FUN FACT: Amstrad was the last company to release a cassette computer (464 plus).

[ChaRleyTroniC]

Prodatron is spot on - the problem with the PCW16 was the terrible (and slow, but generally all round terrible) operating system. There are some in-jokes in various issues of BTL about the word processor's lack of scrollbars, I think. 

The hardware was fast. 8-bit CPU wasn't the issue. But there was a foulup with the mouse/serial interface and some weirdness around reading/writing DOS format discs, which by then was already essential. 

At one point Amstrad got in touch with me because they wanted someone with Z80 experience to finish/rewrite the spreadsheet app. I went down to Brentford one day to talk to them. But it would have meant giving up my (first) fulltime job for something with uncertain prospects. In retrospect I'm very glad I turned it down. 

It could have been a success if it had been project-managed right. But Amstrad lost that knack after their early PCs. 

[Prodatron]

Hi @ChaRleyTroniC (CRTC), thanks for this insights, and cool that you are still here sometimes! I heard similiar things from Rob Buckley/Trebmint, who was directly working for the PCW16 project.
Yes, it seems, that the mouse is not a normal serial one, which you could use for the PC in these days - for the NC100/200 as well. It needs to work in a special mode unfortunately. At least the keyboard was a standard PS/2 keyboard (hope so).

[ChaRleyTroniC]

Oh, and I mean Brentwood. Not Brentford. Other side of London. 

I remember arriving at Brentwood station as a fresh-faced 21 year old and trying to get a taxi to the Amstrad offices. The taxi driver laughed at me. "They're right there, mate." Directly opposite the station...

[Prodatron]

Always nice to see, that we old Amstrad users have been produced nearly at the same time

[Anthony Flack]

I have to assume that "the knack" was Roland Perry. By rights the CPC464 should have been a catastrophe. 

[Shaun M. Neary]

I have to assume that "the knack" was Roland Perry. By rights the CPC464 should have been a catastrophe.

I'm not so sure I'd fully agree with that given that the C64 was still being produced with 64K of RAM up until the early 90s while the Speccy's and Amstrads had moved to 128k in 85 and 86.

The big selling point for the CPC464 in my opinion was the fact that you also got a monitor with it in most cases, which means it could stay in a bedroom or office without the television being hijacked. Growing up, I saw so many arguments in houses when it came to delegating time to hook up their Atari/C64/Speccy and yes, even Amstrads with modulator because they wanted to see the games in colour.

Maybe I'm biased as a 464 guy, but from what I saw, and maybe this was just a Dublin thing too, but 128k and disc drives weren't the be all and end all during the 80s.

[Anthony Flack]

I mean, if you read the development story of the CPC464 it was heading towards being a train wreck, until Roland Perry stepped in and somehow turned everything around within an impossible deadline. 

[Gryzor]

Well yeah, but that was more of a project management thing, I think.

[eto]

Roland Perry saved it from project management side but the engineers saved it from the technical side.

Afaik it was supposed to be in the VIC20/Speccy feature range. The original idea was to have 8 colours - basically what the Speccy has. I am 100% sure we would all not be here if it wouldn't have been for the genius engineers that combined those simple, cheap parts in a really clever way to build something that is much more capable than what you would have expected from the specs alone.

Too sad there is not too much information on how the technical side happened. I am pretty sure there was a magical night shift after which the Arnold became a (C)olour (P)ersonal (C)omputer.

"oh it is so boring to build a computer based on these parts."
"yeah, seriously, who wants to have only 8 colours these days?"
"hey, look. If I keep the RGB pins floating and add the resistors here, we can have 27 colours, not only 8"
"oh nice idea. But we don't have enough bandwidth to get the data and ensure RAM refresh at the same time."
"well, damn ... but... if we rearrange the screen addresses the CRTC will refresh the RAM and the GateArray can get enough data"
"ah, brilliant. Let me think... I think we then can also have 80 columns if we do it like that?!"
"oh - oh - right - niiiice. Hopefully we get a good BASIC so this all makes sense. Oh... just realised ...if we have 80 columns and a Z80, couldn't we run CP/M?"
(phone rings)
"hey guys, Locomotive here, can we come over? We also bring Pizza and then we can show you what nice Firmware and BASIC we have for you! Fits into 32K so no biggy."

[Anthony Flack]

It could have been a success if it had been project-managed right. But Amstrad lost that knack after their early PCs.

Project management yeah, that what I'm saying. Project management on the CPC was a disaster until Roland Perry stepped in and saved it. By rights Amstrad's attempt to break into the computer market should have ended in failure.