I think OpenGL and OpenCL is already tossed out the window?
I meant "like OpenGL and OpenCL" in the general sense, not specifically. The "Open" in those two names. Also the significant amount of deployed code that depends upon those libraries.
As much as folks point to the Mac Pro 2013 'problematical hardware issues, Apple's lazy approach to making OpenCL 'first class' on macOS was also a factor. Metal eventually came and somewhat cover a subset of that ground, but it came so late as to do essentially nothing to help the momentum of the Mac Pro 2013.
It's still there in zombie mode, but Apple has made pretty clear it's going away soon.
Apple itching to rip something out when don't have a fully formed replacement is a problem. If kill off X-Windows then will loose folks who are in camp who were using macOS as a 'better' Unix workstation OS.
Apple is supposedly gearing up to eject scripting tools from macOS bundle. Python . Perl , Ruby etc. A Unix OS and there no Perl or Python ? That is really creditable? Sure folks will be able to use package managers to add them back in but it is yet another 'hole' that Apple is punting on. ( and Microsoft is heading in about 180 direction in terms of developer support with Unix like tools . ).
What's next ... they are going to start ejecting shells too?
Apple can narrow down to a niche where only care about XCode developers and that is it. Long term that isn't gong to help the Mac Pro market grow though.
[doublepost=1563403270][/doublepost]
I don't think the server variant is that important of a metric.
Exact same attitude Apple assigned to the Mac Pro for 2013-2017 period. Which is extremely indicative that their general attitude really hasn't changed much two the product. That more likely this 2019 model is far more a hobby project and that when Apple has finished it they'll go back to almost solely working on their "day jobs" full time..
If they weren't serious about they shouldn't have mentioned it in the first. Just throwing it out there for a dog and pony kicker again is highly indicative that their attitude toward this space really hasn't changed much at all. And that the same behavior is the more likely expectation.
Back in 2017 there were a lot of folks who expected the Mac Pro to come out in 6 months of less because Apple was going slap a somewhat available board into a new case and ship. Well this server model is about just that. They really shouldn't need a new logical board. All they really need is a new container and perhaps some fans with slightly different dimension. If moving the TB ports from the top to the front is throwing them for a long delay loop ....... that doesn't saw much for their collective organizational talent level to deal with future technological improvements in a timely fashion.
Seems like everything is on track. They haven't announced a date, but that's very Apple. We last heard September, and the last thing heard from the rumor mill is that China has already started production runs.
Rumors also said they were ramping up production on AirPower too. That working was just around the corner. Apple had real systems on stage at WWDC , but the rumors were that they were trying to ramp production ; not that they were in a viable production. I don't think it is ApirPower status but until it does ship in volume they still haven't done a real 'do' (as oppose to talk).
One hold up is that the new Mac Pro won't work without 10.14.6. There are features there for things like the Infinity Fabric needed.[/QUOTE]
Given Apple demoed the Infinity Fabric stuff in 10.15 at WWDC, this sound more likely like a backport hack than something going "according to plan". ( unless the 10.15 ones were rigged demos. )
Even if Apple manages to start shipping a small number of Mac Pros in September ( and generate another round of MP shortages stories like in 2013-2014 ) they would only filled in about 1 foot of an 6-8 foot deep hole they had dug themselves. Flaking in the sever model basically would put them at about zero progress by the end of 2019.
