LCD screen manufacturing process simplification?

[Cybersoya]

Hello

I am no expert so i apologize if the answer is trivial.I was wondering, if they can manufacture a transparent transistor on every (sub)pixel of LCD glass screen, why are they using "COF (Chip-On-Flex) or COG (Chip-On-Glass construction)" to connect every (sub)pixels' tiny wire at the end of row/column with LCD driver chip? Wouldn't it be much easyer to align neccesary tiny wires with connectors, if (non)transparent demulitplexer circuit was photolithographically "etched"(at the same time as pixel transistors are etched), so that connetcing wires leading to driver chip could be less tiny, less in number, and so easyer to connect (with less throwaway LCDs)?
I am only talking about physical connection of (individual column/row) LCD glass wires -to- (silicon)chip wires, i know that further down only few signal wires are needed to drive LCD.Thanks for answers.

 

[phyzguy]

The transistors that can be manufactured on glass are of low quality. They are adequate to turn the pixel on and off, but they are too slow and too weak to do the demultiplexing.

 

[Cybersoya]

Thank you for answer.

That's a pity. Wouldnt it be nice, if they were searching for innovation in this direction, maybe to "evaporate" or "spray" extra layer of some substance on edges of LCD screen, that would enable them to put real/fast transistors on top of it.
For instance, I've heard that in Japan they managed to put entire Zilog Z80 microprocessor on a transparent glass substrate, albeit it was bigger than ordinary Z80 die - but "macroscopic" size of transistors wouldn't matter for LCD screen if they were put at the edges, i think.

http://www.z80.info/sharp/z80_glas.htm
https://original.sharpmz.org/z80glass.htm

 

[phyzguy]

That's a pity. Wouldnt it be nice,

If wishes were fishes...