Before there was no such switch in front of the "ring of buffers," you need to show us where did TiVo say before there was this switch (or "valve") in front of the "ring of the buffers."
Ok, it was shown as a valve, not a switch. You discuss it below, so I think you understand what I mean.
Don't make things up when TiVo did not say so, the judge will not hear your made up switch. The functions of the "ring of buffers" were never modified in any way, they were there before and are here now, not changed.
I am neither making things up nor quoting TiVo. I am explaining how the thing works based on my reading of the transcripts, patent applications and discussion threads. The ring of buffers was not changed, but it IS being used in a new way simply because the valve upstream of it is no longer there.
The "valve" TiVo talked about, was there before to control the flow of the parsed start codes when they got filled into "a buffer" by the "source object." That "a buffer" was "obtained" by the "source object" from the "transform object" and used to store the parsed start codes and streams.
No it controlled the video and audio data stream not just the index file. I believe that is a mistake, not making stuff up, but if repeated, we will know it is a fabrication. The buffer obtained is a member of the ring now instead of the intermediate buffer that was disabled. It is still obtained by the source object, as the direct buffer descriptor. It stores streams.
And in the old design, right before this "a buffer" there was a "valve" which TiVo in its graphics used it to depict "automatic flow control" function. That valve turned off if the "a buffer" got full, so to avoid data "spilling onto the floor." None of such had anything to do with the "ring of buffers" while the "ring of buffers" was minding its own business, both in the old design, and now in the new design.
The valve was related to the removed buffer. I called it a switch above, same thing. It turned off if the ring of buffers was about to catch up with itself. If the ring of buffers were to catch up, the data would "spill onto the floor". Spills are messy, a switch allows for the turn off to occur exactly on a start code. That way if data loss occurs, it can be more cleanly managed with the boundaries of data loss set to start codes. So yes, it has everything to do with the ring of buffers, they are the ones that will overflow if not switched off. In the new design, the ring of buffers is much more carefully managed to make sure it doesn't spill by giving the draining process a higher rate or priority. It is not as clean as a switch, but serves the same function, just in a more complicated way.
And the TiVo expert now admitted that "valve" (i.e. automatic flow control) was removed, the code was gone.
Given there are no start codes to use to decide when best to open and close the valve, letting an overflow spill over is an eminently practical way to control the flow. Removing the valve frees up the resources needed to more carefully manage the ring of buffers.
If a software engineer here has a different take, I would much prefer to hear it. Please tell me how allowing uncontrolled ring buffer overflow or preventing it via carefully tuning process priorities and rates is not a form of data flow control. I am interested in a technical, not a legal or "word" driven discussion.