Mixbus Version Mixbus-5.3.4
I suddenly have the situation that there are signal in all my channels and they are assigned to busses, but there is absolutely no sound going from the channels to the assigned busses. The VU's in the Bus section absolutely shows nothing.
These are canned default profiles that is not changed. So the problem must have crept in with mixbus.
So mixbus shows signal in channels but when routed to busses, absolutely no signal.
All I can think of is that mixbus is heavily dependent on Python and since the snake recently crept in and seemingly updated itself and caused problems with another application. The snake also took spdif with it which worked spectacularly until it updated.
Any ideas how I can resolve this, and I want to know if there is any Python dependency in mixbus that can cause the channel/bus disconnect.
I checked with all three the commercial versions of Mixbus I purchased with known working profiles and the result is the same.
To ADD:
I tried to reinstall mixbus and I get this error.
"Architecture is x86_64
Checking for required disk space
WARNING: GCC4/5 libstdc++ ABI Mismatch
Mixbus was compiled with gcc4, your system uses a newer version of the
standard c++ library. Plugins on your system may not load or plugin-UIs
may cause crashes.
"
Part of the problem currently with mixbus is exactly that. Suddenly a lot of plugins do not load and it says so and list them when I open a known working profile.
This is truly weird. I did not install a newer version of GCC and there are no automatic updates I know of.
So mixbus seemingly has an EOL with libraries and this is within one Version of MX !!
It seemingly forces me to upgrade Mixbus again.
I would suggest you contact Harrison support: mixbus@harrisonconsoles.com
@zimbodel -
And, while you're contacting them, see if they have an image of 5.3.4 that was compiled with GCC5.
Most Mixbus releases over the past few years have been offered as gcc4 & gcc5.
If you've been bitten by some general computer OS library update, due to another app's installation,
the gcc5 version of Mixbus might get you back into a working condition.
However, and I speak from experience here... Be prepared for the worst.
That is - if indeed your OS has received a non-intended library update, then you might have to
re-install your OS to the configuration that you were running before you started having issues.
I spent the better portion of 3-4 weeks trying to isolate and revert a similar situation.
I finally said - "F-it!" and re-installed my OS to a more recent version and re-installed all of my
Mixbus releases. My platform has been solid ever since.
Cheers!
Patrick
thanks pburyk, I forgot that mixbus issued a gcc5 version. I have it still. I reinstalled the mixbus gcc5 version and it installed perfectly, but gives me the same trouble. It is clearly a python issue as the snake crept in and updated itself. I made a point of locking down my system and it has been working flawlessly for about 2 years running about any midi device and large interfaces I threw at it. Never could have such a rock solid system with windows.
I had to get rid of systemD completely to make everything work as well as limiting use of Python. Well the Python crept back and it is clearly the reason why things dont work as I have other applications that also started developing trouble.
If Linux can rid itself from systemD and the irresponsible Python development environment it will truly become an unbeatable operating system.
I yet have to find a systemic problem on Linux that is not either a Python or a systemD induced problem. The latter is seemingly trying too convert Linux into poetternix.
I also purchased the latest 32C Mixbus version last night and that runs worse!
Just out of curiosity - What distro & release of Linux are you running?
The systemd Wiki page is an interesting read. Includes a chart of Linux distros and their reliance on systemd.
[
https://en.wikipedia.org/wiki/Systemd ]
But, you sound like you're already familiar with it.
I don't know... pointing more and more to an OS reinstall.
(01-20-2022, 02:30 PM)PBuryk Wrote: [ -> ]The systemd Wiki page is an interesting read. Includes a chart of Linux distros and their reliance on systemd.
[ https://en.wikipedia.org/wiki/Systemd ]
But, you sound like you're already familiar with it.
I don't know... pointing more and more to an OS reinstall.
systemD is a huge mess and tries to take over every system daemon on linux. In the end all the system processes will be just systemD and that spells a lot of danger.
Just a simple example. If you have a lot of cronjobs running, then they run fine in sysVinit but if you boot with systemD most of your cronjobs dont work and does strange things. This created huge trouble for me as I rely a lot on cronjobs. As an example @reboot doesnt work at all so all your system setups after a boot suddenly is not performed.
This is just a small example. The list is huge.
systemD is slowly turning Linux into Win95 and it wont be long before you boot into a blue screen of death. I actually think the latter could be prophetic as systemD tries to become the only system daemon and therefore a single fail error you have no control over to mitigate. You have no idea how much work it was to get rid of systemD and still make everything work, but I did it .... until the damn snake came crawling in.
I will have to revert to an older system backup I have, but I have to move to a newer debian version anyway so I will just redo it. I have to as I need to document all the system admin how to get the snake and poettersys out of Linux for it to work properly as a sound server.
This is not a mixbus problem, but I am a bit alarmed that mixbus is seemingly relying on Python. I explicitly asked Harrison what Mixbus was written in knowing what the snake will do if used, and Python did not come up. well it seems python is used for the plugins and somehow to connect the channels with the busses in mixbus as the snake made those two inoperative.
(01-20-2022, 02:23 PM)PBuryk Wrote: [ -> ]Just out of curiosity - What distro & release of Linux are you running?
I use MX Linux version 18 currently. I use MX as they already went some ways to remove the dependency on systemD and it was a great starting point for me as they have a very clever way to use systemD applications by seemingly providing dummy systemD configs to apps without actually running systemD and then the apps miraculously works. I took it quite a bit further and painstakingly removed systemD completely together with a lot of DBus nonsense (talk about another boondoggle) and making sure python cannot upgrade and systemD cannot reinstall by pinning them in dpkg. Amazing how well linux works as a sound server if you get rid of all that interfering junk, but then you must be very good at sysadmin to get to that stage. It took me 6 months of sysadmin and reading up and custom scripts to get to that point. I tried to start from Devuan (Totally systemD free Debian) , but I found out quickly that I had to redo all the work the MX developers did, so I just moved to MX and took that further.
So I know what happened here. The snake found a hole to crawl back in and I am not sure how it got past me.
Trying to uninstall a python upgrade is like unscrambling an egg. It is surjective software. Good software is injective. Surjective software is basically badly thought out arbitrary junk, which is basically what Python is. Great to develop with but a nightmare for a user to unscramble when something goes wrong due to the idiotic surjectivity.
Again, this is not a mixbus problem per se, but a Pyhton problem and I am very amused that mixbus seemingly uses python to handle plugins and to connect channel data with bus data apparently as that is exactly what died in mixbus when the snake crawled back in.
So my beef with mixbus really is the seeming use of Python. They must be calling python scripts and that is what I want to figure out so I can mitigate future issues regarding Python and see if I cannot create a sandbox (python snakepit) for mixbus so that mixbus only relies on the snake species it needs and is independent from system python upgrades. that will take care of this nonsense. For that of course I would need to know where mixbus calls python and if I can alter the headers about how mixbus calls python.
So if mixbus can confirm my suspicion that python is used in Mixbus for connection between channel data and bus and for handling plugins, then I can put in the work to create a sandbox.
I will contact them and find out why Python influences mixbus so I can devise some method to prevent it in future.
(01-20-2022, 03:26 PM)zimbodel Wrote: [ -> ] (01-20-2022, 02:23 PM)PBuryk Wrote: [ -> ]Just out of curiosity - What distro & release of Linux are you running?
So if mixbus can confirm my suspicion that python is used in Mixbus for connection between channel data and bus and for handling plugins, then I can put in the work to create a sandbox.
[/url]
Quote:[url=https://discourse.ardour.org/u/seablade][color=var(--primary-high-or-secondary-low)]Seablade
[color=var(--primary-med-or-secondary-med)]May '20[/color]
[/color]Ardour uses Python as part of the WAF build system to build it. This has been the case since… oh probably A3 or so, and before that used Scons. It is not used in the operation of Ardour.
https://discourse.ardour.org/t/python-versions/104010
Mixbus/Ardour are using Lua as their script language.
MMM