From Wikipedia, the free encyclopedia

GA Review

Article ( | visual edit | history) · Article talk ( | history) · Watch

Reviewer: PresN ( talk · contribs) 23:38, 25 January 2024 (UTC) reply


Claiming this review; I'll start posting comments soon. -- Pres N 23:38, 25 January 2024 (UTC) reply

Thank you for taking the Herculean task of reviewing the mighty Spectrum, PresN! I have returned in good time, and will begin addressing your comments. I note that some of the technical sections have been altered or degraded since I have been gone - but will tackle that when we come to it. ♦ JAG UAR  22:52, 4 February 2024 (UTC) reply
Lead
  • It's odd for the lead to launch almost immediately into the clones, before discussing what the machine is or that it was released outside the UK. I would have expected that sentence at the end of what is now paragraph 3.
  • I've removed the mention of clones entirely and condensed the rest. I think it reads alright. ♦ JAG UAR  23:19, 4 February 2024 (UTC) reply
  • "The machine was the brainchild of" - oddly informal
  • Inclined to agree, though I think most know British English employs a more colloquial reading experience, even for Wikipedia. I've changed it to 'creation'. ♦ JAG UAR  23:19, 4 February 2024 (UTC) reply
  • The lead says it was released "in some parts of Europe as the Timex Computer 2048.", but the infobox just says Spain in 1985. Sounds like at minimum that should be changed to EU in 1985, unless you want to get more specific about the countries.
  • Yeah, that would be a nightmare. Changed to 'EU'. ♦ JAG UAR  23:19, 4 February 2024 (UTC) reply
In 1985 Spain wasn't in the then EC (the EU didn't exist under that name in 1985 either). Although I can't quickly find a single source which lists release dates in different European countries, I would have expected that it was launched earlier in e.g.
Ireland than Spain. 90.167.254.202 ( talk) 22:29, 29 March 2024 (UTC) reply
'EU' is generally referring to Europe; it is easier to cluster the releases that way. ♦ JAG UAR  21:56, 2 April 2024 (UTC) reply
  • Actually, now it seems like the issue is that lead is playing loose with whether the Timex versions were versions of the Spectrum or were clones of the Spectrum. It doesn't seem right to call the Timex a clone, then turn around and say the Spectrum was released in the US as the 2068. Personally, I'd stick with calling them versions or variants, seeing as they were official (slightly modified) Spectrums.
  • Agree. Since I removed the mention of clones above I hope it negates this issue. I'd call the Timex versions versions, and the Soviet imitations clones. Let me know how it reads? ♦ JAG UAR  23:19, 4 February 2024 (UTC) reply
  • Happier with the lede now! -- Pres N 03:27, 5 February 2024 (UTC) reply
Background/Development
  • This feels like its going way too deep into Sinclair's history. Nothing about Sinclair Radionics is directly relevant to this computer; its important to note they had made the ZX80 and 81, and had been making electronics since the 60s, but the details of his business history is too much.
  • I've cut down on some of the less relevant company history. I took inspiration from the ZX81's background subsection, which to be fair adheres much less to WPVG's manual of style. I think keeping the mention about Sinclair's relationship with the NEB is important, as his breakup prompted him to develop microchips, which ultimately led to the Spectrum. ♦ JAG UAR  11:51, 13 February 2024 (UTC) reply
Hello, Jaguar, and thank you for your interest in the article.
I didn't see this review on time, so I reverted your edit [1]. If you really feel that it doesn't belong to the article, then you can undo my revert. However, I personally find the remove sentence both interesting and relevant. Z80Spectrum ( talk) 16:00, 25 February 2024 (UTC) reply
  • "Sinclair resolved to make his own products obsolete before his rivals developed the products that would do so, thus seeking to make the technology as cheap as possible." - not following how wanting to make a computer better than the ZX80/81 leads to making the Spectrum (or is it computers in general?) cheap.
  • That doesn't read well! I cut the second half entirely. ♦ JAG UAR  11:57, 13 February 2024 (UTC) reply
  • "Architecture from the ZX80 and ZX81 were recycled" - tense issue, but also more clear as "parts of the designs of the ZX80 and ZX81 were reused
  • Rephrased. ♦ JAG UAR  12:18, 13 February 2024 (UTC) reply
The bigger problem with that sentence is that it might be misleading, not factual. Spectrum is a very different machine than ZX81. Reused was the idea to use Z80 CPU, to use a Ferranti ULA, to output a TV signal, bitmapped graphics, a big chunk of code for BASIC.
However, those are just some generalistic issues. Nothing specific in the architecture was reused on Spectrum. Z80Spectrum ( talk) 16:14, 25 February 2024 (UTC) reply
  • "According to Sinclair, the team had the concept of using..." - so... did they do so or just "have the concept"? And what does it mean in practical terms to combine the video and audio RAM?
  • Changed to 'idea'. I'm afraid I don't know what it means, I will change it once I find out. ♦ JAG UAR  12:18, 13 February 2024 (UTC) reply
I can guess what it means. It was probably a part of the Spectrum's design goals to consider using only one bank of RAM. The Spectrum 16K actually uses only one bank (4116 DRAM), and the access to this bank is multiplexed (both CPU and video). Spectrum can store everything in this bank: video, audio, code, etc...
However, the 48K model has a second bank of RAM (and this bank is expensive, 8 chips of 4532, which is a half-malfunctioning 4164 DRAM, where 4164 is a more modern and more expensive tech than 4116).
So, the end design only partially follows the idea of a single bank of RAM, but it can certanly be claimed that the design goal of a product with a single bank of RAM was achieved, in the form of Spectrum 16K. Z80Spectrum ( talk) 16:25, 25 February 2024 (UTC) reply
I have attempted to quickly dig out the DRAM chip prices, just to show the bill of materials.

(Edit: I quoted prices from wrong years (1981 instead of 1982); corrected)

4116 - (16 Kib per chip) - the price fell sharply in 1981, and then leveled off. In 1982, the price was about 0.91 USD/KiB (150 ns)
4164 - (64 Kib per chip) - the price was about 1.1 USD/KiB in 1982 (150 ns). The price was going to drop significantly in the middle of 1983, which contributed to the decrease of Spectrum's price.
4532 - (32 Kib per chip) - the price per chip was from 10% to 30% less than 4164 DRAM (which has double capacity). However, these are salvaged 4164 chips, so Sinclair could probably had made some good bargains.
The DRAM chips make about 30% of the ZX Spectrum's bill of materials (well, depending on the year). Z80Spectrum ( talk) 17:58, 25 February 2024 (UTC) reply
I would estimate the ZX Spectrum's bill of materials at around 125 USD in January 1983 (without assembly, packaging and Q&A costs). However, this is original research. I know approximately the price of each component (individual chips), if anyone is interested. Z80Spectrum ( talk) 18:13, 25 February 2024 (UTC) reply
  • "Much of the code was written by" - given this is a computer, the code for what? You start talking about the interpreter next, but without the context of how early 80s computers worked (e.g. that you loaded in programs that the computer ran, without a fancy operating system GUI), this is really hard for a reader to follow.
  • Vickers wrote the ROM firmware, so clarified. Let me know if this still doesn't make sense? ♦ JAG UAR  21:52, 17 March 2024 (UTC) reply
  • "the ZX Spectrum was, as quoted by Sinclair's marketing manager, essentially a "ZX81 with colour". - given that you just talked a little about the changes and go on to spend another big paragraph talking about hardware changes, this sentence seems out of place. Overall, these two paragraphs jump around between hardware changes from the ZX81 and how the software works, with this line in the middle; it would read smoother if you reorder the sentences so it's "these were the changes made, but ultimately it was a ZX81 with some changes made to support colour. The operating software for the machine to run programs was written by Vickers. It handled color like X. It took up 7kb of the system's total 16kb of memory." (also note that 16kb was the minimum/default memory amount, since there was a 48kb version at launch)
  • I've shifted the 'ZX81 with colour' sentence to the middle so flow is improved. ♦ JAG UAR  22:56, 17 March 2024 (UTC) reply
As far as I understand it, the sentence is still wrong. The sentence claims that ZX Spectrum was "ZX81 with colour", which is terribly wrong. If the sentence indicated this claim as an opinion of the marketing manager, it would be much less wrong, but still misleading. Z80Spectrum ( talk) 23:20, 18 March 2024 (UTC) reply
Aldo, that entire paragraph is quite badly written, confusing and inaccurate. Unfortunately, I don't have spare time at this moment to fix it, maybe some other day. Z80Spectrum ( talk) 23:43, 18 March 2024 (UTC) reply
  • The paragraph that starts with "A divergence of perspectives between Nine Tiles" uses about 50% too many words to explain that Sinclair wanted to reuse the ZX80 software architecture with additions in order to launch quickly, but Nine Tiles (and you haven't previously mentioned that it was more than just Vickers, but was the company as well) thought that it would make a mess as the ZX80 design was optimized for less memory and processing demands. It just reads really stilted, and a quick rewrite would sort it out. Also, you don't actually say which side won, just that they disagreed, so it doesn't really go anywhere in the end.
  • "brainchild" is still informal
  • Changed. ♦ JAG UAR  12:18, 13 February 2024 (UTC) reply
Point 1. Jaguar wrote: Much of the code was written by - I think that you have nailed the problem there: the article is missing an entire section describing how the Spctrum was operated in practice. I think that some sections from the article on ZX81 can be reused for this purpose. However, this will further expand the article, so I propose splitting the history section into another article.
Point 2. Jaguar wrote the ZX Spectrum was, as quoted by Sinclair's marketing manager, essentially a "ZX81 with colour - the problem here is the following: from a perspective of initial concept of ZX Spectrum, it was a "ZX81 with color". That's what Clive Sinclair wanted, and that has been cited everywhere. However, the end-design, by engineer Altwasser, is much more than that. Alwasser has obviously put a lot of sweat into the design, even whan working under the extreme time pressure that was obviously imposed by Clive Sinclair. To cut it short, the end design is much more than just "ZX81 with color", and in that sense the sentence is misleading. Z80Spectrum ( talk) 16:41, 25 February 2024 (UTC) reply
Point 3. Jaguar wrote The paragraph that starts with "A divergence of perspectives between Nine Tiles - the development of the firmware ended when the ROM was full, and development time had also 'run out'. It was essentially the BASIC from ZX81, plus floating point math (demanded by Clive Sinclair), additional graphics routines and a few extras (new cassete tape routines). I don't know how should that be interpreted in the article as somebody's 'win'. Perhaps it's more towards Sinclair's viewpoint. Z80Spectrum ( talk) 16:53, 25 February 2024 (UTC) reply
  • Was the Designing of the keyboard "minimalised", or was the design "simplified"?
  • Simplified is better, changed. ♦ JAG UAR  12:26, 13 February 2024 (UTC) reply
  • I have an electrical engineering degree, and the paragraph on the cassette reader makes perfect sense to me. I'm not convinced the average reader can follow it unless they dealt with cassette-based computing. I don't think it's a major problem for GA, but you should consider if there's way to make it more clear.
  • I'm a Luddite, me no code... still, I'm glad one of us knows what I've written about! To make things hopefully clearer, I've linked instruction cycles and inserted some plain English to make the jargon easier to understand. ♦ JAG UAR  14:59, 26 March 2024 (UTC) reply
  • I'm a little confused about the Timex paragraph- you say that it wasn't an obvious manufacturing partner as they made watches, but then sneak in that it manufactured the ZX80. 1) It also manufactured the ZX81, which isn't mentioned, but more importantly 2) if it had been the manufacturer for the previous 2 computers, then at that point it was a very obvious manufacturing partner.
  • I've rephrased and rearranged parts of the Timex paragraph to make it chronologically clearer. ♦ JAG UAR  23:25, 2 April 2024 (UTC) reply
Launch
  • "With the arrival of the more inexpensive Issue 2 motherboard, production rapidly increased." - this is the first mention of "Issue 2"; what is that? Also the first mention of a motherboard, so link and maybe namedrop earlier in the article. And why did a cheaper motherboard mean they could suddenly produce so many more units a month, or was it just that they increased production when they launched a second version?
  • Good question. I have determined from the Adamson and Kennedy source that the former is correct, as Sinclair started producing more units with a streamlined, cheaper motherboard. As such I've rephrased this to clarify. Also linked motherboard. ♦ JAG UAR  23:25, 2 April 2024 (UTC) reply
  • Again with Issue 3- why did the introduction of a new computer part boost sales?
  • The source doesn't specify if the Issue 3 motherboard itself boosted sales, but I've rephrased it to separate. ♦ JAG UAR  23:25, 2 April 2024 (UTC) reply
  • You give sales numbers of 200k in the first 9 months, 300k in the first year, and 500k by August 1983... but you started the paragraph by already saying that "a further" 500k had been sold in 1982 and into 1983. So... which is it?
  • Oops, doesn't make sense to me upon reading it again. I've removed the "a further 500k" sentence to cut confusion. ♦ JAG UAR  11:38, 3 April 2024 (UTC) reply
  • What was unusual about the marketing campaign, other than the weird lack of budget?
  • This seems to have been removed by another editor while I've been gone. ♦ JAG UAR  21:48, 23 April 2024 (UTC) reply
  • I hate to ask for more content, given the length of the article, but... it is really jarring to spend so much time talking about how successful the computer was in the UK and then just have a one sentence "oh yeah, it was a failure in the US btw." Do we know why?
  • I have summarised the Timex's limited success in a not-so-meaty paragraph. ♦ JAG UAR  22:27, 15 April 2024 (UTC) reply
Success and market domination
  • Oh, you talk about the marketing campaign here. So, that little paragraph needs to move down with the rest of the marketing talk.
  • " in due part to saturation of home computers such as the ZX Spectrum." - to clarify, is this trying to say that the UK market was into microcomputers instead of game consoles, so the crash of the game console market didn't affect the UK industry the same way?
  • Yes, clarified. The dominance of microcomputers was unequivocal, especially outside the US and Japan. ♦ JAG UAR  21:48, 23 April 2024 (UTC) reply
  • Alternately, you could just drop the microcomputer bit entirely and just say that the crash of 83 only affected America and not the UK (it didn't much affect Japan either- the main cause of the crash was oversaturation of supply in US retailers by publishers aka Warner due to a misreading of increased store demand (due to an increase of retailers carrying video games) with increased consumer demand (in turn due to nonsensical supply chains causing a complete lack of clarity of what was going on), but that only happened in the US).
  • While I agree the crash of '83 was primarily a 'US thing', it did send ripples through the video game market worldwide. There was a short-lived console market in the UK between 1980-82 which the crash may or may not have killed off. I'd prefer its inclusion, but I have tweaked it to make it clear it affected American markets. ♦ JAG UAR  21:48, 23 April 2024 (UTC) reply
Later years and company decline
  • "The Spectrum+ retained the identical technical specifications as the original Spectrum." - you already said it was just a rebranded Spectrum, so this sentence can be cut.
I appologize for previously attributing Jaguar instead of PresN, sorry.
Point 4. PresN wrote With the arrival of the more inexpensive Issue 2 motherboard, production rapidly increased - 'Issue 2' is the version of the motherboard. There were multiple revisions of the motherboard, because ZX Soectrum was initially full of hardware bugs. That means many Spectrums were malfunctioning when tested brand new out from the factory line. New motherboard issues fixed those problems. It also makes the production cheaper, by not creating a huge pile of malfunctioning products. Z80Spectrum ( talk) 17:05, 25 February 2024 (UTC) reply
  • "Despite the continued domination of home computer market with the ZX Spectrum" - needs a "the" before home, but also the whole phrase is odd- "despite his success, he hoped to repeat his success"? Also we just spent 2 paragraphs talking about how by 1985 things weren't going well, so it's a little jarring to jump back to them being dominant.
  • Did this myself

I'm going to take a break here, and pick back up with the rest of the article later. -- Pres N 21:55, 26 January 2024 (UTC) reply

@ Jaguar: Are you going to be able to work on this article? -- Pres N 23:31, 3 February 2024 (UTC) reply
@ PresN: I can take over this review if he's gone AWOL. DigitalIceAge ( talk) 00:48, 11 February 2024 (UTC) reply
He last edited this article a few days ago and I've been talking to him on Discord about it. -- Pres N 01:27, 11 February 2024 (UTC) reply
Thumbs up icon DigitalIceAge ( talk) 01:58, 11 February 2024 (UTC) reply
I won't be editing the article (for the time being), but if anyone needs my suggestions, or advices, feel free to ask. Z80Spectrum ( talk) 17:09, 25 February 2024 (UTC) reply

@ Jaguar: Poke. -- Pres N 01:41, 23 February 2024 (UTC) reply

You have the patience of a saint, PresN! I have returned from abroad and am addressing your comments now. Let's hit this phase on the head. ♦ JAG UAR  21:54, 15 March 2024 (UTC) reply

Okay, this long-running review has gotten a bit messy. I've gone through and made tweaks to the sections above, so we're on to "Hardware" and below. -- Pres N 01:14, 24 April 2024 (UTC) reply

Hardware
  • "where a desired colour of a specific pixel could not necessarily be selected." -> "where a desired colour of a specific pixel could not be selected, but only the colour attributes of an 8x8 block." (Since this is a restatement ("In practical terms"), you should re-emphasize the 8x8 block concept.)
  • "It is controlled by a single EAR bit, located on port 0xFE." Waaaay too overdetailed. This is an article in a generalist encyclopedia; specific bit names and port numbers are too much. This extraneous detail is then repeated in the next sentence for some reason.
  • That color palette image would make more sense in the graphics section; maybe swap with the motherboard photo.
  • Firmware restates the stuff about Vickers/Nine Tails for some reason
  • I'm on the fence about if this whole section is a little too detailed... I think it's okay for GA, though.
Sinclair Research models
  • "An "Issue 1" ZX Spectrum can be distinguished from later models" - there were only 3 issues, right? Or at least that's all that's mentioned in Development, so this can state "can be distinguished from Issue 2 or 3 models" to be clear. Also, if you don't solve it in Development, this first paragraph should explicitly state how many issues there were, since it talks about the first.
  • "Within the original iterations of the 16 and 48K models, an internal speaker with severely restricted capabilities served as the audio output. This speaker, capable of producing just one note at a time" - this was already discussed above in "Sound"
  • "outsold the rubber-key model 2:1" - ugh, "two to one", please.
  • "RAM disc commands save !"name"" - the specific command used is meaningless to 99.9% of readers
  • "Sinclair unveiled the ZX Spectrum 128 at The May Fair Hotel's Crystal Rooms in London" - the preceding sentence said it was presented at SIMO '85, so how could it be "unveiled" later? Also, when is later?
  • " it has no internal speaker, being produced from the television speaker instead." - " it has no internal speaker, and can only produce sound from the television speaker."
Amstrad models
  • "The ZX Spectrum +2 used a power supply..." - don't do one-sentence paragraphs. Integrate this into the first ZX Spectrum +2 paragraph with the rest of the hardware.
  • Not a real comment, but I cannot believe they thought it was good marketing to have a +2, +3, +2A, +2B, and +3B models. And the original Spectrum. At the same time. What a mess.
  • In general, I think this whole section is probably too detailed on the specifics for a generalist encyclopedia, which I won't complain about at GA but would be an issue at FAC.
Clones and re-creations
  • Paragraph 1 has no citation, and so does much of what follows.
  • This whole "Official licences" subsection is messy- we already talk about the Timex Sinclair 2068 in "Launch" as just... the Spectrum in America, and no longer call it a clone in the lede, so listing out the US and European models as clones contradicts. I think you should scrap this whole thing and write a one or two paragraph summary of the non-UK models. I would also move it out from under Clones and re-creations to be "Officially licenced models" right after "Amstrad models".
  • Most of "Unofficial clones" is uncited
  • I feel like "Recreations" is too long, but I guess 1 paragraph for 5 things adds up.
From Wikipedia, the free encyclopedia

GA Review

Article ( | visual edit | history) · Article talk ( | history) · Watch

Reviewer: PresN ( talk · contribs) 23:38, 25 January 2024 (UTC) reply


Claiming this review; I'll start posting comments soon. -- Pres N 23:38, 25 January 2024 (UTC) reply

Thank you for taking the Herculean task of reviewing the mighty Spectrum, PresN! I have returned in good time, and will begin addressing your comments. I note that some of the technical sections have been altered or degraded since I have been gone - but will tackle that when we come to it. ♦ JAG UAR  22:52, 4 February 2024 (UTC) reply
Lead
  • It's odd for the lead to launch almost immediately into the clones, before discussing what the machine is or that it was released outside the UK. I would have expected that sentence at the end of what is now paragraph 3.
  • I've removed the mention of clones entirely and condensed the rest. I think it reads alright. ♦ JAG UAR  23:19, 4 February 2024 (UTC) reply
  • "The machine was the brainchild of" - oddly informal
  • Inclined to agree, though I think most know British English employs a more colloquial reading experience, even for Wikipedia. I've changed it to 'creation'. ♦ JAG UAR  23:19, 4 February 2024 (UTC) reply
  • The lead says it was released "in some parts of Europe as the Timex Computer 2048.", but the infobox just says Spain in 1985. Sounds like at minimum that should be changed to EU in 1985, unless you want to get more specific about the countries.
  • Yeah, that would be a nightmare. Changed to 'EU'. ♦ JAG UAR  23:19, 4 February 2024 (UTC) reply
In 1985 Spain wasn't in the then EC (the EU didn't exist under that name in 1985 either). Although I can't quickly find a single source which lists release dates in different European countries, I would have expected that it was launched earlier in e.g.
Ireland than Spain. 90.167.254.202 ( talk) 22:29, 29 March 2024 (UTC) reply
'EU' is generally referring to Europe; it is easier to cluster the releases that way. ♦ JAG UAR  21:56, 2 April 2024 (UTC) reply
  • Actually, now it seems like the issue is that lead is playing loose with whether the Timex versions were versions of the Spectrum or were clones of the Spectrum. It doesn't seem right to call the Timex a clone, then turn around and say the Spectrum was released in the US as the 2068. Personally, I'd stick with calling them versions or variants, seeing as they were official (slightly modified) Spectrums.
  • Agree. Since I removed the mention of clones above I hope it negates this issue. I'd call the Timex versions versions, and the Soviet imitations clones. Let me know how it reads? ♦ JAG UAR  23:19, 4 February 2024 (UTC) reply
  • Happier with the lede now! -- Pres N 03:27, 5 February 2024 (UTC) reply
Background/Development
  • This feels like its going way too deep into Sinclair's history. Nothing about Sinclair Radionics is directly relevant to this computer; its important to note they had made the ZX80 and 81, and had been making electronics since the 60s, but the details of his business history is too much.
  • I've cut down on some of the less relevant company history. I took inspiration from the ZX81's background subsection, which to be fair adheres much less to WPVG's manual of style. I think keeping the mention about Sinclair's relationship with the NEB is important, as his breakup prompted him to develop microchips, which ultimately led to the Spectrum. ♦ JAG UAR  11:51, 13 February 2024 (UTC) reply
Hello, Jaguar, and thank you for your interest in the article.
I didn't see this review on time, so I reverted your edit [1]. If you really feel that it doesn't belong to the article, then you can undo my revert. However, I personally find the remove sentence both interesting and relevant. Z80Spectrum ( talk) 16:00, 25 February 2024 (UTC) reply
  • "Sinclair resolved to make his own products obsolete before his rivals developed the products that would do so, thus seeking to make the technology as cheap as possible." - not following how wanting to make a computer better than the ZX80/81 leads to making the Spectrum (or is it computers in general?) cheap.
  • That doesn't read well! I cut the second half entirely. ♦ JAG UAR  11:57, 13 February 2024 (UTC) reply
  • "Architecture from the ZX80 and ZX81 were recycled" - tense issue, but also more clear as "parts of the designs of the ZX80 and ZX81 were reused
  • Rephrased. ♦ JAG UAR  12:18, 13 February 2024 (UTC) reply
The bigger problem with that sentence is that it might be misleading, not factual. Spectrum is a very different machine than ZX81. Reused was the idea to use Z80 CPU, to use a Ferranti ULA, to output a TV signal, bitmapped graphics, a big chunk of code for BASIC.
However, those are just some generalistic issues. Nothing specific in the architecture was reused on Spectrum. Z80Spectrum ( talk) 16:14, 25 February 2024 (UTC) reply
  • "According to Sinclair, the team had the concept of using..." - so... did they do so or just "have the concept"? And what does it mean in practical terms to combine the video and audio RAM?
  • Changed to 'idea'. I'm afraid I don't know what it means, I will change it once I find out. ♦ JAG UAR  12:18, 13 February 2024 (UTC) reply
I can guess what it means. It was probably a part of the Spectrum's design goals to consider using only one bank of RAM. The Spectrum 16K actually uses only one bank (4116 DRAM), and the access to this bank is multiplexed (both CPU and video). Spectrum can store everything in this bank: video, audio, code, etc...
However, the 48K model has a second bank of RAM (and this bank is expensive, 8 chips of 4532, which is a half-malfunctioning 4164 DRAM, where 4164 is a more modern and more expensive tech than 4116).
So, the end design only partially follows the idea of a single bank of RAM, but it can certanly be claimed that the design goal of a product with a single bank of RAM was achieved, in the form of Spectrum 16K. Z80Spectrum ( talk) 16:25, 25 February 2024 (UTC) reply
I have attempted to quickly dig out the DRAM chip prices, just to show the bill of materials.

(Edit: I quoted prices from wrong years (1981 instead of 1982); corrected)

4116 - (16 Kib per chip) - the price fell sharply in 1981, and then leveled off. In 1982, the price was about 0.91 USD/KiB (150 ns)
4164 - (64 Kib per chip) - the price was about 1.1 USD/KiB in 1982 (150 ns). The price was going to drop significantly in the middle of 1983, which contributed to the decrease of Spectrum's price.
4532 - (32 Kib per chip) - the price per chip was from 10% to 30% less than 4164 DRAM (which has double capacity). However, these are salvaged 4164 chips, so Sinclair could probably had made some good bargains.
The DRAM chips make about 30% of the ZX Spectrum's bill of materials (well, depending on the year). Z80Spectrum ( talk) 17:58, 25 February 2024 (UTC) reply
I would estimate the ZX Spectrum's bill of materials at around 125 USD in January 1983 (without assembly, packaging and Q&A costs). However, this is original research. I know approximately the price of each component (individual chips), if anyone is interested. Z80Spectrum ( talk) 18:13, 25 February 2024 (UTC) reply
  • "Much of the code was written by" - given this is a computer, the code for what? You start talking about the interpreter next, but without the context of how early 80s computers worked (e.g. that you loaded in programs that the computer ran, without a fancy operating system GUI), this is really hard for a reader to follow.
  • Vickers wrote the ROM firmware, so clarified. Let me know if this still doesn't make sense? ♦ JAG UAR  21:52, 17 March 2024 (UTC) reply
  • "the ZX Spectrum was, as quoted by Sinclair's marketing manager, essentially a "ZX81 with colour". - given that you just talked a little about the changes and go on to spend another big paragraph talking about hardware changes, this sentence seems out of place. Overall, these two paragraphs jump around between hardware changes from the ZX81 and how the software works, with this line in the middle; it would read smoother if you reorder the sentences so it's "these were the changes made, but ultimately it was a ZX81 with some changes made to support colour. The operating software for the machine to run programs was written by Vickers. It handled color like X. It took up 7kb of the system's total 16kb of memory." (also note that 16kb was the minimum/default memory amount, since there was a 48kb version at launch)
  • I've shifted the 'ZX81 with colour' sentence to the middle so flow is improved. ♦ JAG UAR  22:56, 17 March 2024 (UTC) reply
As far as I understand it, the sentence is still wrong. The sentence claims that ZX Spectrum was "ZX81 with colour", which is terribly wrong. If the sentence indicated this claim as an opinion of the marketing manager, it would be much less wrong, but still misleading. Z80Spectrum ( talk) 23:20, 18 March 2024 (UTC) reply
Aldo, that entire paragraph is quite badly written, confusing and inaccurate. Unfortunately, I don't have spare time at this moment to fix it, maybe some other day. Z80Spectrum ( talk) 23:43, 18 March 2024 (UTC) reply
  • The paragraph that starts with "A divergence of perspectives between Nine Tiles" uses about 50% too many words to explain that Sinclair wanted to reuse the ZX80 software architecture with additions in order to launch quickly, but Nine Tiles (and you haven't previously mentioned that it was more than just Vickers, but was the company as well) thought that it would make a mess as the ZX80 design was optimized for less memory and processing demands. It just reads really stilted, and a quick rewrite would sort it out. Also, you don't actually say which side won, just that they disagreed, so it doesn't really go anywhere in the end.
  • "brainchild" is still informal
  • Changed. ♦ JAG UAR  12:18, 13 February 2024 (UTC) reply
Point 1. Jaguar wrote: Much of the code was written by - I think that you have nailed the problem there: the article is missing an entire section describing how the Spctrum was operated in practice. I think that some sections from the article on ZX81 can be reused for this purpose. However, this will further expand the article, so I propose splitting the history section into another article.
Point 2. Jaguar wrote the ZX Spectrum was, as quoted by Sinclair's marketing manager, essentially a "ZX81 with colour - the problem here is the following: from a perspective of initial concept of ZX Spectrum, it was a "ZX81 with color". That's what Clive Sinclair wanted, and that has been cited everywhere. However, the end-design, by engineer Altwasser, is much more than that. Alwasser has obviously put a lot of sweat into the design, even whan working under the extreme time pressure that was obviously imposed by Clive Sinclair. To cut it short, the end design is much more than just "ZX81 with color", and in that sense the sentence is misleading. Z80Spectrum ( talk) 16:41, 25 February 2024 (UTC) reply
Point 3. Jaguar wrote The paragraph that starts with "A divergence of perspectives between Nine Tiles - the development of the firmware ended when the ROM was full, and development time had also 'run out'. It was essentially the BASIC from ZX81, plus floating point math (demanded by Clive Sinclair), additional graphics routines and a few extras (new cassete tape routines). I don't know how should that be interpreted in the article as somebody's 'win'. Perhaps it's more towards Sinclair's viewpoint. Z80Spectrum ( talk) 16:53, 25 February 2024 (UTC) reply
  • Was the Designing of the keyboard "minimalised", or was the design "simplified"?
  • Simplified is better, changed. ♦ JAG UAR  12:26, 13 February 2024 (UTC) reply
  • I have an electrical engineering degree, and the paragraph on the cassette reader makes perfect sense to me. I'm not convinced the average reader can follow it unless they dealt with cassette-based computing. I don't think it's a major problem for GA, but you should consider if there's way to make it more clear.
  • I'm a Luddite, me no code... still, I'm glad one of us knows what I've written about! To make things hopefully clearer, I've linked instruction cycles and inserted some plain English to make the jargon easier to understand. ♦ JAG UAR  14:59, 26 March 2024 (UTC) reply
  • I'm a little confused about the Timex paragraph- you say that it wasn't an obvious manufacturing partner as they made watches, but then sneak in that it manufactured the ZX80. 1) It also manufactured the ZX81, which isn't mentioned, but more importantly 2) if it had been the manufacturer for the previous 2 computers, then at that point it was a very obvious manufacturing partner.
  • I've rephrased and rearranged parts of the Timex paragraph to make it chronologically clearer. ♦ JAG UAR  23:25, 2 April 2024 (UTC) reply
Launch
  • "With the arrival of the more inexpensive Issue 2 motherboard, production rapidly increased." - this is the first mention of "Issue 2"; what is that? Also the first mention of a motherboard, so link and maybe namedrop earlier in the article. And why did a cheaper motherboard mean they could suddenly produce so many more units a month, or was it just that they increased production when they launched a second version?
  • Good question. I have determined from the Adamson and Kennedy source that the former is correct, as Sinclair started producing more units with a streamlined, cheaper motherboard. As such I've rephrased this to clarify. Also linked motherboard. ♦ JAG UAR  23:25, 2 April 2024 (UTC) reply
  • Again with Issue 3- why did the introduction of a new computer part boost sales?
  • The source doesn't specify if the Issue 3 motherboard itself boosted sales, but I've rephrased it to separate. ♦ JAG UAR  23:25, 2 April 2024 (UTC) reply
  • You give sales numbers of 200k in the first 9 months, 300k in the first year, and 500k by August 1983... but you started the paragraph by already saying that "a further" 500k had been sold in 1982 and into 1983. So... which is it?
  • Oops, doesn't make sense to me upon reading it again. I've removed the "a further 500k" sentence to cut confusion. ♦ JAG UAR  11:38, 3 April 2024 (UTC) reply
  • What was unusual about the marketing campaign, other than the weird lack of budget?
  • This seems to have been removed by another editor while I've been gone. ♦ JAG UAR  21:48, 23 April 2024 (UTC) reply
  • I hate to ask for more content, given the length of the article, but... it is really jarring to spend so much time talking about how successful the computer was in the UK and then just have a one sentence "oh yeah, it was a failure in the US btw." Do we know why?
  • I have summarised the Timex's limited success in a not-so-meaty paragraph. ♦ JAG UAR  22:27, 15 April 2024 (UTC) reply
Success and market domination
  • Oh, you talk about the marketing campaign here. So, that little paragraph needs to move down with the rest of the marketing talk.
  • " in due part to saturation of home computers such as the ZX Spectrum." - to clarify, is this trying to say that the UK market was into microcomputers instead of game consoles, so the crash of the game console market didn't affect the UK industry the same way?
  • Yes, clarified. The dominance of microcomputers was unequivocal, especially outside the US and Japan. ♦ JAG UAR  21:48, 23 April 2024 (UTC) reply
  • Alternately, you could just drop the microcomputer bit entirely and just say that the crash of 83 only affected America and not the UK (it didn't much affect Japan either- the main cause of the crash was oversaturation of supply in US retailers by publishers aka Warner due to a misreading of increased store demand (due to an increase of retailers carrying video games) with increased consumer demand (in turn due to nonsensical supply chains causing a complete lack of clarity of what was going on), but that only happened in the US).
  • While I agree the crash of '83 was primarily a 'US thing', it did send ripples through the video game market worldwide. There was a short-lived console market in the UK between 1980-82 which the crash may or may not have killed off. I'd prefer its inclusion, but I have tweaked it to make it clear it affected American markets. ♦ JAG UAR  21:48, 23 April 2024 (UTC) reply
Later years and company decline
  • "The Spectrum+ retained the identical technical specifications as the original Spectrum." - you already said it was just a rebranded Spectrum, so this sentence can be cut.
I appologize for previously attributing Jaguar instead of PresN, sorry.
Point 4. PresN wrote With the arrival of the more inexpensive Issue 2 motherboard, production rapidly increased - 'Issue 2' is the version of the motherboard. There were multiple revisions of the motherboard, because ZX Soectrum was initially full of hardware bugs. That means many Spectrums were malfunctioning when tested brand new out from the factory line. New motherboard issues fixed those problems. It also makes the production cheaper, by not creating a huge pile of malfunctioning products. Z80Spectrum ( talk) 17:05, 25 February 2024 (UTC) reply
  • "Despite the continued domination of home computer market with the ZX Spectrum" - needs a "the" before home, but also the whole phrase is odd- "despite his success, he hoped to repeat his success"? Also we just spent 2 paragraphs talking about how by 1985 things weren't going well, so it's a little jarring to jump back to them being dominant.
  • Did this myself

I'm going to take a break here, and pick back up with the rest of the article later. -- Pres N 21:55, 26 January 2024 (UTC) reply

@ Jaguar: Are you going to be able to work on this article? -- Pres N 23:31, 3 February 2024 (UTC) reply
@ PresN: I can take over this review if he's gone AWOL. DigitalIceAge ( talk) 00:48, 11 February 2024 (UTC) reply
He last edited this article a few days ago and I've been talking to him on Discord about it. -- Pres N 01:27, 11 February 2024 (UTC) reply
Thumbs up icon DigitalIceAge ( talk) 01:58, 11 February 2024 (UTC) reply
I won't be editing the article (for the time being), but if anyone needs my suggestions, or advices, feel free to ask. Z80Spectrum ( talk) 17:09, 25 February 2024 (UTC) reply

@ Jaguar: Poke. -- Pres N 01:41, 23 February 2024 (UTC) reply

You have the patience of a saint, PresN! I have returned from abroad and am addressing your comments now. Let's hit this phase on the head. ♦ JAG UAR  21:54, 15 March 2024 (UTC) reply

Okay, this long-running review has gotten a bit messy. I've gone through and made tweaks to the sections above, so we're on to "Hardware" and below. -- Pres N 01:14, 24 April 2024 (UTC) reply

Hardware
  • "where a desired colour of a specific pixel could not necessarily be selected." -> "where a desired colour of a specific pixel could not be selected, but only the colour attributes of an 8x8 block." (Since this is a restatement ("In practical terms"), you should re-emphasize the 8x8 block concept.)
  • "It is controlled by a single EAR bit, located on port 0xFE." Waaaay too overdetailed. This is an article in a generalist encyclopedia; specific bit names and port numbers are too much. This extraneous detail is then repeated in the next sentence for some reason.
  • That color palette image would make more sense in the graphics section; maybe swap with the motherboard photo.
  • Firmware restates the stuff about Vickers/Nine Tails for some reason
  • I'm on the fence about if this whole section is a little too detailed... I think it's okay for GA, though.
Sinclair Research models
  • "An "Issue 1" ZX Spectrum can be distinguished from later models" - there were only 3 issues, right? Or at least that's all that's mentioned in Development, so this can state "can be distinguished from Issue 2 or 3 models" to be clear. Also, if you don't solve it in Development, this first paragraph should explicitly state how many issues there were, since it talks about the first.
  • "Within the original iterations of the 16 and 48K models, an internal speaker with severely restricted capabilities served as the audio output. This speaker, capable of producing just one note at a time" - this was already discussed above in "Sound"
  • "outsold the rubber-key model 2:1" - ugh, "two to one", please.
  • "RAM disc commands save !"name"" - the specific command used is meaningless to 99.9% of readers
  • "Sinclair unveiled the ZX Spectrum 128 at The May Fair Hotel's Crystal Rooms in London" - the preceding sentence said it was presented at SIMO '85, so how could it be "unveiled" later? Also, when is later?
  • " it has no internal speaker, being produced from the television speaker instead." - " it has no internal speaker, and can only produce sound from the television speaker."
Amstrad models
  • "The ZX Spectrum +2 used a power supply..." - don't do one-sentence paragraphs. Integrate this into the first ZX Spectrum +2 paragraph with the rest of the hardware.
  • Not a real comment, but I cannot believe they thought it was good marketing to have a +2, +3, +2A, +2B, and +3B models. And the original Spectrum. At the same time. What a mess.
  • In general, I think this whole section is probably too detailed on the specifics for a generalist encyclopedia, which I won't complain about at GA but would be an issue at FAC.
Clones and re-creations
  • Paragraph 1 has no citation, and so does much of what follows.
  • This whole "Official licences" subsection is messy- we already talk about the Timex Sinclair 2068 in "Launch" as just... the Spectrum in America, and no longer call it a clone in the lede, so listing out the US and European models as clones contradicts. I think you should scrap this whole thing and write a one or two paragraph summary of the non-UK models. I would also move it out from under Clones and re-creations to be "Officially licenced models" right after "Amstrad models".
  • Most of "Unofficial clones" is uncited
  • I feel like "Recreations" is too long, but I guess 1 paragraph for 5 things adds up.

Videos

Youtube | Vimeo | Bing

Websites

Google | Yahoo | Bing

Encyclopedia

Google | Yahoo | Bing

Facebook