awesome!
Your modules make VCV Rack for me.
I haven’t posted much in the last few weeks, but I did work pretty intensive on my codebase. I have several modules in the pipeline, some of which are almost complete, but I am not so good at actually finishing things and getting them released… 
My plan was to publish an update to the VCV Library in the end of July, which should include SPLICE-KIT, but I got a bit obsessed with automated testing and fixing bugs no one ever reported… and then a question on MIDI-KIT came by, a module which I ignored for a very long time. I started re-working its complete code and scripting API for two or three weeks and I am finally happy with it, so it might arrive in the VCV Library sometime in the future…
What I was working on:
- SPLICE-KIT
- TRANSIT-PAD - an expander for TRANSIT with an X/Y-pad, which allows morphing between up to 8 snapshots (similar to Phase-mode). I teased this a few times already.
- TRIO - the third module which completes the trilogy of MAZE and HIVE… you probably cannot guess its shape…
- MIDI-KIT
- patch & selection browser for MB
- A dozen more modules, which are work-in-progress and in concept-phase
Still getting a handle on my windows rack setup and workflow. Ive been editing my template, and since the switch to windows from mac I’ve discovered more of a case for caching my module browser (using the deep cache module), unfortunately for me, this means i cannot use MB, as one has to be “owner”.
Is there any chance MB might ever have a caching option, or something like it for improving browser load times?
Cheers, and thanks for all you do for the VCV ecosystem!
I saw this module and I’m not convinced on it. Are you saying that your module browser is working faster because of this module?
It certainly does on my Windows computer. There’s a slight freeze when modules come into view with the default module browser and MB, and Deep Cache fixes that, since it caches the renders to disk.
(The main issue with Deep Cache is that it renders everything at like 4x res and takes up a an obscene amount memory and VRAM because of that 
It also needs its own browser replacement to do its thing)
I’ve added my own kind of “pre-warming”, for testing: when MB’s browser is open, the off-screen modules are rendered in the background while it is idle. Nothing is persisted to disk, restarting Rack starts this process from the beginning. On my 7-year-old Surface Laptop it takes ca. 15s to pre-warm ca. 800 modules.
Try it out and let me know what you think…
Edit: There is a new option to enable it…
It doesn’t feel very effective on my machine. It seems to take too long to prewarm, and doesn’t do anything while I’m scrolling the browser. (which the code seems to indicate is intentional?)
It also doesn’t really help with freezes when using the search.
Does it only prewarm previews while the browser is visible?
IMO the main reason Deep Cache works so well is that it does store the previews on disk.
Here’s my specs, for the record:
CPU: Ryzen 7 5700x @ 3.2 GHz (4.6 GHz boost)
RAM: 64GB DDR4 @ 3600 MHz
GPU: Nvidia RTX 3060 12GB (latest drivers)
Storage drive Rack + user folder are on: WD Blue SN580 2TB NVMe SSD
OS: Windows 10
VCV Rack Free 2.6.6
Yes. Sorry for asking in advance, but you did enable the option for it?
Appreciate you picking this up! if you mean pre-render previews when idle, yes, it does not seem to behave as deep cache does. Behavior on my machine can be described as an initial delay before pulling up the browser, then it seems to be fairly unresponsive, almost like it does the pre-warm for modules only after they are initially shown in the actual browser. So effectively it acts like it hasn’t cached when I search for something, or whenever the modules on the browser change (or reveal modules that have not been seen yet?), i.e. scrolling, UNTIL I’ve done it and waited the first time?
What I can tell you of deep cache is it seems to take a bit to load on load, but then the browser is instant and responsive on any search etc. Hope that is helpful.
My Computer Specs:
CPU: 11th Gen Intel(R) Core™ i7-11370H @ 3.30GHz (3.30 GHz)
RAM: 32 GB
GPUs: NVIDIA GeForce RTX 3050 Ti Laptop GPU (4 GB), Intel(R) Iris(R) Xe Graphics (128 MB)
Windows 11 pro
VCV 2.6.6 pro
I forgot to reply to this, but yes, I did.
Testing here with the latest nightly, it seems to take less time if you’ve already let it prewarm on a previous run of VCV, most certainly due to OS file caching.
I’ve got 2681 modules according to MB.
It took about 1:18 minutes to prewarm the whole list on the first run I timed, but this was after I had already run VCV without timing it. About the first third or so of the bar went faster than the rest. (makes sense, that’s about where I closed VCV on that first untimed run)
Doing it a second time took about 37 seconds to prewarm the whole list.