VEPro 7 on i9 iMac 32GB RAM SSD latest OS and Logic versions, everything updated latest. Clean system, using defaults (other than enabling autosave, given the constant crashiness), fairly simple template with Synchron, Kontakt, and Aria players, no other plugs. This should work, and it kinda does, for a while, anyway.
But getting this template together VEPro (and sometimes even Logic) crashed about every half-hour, doing various things that you would think would work, kinda worked, seemed to work for a few times and then died of memory leak or whatever.
Software design looks about 2004 era. Syncrhon Player looks current. That may have been built anew recently with a contract team or the better front end people who dread working on the legacy codebase of VEPro. But VEPro is the essential piece the company offers and it just can't go on like this.
Suggestion is first to adopt American software discipline where if a feature doesn't work, it is removed. European software management typically prioritizes features over quality, this is a total mistake, as most features are gratuitous for most. Pick only the features a professional uses and delete the rest, as you simply won't be able to QA them given the enormous matrix of 2 OSes, 2 plugin formats, half a dozen hosts, and thousands of plugins.
For instance, get rid of the "coupled" mode and do everything decoupled with instance preserved as a pro template user will. Every "mode" you add adds to that matrix and swamps your QA. Have only one pro-level mode and face down the tears and shrieking internally and externally for the transition.
The thing crashes with simple commands such as Insert Channel Set or Load Channel Set from Disk. Remove these commands then. You can't make them work on all platforms, they get removed until you can.
Implement automated E2E testing across the entire matrix, operating at the GUI level. If you're going to blow off technical debt, you remove those features. You don't make the software a minefield, especially as your autosave sucks and is noisy (makes the dock icon bounce every minute! Does anyone there actually use this software?). Resist the urge to just spruce up the GUI, the most urgent issues are always the crashers. Software in 2022 shouldn't crash.
Implement stringent memory leak detection and other iteration testing. One of the nice things about the level of indirection VEPro offers is the ability to swap instances to swap libraries being used. I got this in my template but it only works 8 times or so back and forth until it just instantly brings down VEPro. That durability has to be tested with automation, again across the matrix including other plugin vendors.
I would probably do this on the existing code base, no matter how gruesome and outdated it is, before attempting a ground-up rewrite. As people are relying on this in the here and now and you have no idea when a replacement will actually settle into working no matter how optimistic you are now. The measures I'm suggesting are near-term as long as you can achieve buy-in down the org chart.
I will try to continue working with this; but people there ought to be humiliated at the state of their flagship software product. Single worst piece of code on my machine. If there was a legit competitor, the transition would not only be swift monetarily, it would likely take out the cash cow of the sampling business along with it. Pay attention.
