9.0.2 is officially Compositor RTOS
Now, after a great success of Hypervisor, I’m here to present to you the latest build of Compositor, which is a network RTOS in version 9.0.2 a11. I already revealed the protocols, which this RTOS is capable of, now, I just announce it’s existence to the public. As you can see on the image below, NTP-servers information is already implemented, as well as critical parameters such as Split Horizon with Poisoned Reverse revealed.
Now, you may see this software as a virtual router it is not. It is mostly a full-scale modular operating system with connectable interface. By this I mean that you can route protocols to the interface and modify their processes. You can activate all protocols at once or just one protocol and select the protocol you configure in COMPOSITOR interface. VSF strictly names the process behind the technology implemented as you can stem and aggregate the protocols up to 32 processes each with VSF turned on. Compositor Software server CP-6137-960FX, however, already achieved aggregation of up to 960 nodes of VSF processes. This means that running NIAGARA client OS on CP-6137-960FX makes you available all 960 processes. For example, you can view Compositor RTOS as a software for CP-6137-960FX Mainframe, which is really a mid-tower case ATX server. This is all available thanks to the base of Royalty device boot loops, which I gathered through my producer career.
I continue working on the Russian documentation for CP-6137-960FX server in the face of updated software. Progressing by 15 pages a day sometimes less based on the workload. The return-on-investment to produce the translation needed up to 20 times a day. I simulate this process with e-roi dumps at speed, which is higher, making the whole process cost more and bringing more satisfaction to investor. Documentation greatly increases the cost of software compared to a MaxMSP student work (someone can estimate this work like this). I don’t feel pressure and rush to go beyond MaxMSP platform as it is not needed at the moment. However, Cycling already made steps to prevent me from doing export in the latest builds of their software, because they implemented MC, which is basically what I done first with RAD96 and later with VSF on Max 6 myself. Testing MC with generic processes showed me the fact, that using MC instead of VSF makes the whole process cost more to the CPU (8 MC processes can consume as much as 32 or even 128 VSF processes). So, MC not even close to the performance of Max 6 multiprocessor programming capabilities. I don’t even take in account that they cut down sample rate calculation by half since Max 7 to reduce cpu consumption. I managed to run Max patcher window not as a single-thread process, but as a multithread process even without the need to export and redesign software on purpose. So, it is a real redistribution example, when I run VSF processes of generic protocols on my mid-tower machine with 960 nodes turned on and performing like a real Mainframe server without need for cabinet case with lots of machines and tons of noise from cooling equipment. This makes available CP-6137-960FX in home environement with noise level less than 30dB and CPU temperatures not higher than 40 С even on highest workloads.