instability when stressing a transfer, BC4 is usable and developed on M-series hardware with Rosetta2. This is why I still try to troubleshoot and gather information about BC4 and different hardware combinations, which seem to give different levels of stability on how far you can push it. Our developers and support staff have a varied set of M-series hardware they use, but not every combination so it's helpful to compare and contrast with what our customer base is using and reports.