VMD-L Mailing List

From: Gumbart, JC (gumbart_at_physics.gatech.edu)
Date: Thu Aug 29 2019 - 17:03:28 CDT

Wow, I knew it was going to be bad, but not this bad. Thanks for your efforts. I may need to look into installing Linux as well on my laptop.

Thanks!
JC

> On Aug 29, 2019, at 4:24 PM, John Stone <johns_at_ks.uiuc.edu> wrote:
>
> Hi,
> This is on my TODO list. I have to build a new FLTK, new Tcl/Tk
> and various other libraries, and in particular FLTK and Tcl/Tk
> have to be custom-patched to make them work. Since that work is
> not really VMD-specific, I'm leaving to the end since VMD itself
> has compiled in 64-bit mode on MacOS X for many years, but always
> with issues from FLTK and Tcl/Tk GUI code (not ours). This is really
> going to hinge entirely on the stability of the GUI toolkits.
>
> A much more serious question about the long-term viability
> of MacOS X now rests on Apple's announced deprecation of OpenGL
> in favor of their proprietary (read non-portable) Metal graphics API.
>
> There are some open source efforts underway to try and make OpenGL
> code run on MacOS X through an emulation layer on top of the
> proprietary Metal APIs, but so far ther're in their infancy.
> Another option would be to run a Vulkan renderer on top of MetalVK,
> a similar emulation layer that sits on top of Metal. That is more
> apt to be a successful route, but requires VMD to have a completed
> reimplementation of the entire interactive rasterization pipeline
> in pure Vulkan, and I'm still only in the early stages of writing
> that code since Vulkan itself has been coming along a bit slower than
> I had originally anticipated.
>
> You may have also noticed that Apple is also dropping support for OpenCL,
> and they haven't had CUDA-capable GPUs in their machines for years.
>
> To be honest, the future direction of the MacOS X OS platform is
> starting to look like a poor match for scientific visualization
> codes going forward. It may be that booting Windows or Linux on a
> Mac ends up being a better way of utilizing the hardware to its
> potential given what's been happening with Apple deprecating
> the major cross-platform rasterization and computing APIs.
>
> Anyway, one issue at a time. We've only recently gotten a Catalina
> test box setup to work on ourselves, and we also discovered that
> basic things like NIS/NFS etc don't work like they used to either,
> so even our usual in-house development approach for MacOS X, with
> use of network-based filesystems is problematic going forward.
>
> Best,
> John
>
>
> On Thu, Aug 29, 2019 at 07:53:05PM +0000, Gumbart, JC wrote:
>> Hi John et al.,
>>
>> The next MacOS is going to eliminate all support for 32-bit applications. Will there be a 64-bit VMD compilation made available for download in the near future, or should we compile it ourselves? (I have no idea how difficult it is to compile VMD on Macs!)
>>
>> I???ve already run into one person running a beta of Catalina who can???t install VMD. They were thinking of putting Windows on their Mac as a work-around???horrifying...
>>
>> Thanks!
>> JC
>
> --
> NIH Center for Macromolecular Modeling and Bioinformatics
> Beckman Institute for Advanced Science and Technology
> University of Illinois, 405 N. Mathews Ave, Urbana, IL 61801
> http://www.ks.uiuc.edu/~johns/ Phone: 217-244-3349
> http://www.ks.uiuc.edu/Research/vmd/