Talk:ZX Spectrum

Memory
""The Z80 processor used in the Spectrum has a 16-bit address bus, which means only 64 KB of memory can be directly addressed. To facilitate the extra 80 KB of RAM the designers used bank switching so the new memory would be available as eight pages of 16 KB at the top of the address space. The same technique was used to page between the new 16 KB editor ROM and the original 16 KB BASIC ROM at the bottom of the address space""

It is the 8 bit processor which limits the amount of memory which can be addressed, not the address bus, which can be latched (as in the IBM PC). And I'm fairly sure because of the kernal and the way the pins on the Z80 are tied up, only 48 kB can be accessed in the Spectrum. I believe the extra memory on the +2 is accessed by paging in 16 kB at a time and swapping the original contents out "high" above 48K, using the offset addressing method available on the Z80.101.178.163.92 (talk) 04:02, 16 March 2018 (UTC)

In the case of ZX Spectrum 48K it is actually 16 kB ROM and 48 kB RAM = 64 kB. They are occupying entire 64 kB address space. — Preceding unsigned comment added by 96.43.227.182 (talk) 04:07, 18 November 2020 (UTC)
 * Yep, the RAM starts at address 16384 (decimal). There were some add-ons for the earlier Spectrums that allowed the ROM to be bypassed, effectively swapping in the contents of a programmable aftermarket chip.  Presumably that's also how the Hobbit managed to use 64K RAM without breaking compatibility. Pastychomper (talk) 10:20, 20 November 2020 (UTC)

Talk page archiving
I've archived threads from this page, it's been 10 years, I thought it was about time. Have archived anything over 2 years or where the discussion had been closed. - X201 (talk) 16:06, 29 March 2018 (UTC)

I don't think this is representative
"new software titles continue to be released – over 40 were released in 2019." I don't think this is representative of the actual success of the Spectrum. The software released in 2019 are part of the retro craze, which is seperate from the longevity of the product itself. And the word "software" is misleading. Are people still using the Spectrum for office work or business applications? How many of these are games? 80.98.184.139 (talk) 21:50, 1 March 2020 (UTC)
 * Seems a reasonable claim to me. Software has been released regularly for the Spectrum (and C=64 and others) for the last 20 years.  I don't think it falls into the "craze" criteria.  Anyway, the fact that software is released is testament to the longevity, no matter how you see it.  Software is a catch-all statement.  It's not misleading (all releases are software - whether games or business, or utilities), but it's not meant to specify the type of software, only that it was released.
 * Why don't you think it representative? What is the issue with it?  As I said, it seems fine to me.  Chaheel Riens (talk) 22:26, 1 March 2020 (UTC)
 * I agree with the original poster 100%. This statement is incredibly misleading and lacks context. A reasonable reader of the article who knows nothing about computers could easily assume that these are commercial releases rather than enthusiast freeware and that they are evidence of a continued market for the computer when in fact the world at large has moved on from it decades ago.  Also, I doubt this really says anything about the longevity of the computer, as I presume this software is being played by its relatively small audience on emulators and modern hardware clones, not actual Spectrums. Furthermore, a forum post is not a reliable source, and this info is not expanded upon in the body of the article, which violates WP:Lead.  If someone wants to find significant coverage in reliable sources regarding the retro Spectrum game scene and discuss it in some appropriate spot on Wikipedia, then they are welcome to explore that.  This misleading and unsupported statement given undue attention in the lead is not the way to go about it. Indrian (talk) 02:41, 2 March 2020 (UTC)
 * Why does it matter if the software is commercial or homebrew? Or if it's used on an emulator or real hardware?  It's still a software release.  Other points accepted, but not those I mention.  Chaheel Riens (talk) 07:34, 2 March 2020 (UTC)
 * To clarify, I don't think it matters in terms of what topics to cover on Wikipedia, but I do think these facts should be clearly stated in any writing on the subject to avoid confusion. It's the context that matters. Indrian (talk) 08:20, 2 March 2020 (UTC)

Successor
As far as I can see, the Next is a (mostly? fully?) Spectrum-compatible machine with added capabilities. The same is true of the Hobbit and the Sam Coupe, both of which were sold while the Spectrum still had a market beyond retro enthusiasts. I take the anon editor's point that the QL wasn't much of a successor since several Spectra were released after it, but it was at least intended by Sinclair to be one, which isn't true of any of the others. How about we remove the "successor" heading altogether? Pastychomper (talk) 12:22, 28 June 2022 (UTC)


 * The unlike the Sam Coupe, the Spectrum Next is an official Sinclair product, branded as such. This sets it appart from the Sam Coupe and other community-driven projects such as the UNO etc.
 * The editor MrMajors has been engaged in endless editing wars on the ZX Spectrum Next page itself, has put forward for deletion the Wikipedia entry of the people running the project and has overall pushed for a personal vendetta against the Spectrum Next. I suggest disabling his ability to edit the entry for the time being. 37.157.32.210 (talk) 15:43, 28 June 2022 (UTC)
 * I am gonna have to side with this. MrMajors seems to be simply sabotaging other peoples work to match his personal agenda, without any real goal of actually improving the articles. 2001:14BA:16F7:5700:E54B:DFB1:75E1:3F00 (talk) 11:23, 24 July 2022 (UTC)
 * The only agenda here is that of a small number of Spectrum Next fanboys attempting to promote their obscure toy. MrMajors (talk) 11:32, 24 July 2022 (UTC)
 * Why would something that is no longer commercially available need any promoting? What purpose would the promotion serve? I think it is more just the case of documenting that such a reproduction version exists. 2001:14BA:16F7:5700:E54B:DFB1:75E1:3F00 (talk) 11:48, 24 July 2022 (UTC)
 * As an addendum, the rationale that the QL wasn't an 8 bits computer and therefore can't be a successor doesn't stack up. No one would dispute that the Mega Drive was a successor to the Master System, or that the 80286 was the successor of the 8088 on such grounds. 37.157.32.210 (talk) 16:06, 28 June 2022 (UTC)


 * The QL was the computer created by Clive Sinclair after the Spectrum, it is therefore the successor. The Next is not an official Sinclair product, the name is used with permission of the rights holder. Even the Sam and the Vega's have a greater claim to be the successor as the Coupe was created by Miles & Gordon who were ex-Sinclair, and Sinclair Research was a director of RCL, the company behind the Vega consoles. MrMajors (talk) 16:19, 28 June 2022 (UTC)
 * This is strange. You've claimed earlier in the history that the QL couldn't be a successor because it's 8 bits. Now that's no longer an issue?
 * As you've mentioned, the Next is a licenced product from the rights holder, therefore an official Sinclair product -- it even carries the brand, something none of the examples you've mentioned do.
 * This seems very much like a personal issue of yours, not an objective reason that applies to the readers of Wikipedia. 37.157.32.210 (talk) 16:38, 28 June 2022 (UTC)
 * 1. It was another editor that said the QL couldn't be a successor.
 * 2. from the Nexts' own page: "Despite the name, the machine is not directly affiliated with Sinclair Research Ltd., Sir Clive Sinclair or the current owner of the trademarks, Sky Group."
 * 3. the Vega also used the brand Sinclair, licenced from Sky, so is "therefore an official Sinclair product" according to your own criteria, and even moreso due to the involvement of Sinclair Research and Clive Sinclair himself
 * MrMajors (talk) 17:41, 28 June 2022 (UTC)
 * I'm kind of with MrMajors and his rationale on this one. The QL is the branded successor to the Spectrum despite many differences and obvious incompatibilities WP:OTHERSTUFF notwithstanding, but the Commodore 64 article includes the Amiga as a successor, which I would also agree with, and the thinking that includes it there applies here as well.
 * IP 37.157 asked for another editor to weight[sic] in - so I have. Chaheel Riens (talk) 17:58, 28 June 2022 (UTC)
 * First of all, sory for this wall of text. I must say, that I find the question of "successors" somewhat hard to define when it comes to early home computers. It all depends on what you mean by "successor". Is it simply a product brought to the market by the same company and cronologically following the "predecessor"? In this way QL is the successor of ZX Spectrum, Amiga is the successor of Commodore 64 and Atari ST is the successor of Atari XL/XE.
 * However I, for one, do find this approach overly corporate-centric and disregarding large parts of computing history. So should we focus on the people involved in the development of machine instead? In this case Amiga could be seen as a successor of Atari 2600 but not Commodore 64/128. Atari ST is a botchjob thrown together with Jack Tramiel in the lead, precicely like Commodore 64 before it. Does that make Atari ST the successor of Commodore 64? Probably not, but you might get a different answer if you asked Jack Tramiel. I know I am oversimplifying hugely here abd bringing other machines to the discussion. The reason for that is just to show that it's all very vague when it comes to early home computers, including, but not limited to ZX Spectrum.
 * We could go further, down the line, ZX80->ZX81->ZX82/Spectrum (or Pet->MicroPet/Vic20->MaxMachine/C64->C128) and there would always be some level of shared architecture, and in many cases even some level of compatibility (albeit not necessarily very extensive level) involved. I'm actually not sure to what extent does QL built on top of an old ZX architecture, but at least the processor is different. Amiga has nothing whatsoever to do with older Commodore machines, same as Atari ST has nothing to do with older Atari 8 machines. They are simply products that ended up being released by the same company as their claimed "predecessors", more often than not by the whim of destiny more than anything else. They are in-laws more than relatives and their real family is elsewere.
 * So is ZX Spectrum Next the successor of ZX Spectrum? Well yes and no. It does have the license for name and branding, it is software and hardware compatible, it has the original Spectrum ROMs and Rick Dickinson from the original ZX/QL team has been present in the development, so in that way maybe yes? However, while the Next can be seen to build on top of ZX architecture, it is still an FPGA implementation that incorporates much of the development done in the ZX Spectrum/Pentagon user community during the past 40 years. Predecessor/successor? Go figure... 2001:14BA:16F7:5700:E54B:DFB1:75E1:3F00 (talk) 13:31, 24 July 2022 (UTC)


 * Successors must be computers released back in the day, with some connection to the predecessor, and positioned as such on advertising / commercially. Usually they follow each other on the market (ex: computer A is discontinued at the same time computer B is introduced by the same or related company), occupy the same niche (ex: home/games computer that you connect to a TV) and have a similar price range (ex: low cost entry model). So for the Spectrum this would be the SAM (I favor this due to compatibility) and the QL (but the infobox might show both), but not any clones or modern devices. 4throck (talk) 15:42, 24 July 2022 (UTC)
 * But QL was in no way intended for the same audience as ZX Spectrum, nor was it anywere near the same price range. QL was meant to be Sinclair take on a business computer to compete with IBM and Apple products, while ZX Spectrum was clearly targeted to the home market for hobbyists, education and gaming. This also becomes apparent in the QL commercials of the time. Now I'm not saying that QL isn't a successor of ZX Spectrum, but your grantedly valiant attempt to define "successor" doesn't seem to work in this case. 2001:14BA:16DC:3300:9DEA:DBA9:6DA8:D0A5 (talk) 11:55, 30 August 2022 (UTC)

IEC instead of legacy units, e.g. 48 KB -> 48 KiB
Would it make sense to update the page to use IEC units (KiB, MiB, GiB) instead of legacy units (KB, MB, GB)? HenrikB (talk) 09:06, 16 July 2022 (UTC)


 * No. There are specific guidelines that say they should not be used, except in a few specific circumstances. See the last paragraph of WP:COMPUNITS - X201 (talk) 10:18, 16 July 2022 (UTC)
 * I'm just reading the guideline, and I almost can't believe it. A valid international standard is being ignored by Wikipedia.
 * And, the wording of the "guideline", it is not a guideline at all, it is a straight prohibition.
 * Where can I complain about this? Can someone point me to the proper venue for complaints? Z80Spectrum (talk) 18:59, 11 January 2024 (UTC)

The article is too long
The article contains too much information. It is confusing. Too much is duplicated, or it belongs to other articles. I'm going to work on trimming down the article, removing redundant and unnecessary information, and on moving the information into other pages. 80.80.52.211 (talk) 09:40, 11 January 2024 (UTC)


 * I forgot to log in... I added the topic. Z80Spectrum (talk) 09:41, 11 January 2024 (UTC)
 * The "Sound" sub-section is too technical for the article at this level.
 * It could, perhaps, be moved to another new article, perhaps "ZX Spectrum integrated peripherals" Z80Spectrum (talk) 10:27, 11 January 2024 (UTC)
 * Here are statistics and recommendations:
 * The current article size: 8,622 words
 * WP:TOOBIG recommends:
 * over 9,000 words: Probably should be divided or trimmed...
 * over 8,000 words: May need to be divided or trimmed; likelihood goes up with size.
 * I have only one important issue concerning that:
 * the "History" section is too long; there is more than sufficient material for a separate article; I recommend title "History of ZX Spectrum and compatibles"; the material from the current "History" section should form the basis of the new article, then the "History" section in this article should be summarized
 * Z80Spectrum (talk) 13:03, 14 January 2024 (UTC)

The three major issues (with the article)
In the discussions during previous two months, two more major issues have surfaced, so I'm just highlighting those issues here. These are only my opinions, and only advisory at this moment:
 * Issue 1. Article Size. The article is getting too long. The current size is 8,527 words. I suggest that the section "History" should be split into a new article, while only a summary should remain in this article.
 * Issue 2. Operation. I propose a section named "Operation", which would explain to the reader of the article how the machine was operated in practice, and what were the most common uses of the machine. Today's home computers are operated in a significantly different way compared to home computers from the 80's. I consider it highly likely that readers unfamiliar with operating the ZX Spectrum would find the article confusing and misleading, unless the "Operation" section is added.
 * Issue 3. History after 1992. Since computers compatible with ZX Spectrum were still widely used after ZX Spectrum line was discontinued, I propose expanding the section "Clones" with additional explanations about history after year 1992. It doesn't have to be very long, but the article is missing at least a few paragraphs about history of ZX Spectrum clones after 1992. Z80Spectrum (talk) 22:41, 27 March 2024 (UTC)
 * 8500 words is borderline, not obviously too long. I suggest tighter copyediting may be better than a pure size split at this point.
 * Expansion of operation should be limited per WP:NOTHOWTO, and general description of features common to 1980's computers (as opposed to unique to the subject) is probably best covered elsewhere.
 * Issue 3 seems only tangentially related to the subject of the article and probably should occur elsewhere. VQuakr (talk) 22:50, 27 March 2024 (UTC)
 * I don't agree with most of VQuakr's comments due to the following:
 * Issue 1b. (Article size / split of History) - The "History" section currently dominates the article. It is the first section, and it is very long (for a section). This article is not a history article, it is a technology article. But, for a technology article, the "Hardware" section is too short (and I'll likely be adding material to the "Hardware" section). I suggest that it would be advantageous to do the split sooner rather than later.
 * Issue 2b. (Operation / How To) - A description of operation is not a "how-to". A "how-to" is a detailed description instructing the reader how to do something. A description of operation is a general outline, an overview. For example, the sentence "cars are operated by turning the driving wheel and pressing the pedals" is not a how-to, it is a description of operation.
 * Issue 3b. (History after 1992.) - The article already includes some history after 1992., but talks mostly about irrelevant details and events related to minor companies (see the "Recreations" section). On the other hand, the article stays mostly silent regarding major developments in history regarding clones. In total, the clones probably sold in over 2 million units, and were continually in use after 1992. Z80Spectrum (talk) 14:28, 28 March 2024 (UTC)
 * Tighten the prose, add sourced content, and when it's stable then we can reassess if and where it makes sense to split. Splitting preemptively isn't a good approach. This is a historical computer; the computer's history is going to be a significant portion of the article. VQuakr (talk) 16:19, 28 March 2024 (UTC)

Casette tape speed
(At the end of "Firmware" section)

The preferred unit for the tape speed should be B/s.

The approximations given for casette tape speed are incorrect. The signal "0" takes 2×855 T-states, while "1" uses 2× 1710 T-states. From those values, the exact bitrate can be computed. I'm going to consider that this trivial computation is allowed by WP:CALC.

I'm going to find the sources for the T-cycle timings, and them I'm going to re-compute it into B/s. The biggest problem is finding sources, but I do have them somewhere. I'm just writing this as an overview of what needs to be done to improve the article. Z80Spectrum (talk) 10:52, 11 January 2024 (UTC)


 * Everyone is welcome to help me by finding the sources for the timings (If I remember correctly, it is 855 and 1710 T-states). Anyway, the timings must be in T-states, which are the same thing as Z80 clock cycles. Z80Spectrum (talk) 11:01, 11 January 2024 (UTC)
 * Here is the reference, it has already been added to the page:
 * https://sinclair.wiki.zxnet.co.uk/wiki/Spectrum_tape_interface Z80Spectrum (talk) 12:13, 11 January 2024 (UTC)
 * Therefore:
 * - "0" bit is transmitted at: 3500 kHz ÷ 2 ÷ 855 = 2.046783626 kHz
 * - "1" bit is transmitted at: 3500 kHz ÷ 2 ÷ 1710 = 1.023391813 kHz
 * Peak speed:
 * - "0" bits only: 3500 kHz ÷ 2 ÷ 855 × 1 b/Hz = 2.046783626 kb/s = 255.848 B/s
 * - "1" bits only: 3500 kHz ÷ 2 ÷ 1710 × 1 b/Hz = 1.023391813 kb/s = 127.924 B/s
 * Lengths:
 * - "0" bit: 2 × 855 ÷ 3500 kHz  = 0.488571 ms
 * - "1" bit: 2 × 1710 ÷ 3500 kHz = 0.977143 ms
 * Average Length of one transmitted bit:
 * (0.488571 ms + 0.977143 ms) × 0.5 = 0.732857 ms
 * Avarage speed:
 * 1 s % 0.732857 ms × 1 b = 1.36452 kb/s = 170.565 B/s
 * Computation ends Z80Spectrum (talk) 12:36, 11 January 2024 (UTC)
 * Now, it is obvious that the current value in the article is incorrect.
 * The error is a whooping 10% (result by old source: 1500 b/s; my result: 1365 b/s).
 * The problem is in the source of data - he incorrectly computed the harmonic mean.
 * Therefore, I'm going to correct the value in the article now. Z80Spectrum (talk) 12:39, 11 January 2024 (UTC)
 * I have completely missed that the newly provided link of mine actualy did the same computation as I did above.
 * All the values are the same, except that the source arrived at 1364 b/s, but I got 1365 b/s.
 * The source is incorrect and I am correct, because the more precise value is 1364.52, but the source has incorrectly rounded the value. Z80Spectrum (talk) 13:05, 11 January 2024 (UTC)

I suspect this is a mistake but don't know what the true figure is
"In 1981, Sinclair Research had spent just over £5 on advertising, whereas after the launch of the ZX Spectrum it was projected that they would invest more than £10 million"

Now I know I saw at least one full page newspaper ad for the ZX81, so I'm finding it improbable, even taking into account inflation, that the advertising budget was that low. Indeed, Sir Clive Sinclair's main selling method until the ZX81 and Spectrum was through magazine ads. So £5 seems... low? I mean, that wouldn't even have gotten you a classified ad in the back of Private Eye in 1981.

At the same time, I don't know what the true figure is. If it was a typo and meant to be £5M, then it seems like £10M the next year isn't the big deal the rest of the paragraph seems to imply it was.

Anyone know the right numbers? 2601:584:300:345E:7D19:F07E:F0C3:ABAC (talk) 20:59, 22 January 2024 (UTC)

The Spectrum is not the second most sold microcomputer
The PC is easily the most sold microcomputer. Commodore 64 second. Maybe Amstrad PCW third with 8 million sold. 178.74.0.136 (talk) 21:07, 6 March 2024 (UTC)

The leading paragraph
The leading paragraph currently claims: In addition to being one of the most influential computers ever made, the Spectrum is also one of the bestselling models of computer ever with over five million units sold.

Those are some overarching and grandiose statements. I would like them sourced, but @User:Chaheel Riens said that they are supported by the body of the article. However, I cannot find the support in the rest of the article. Also relevant might be MOS:LEAD, which says As a general rule of thumb, a lead section should [...] be carefully sourced as appropriate, although it is common for citations to appear in the body and not the lead. So, MOS is ambiguous, but it does require the lead to be supported by sources. So, where are those reliable sources that support the quite grandiose claims of the leading paragraph? I'm also pinging @User:Jaguar who recently edited the leading paragraph. Z80Spectrum (talk) 22:35, 26 March 2024 (UTC)
 * MOS is not ambiguous at all. It's perfectly clear:  As a general rule of thumb, a lead section should [...] be carefully sourced as appropriate, although it is common for citations to appear in the body and not the lead.  The lede is supported by sources because the lede summarises the article body which contains the required sources and references, as well as more detail on the subject.  The lede should not be used to introduce new material that has no mention in the main body.  Some contentious articles will have sources in the lede, but that's not a requirement.
 * This is a particularly weird thing for you to obsess over, given that you've just reinstated this phrase into the lede: - which is unsupported by the rest of the article.  Chaheel Riens (talk) 23:07, 26 March 2024 (UTC)
 * Thank you, User:Chaheel Riens, but I have asked: where are the sources? So, could you point me to the sources (mentioned in the article) that support the two grandiose statements in the first paragraph. I'm not saying that those statements are not true, but I don't see the sources for them. Z80Spectrum (talk) 23:33, 26 March 2024 (UTC)
 * You need to look harder, although not much harder as they're not exactly hidden. There are two clearly marked sources for the 5 million number and an entire - sourced - section commenting on the importance and legacy of the machine itself.  Chaheel Riens (talk) 07:58, 27 March 2024 (UTC)
 * Inserted an existing reference to "one of the most influential computers ever made". It is not a very reliable source, but at least someone has said it and published it.
 * Still can't find "one of the bestselling models of computer ever". It would be good if it is a reliable source, because the claim in the leading paragraph is quite significant. Z80Spectrum (talk) 08:30, 27 March 2024 (UTC)
 * @VQuakr : Hope that you won't mind me pinging you and dragging you into this discussion. You are under no obligation to participate.
 * Anyway, when reading the VQuakr's version (current version) of the probematic sentence in the first paragraph, I must say that I'm still not quite happy with retention of the vague phrase "one of the bestselling models of computer", and I think that adding the "five million" phrase doesn't quite fix it. Also, I don't like that there is no inline citation to those claims.
 * I suggest that those disputed parts should be re-phrased. However, I'm not a native English speaker, and at this moment I can't offer ideas about how to improve the phrasing. So, I'm open to suggestions.
 * Given my lack of a better ideas there, I'll probably let it stay as it is, untill I think of something better.
 * Currently, the sentence reads:
 * In addition to being one of the most influential computers ever made, the Spectrum is also one of the bestselling models of computer ever with over five million units sold.
 * - Z80Spectrum (talk) 20:20, 27 March 2024 (UTC)
 * As already discussed above, we generally avoid in-line citation clutter in the lead except for cases where WP:BLP is a factor or for direct quotes, things like that. I'm not seeing any verifiability issues with the sentence as it sits. Commodore 64 holds the title by a mile and the other ~5 million selling computers including Spectrum are quite reasonably described as "one of the". Anything is open for improvement, of course, but lacking concrete suggestions for a change it seems fine as-is. If we feel the existing sourcing and phrasing in the Success and market domination section is insufficient, I see plenty of sources available online so we could tweak a sentence a bit to better support the lead if someone found that important. VQuakr (talk) 21:23, 27 March 2024 (UTC)
 * Thank you for the explanation. I'll accept it in the sense that I now consider the disputed sentence as being of satisfactory quality. However, I still think that the disputed sentence can be improved, by some tweaking and re-wording. But, I'm not going to be the one who is going to change it further. Z80Spectrum (talk) 21:35, 27 March 2024 (UTC)
 * @User:Paradoctor I just wanted to mention that the reference you have added is not very reliable, because the list of the best-selling computers of all time would certaily include the Raspberry Pi (multiple models, some over 20 million units sold). Also, C-64 sold in about 13 million units; the 17 million figure is an overinflated number based on Tramiel's words. There are other problematic numbers in the cited source. Z80Spectrum (talk) 03:01, 28 March 2024 (UTC)
 * As you say, Raspberry Pi is diluted over various models. More generally, "the source is not reliable because I disagree with the information with it" is not a strong argument. VQuakr (talk) 04:16, 28 March 2024 (UTC)
 * ( WP:EDITCONFLICT )
 * Two issues here.
 * The source says "9 the best selling", not " 9 best selling". So the existence of one high seller not on the list does not in itself negate its correctness. Even finding a Top Ten list in which every single computer sold more than any of the 9 would would not invalidate the claim. The Spectrum would still be in the Top Twenty.
 * "Raspberry Pi" / "C-64" Sources. Always sources. Right now, this is just something you say, and therefore ineligible to influence article content.
 * WP:V: Even if you are sure something is true, it must have been previously published in a reliable source before you can add it. If reliable sources disagree with each other, then maintain a neutral point of view and present what the various sources say, giving each side its due weight. Tatto it onto your eyelids, because this you will encounter all the time.
 * Should you dig up reliable sources supporting your claims, which I have no doubt you can, a third issue will arise:
 * WP:SYNTHESIS: Do not combine material from multiple sources to reach or imply a conclusion not explicitly stated by any source.
 * Which means you need a reliable source that the Spectrum is not among the best selling computers in history, or equivalently, claiming to be comprehensive, but not including the Spectrum. In that case, we report on the discrepancy, as guided by WP:DUE weight. Paradoctor (talk) 04:26, 28 March 2024 (UTC)
 * Or, I can simply attempt to prove that the source is not a reliable source (I mean not just the website, but the linked page), which will probably be the path that I will take. However, I do believe that ZX Spectrum actually is "one of the best-selling computer models". I just want a source better than that faulty list at "interestingengineering.com". And I would like it re-worderd, but I'm not insisting on it anymore. Z80Spectrum (talk) 04:51, 28 March 2024 (UTC)
 * Or or, you can simply move on and quit wasting editor time on this since you don't think the statement is untrue and it really doesn't need a source in the lead anyways. VQuakr (talk) 04:55, 28 March 2024 (UTC)
 * I propose this compromise: the Paradoctor's source has to be removed, and the sentence stays as-is for the time being, until a better source is found, or until someone figures out a better way to solve this dilemma. Z80Spectrum (talk) 05:37, 28 March 2024 (UTC)
 * ( and another edit conflict )
 * prove that the source is not a reliable source Well, make sure to closely read what WP:RS has to say. You might save yourself some effort by raising your concerns at WP:RSN. If they say it's ok, then you have done all the due diligence anyone can ask of you.
 * re-word Well, "not insisting" does not sound like your satisfied. What's your concern there, if I may ask?
 * @VQuakr I put the reference in the lead because the claim was not in the article body. That's fixed now.
 * compromise My intervening edits mooted this. Let me know if you still see issues. Paradoctor (talk) 05:48, 28 March 2024 (UTC)
 * P. S.: I'm watching this page, so please don't ping me here. Paradoctor (talk) 05:52, 28 March 2024 (UTC)
 * I still don't agree with the current state of the article. 2 issues:
 * Iss1. An inline reference is needed for "It is one of the most influential computers ever made" in the lead. I suggest the inline reference that I added (and Paradoctor has removed it). It is not a stellar reference, but something is better than nothing there.
 * Iss2. The reference 225. interestingengineering dot com/innovation/9-of-the-best-selling-computers-of-all-time and all inline references to it have to be removed from the article.
 * When I say "inline references", I mean those numbers in superscript brackets, tagged by "ref". Z80Spectrum (talk) 06:18, 28 March 2024 (UTC)
 * Iss1 An inline reference is needed That is incorrect. The claim is not in contention, and per MOS:LEADCITE, the presence of citations in the lead is neither required in every article nor prohibited in any article [...] should be determined on a case-by-case basis by editorial consensus. So the decisive point is what the consensus here is. We have you in favor, and three naysayers, all experienced editors. This situation permits only two interpretations: either there is consensus, then it is clearly against. Or if, for the sake of discussion, one reads this as WP:NOCONSENSUS. In that case, we retain the version of the article as it was prior to the proposal or bold edit. Which is the one without ref.
 * If this doesn't do it for you, and you still want to pursue it, you'll have to start WP:dispute resolution. That's going to be a of work. And I can promise you that there is no chance that you'll find consensus in your favor. OTOH, it could be a good learning experience for you, if you manage to keep your temper. These things can be stressful. Trust me, I know.
 * Iss2 Hell no. The source is reliable and it is needed to support its claim. The only way this source will be removed is either replacement by a superior source, or removal of the claim. Which I will object to, as it is a salient feature of the article's subject.
 * Last but not least, you haven't said anything about the rewording you desired. Can I take this as being resolved? Paradoctor (talk) 07:36, 28 March 2024 (UTC)
 * Iss1b - What you are saying is not what I desired. I'll later explore whether Wikipedia has policies that would force a reference to be inserted. For example, most other articles have inline citations in the lead, and I don't quite understand why this article would be an exception. I need to examine some policies in more detail.
 * Iss2b - I'll dispute the reliability of the source at Reliable Sources Noticeboard.
 * Rewording - I liked the phrase "computer model" better than just "computer". Besides that issue, the new sentence is better than the old one.
 * Overall, from my point of view, you made the lead paragraph worse than it was before you started editing. Z80Spectrum (talk) 07:50, 28 March 2024 (UTC)
 * examine some policies in more detail [...] dispute the reliability of the source at Reliable Sources Noticeboard I'm happy to hear that. That's the Wikipedian spirit. :) Make sure to refer to WP:RS in your arguments, these people are scary.
 * I liked the phrase "computer model" better than just "computer" Feel free to change it, then. I prefer the shorter version, but I can see your point.
 * made the lead paragraph worse Believe it or not, I understand your frustration. Like Churchill said, "democracy is the worst form of government" Winston Churchill Paradoctor (talk) 08:16, 28 March 2024 (UTC)
 * Iss1c - WP:V states: any material whose verifiability has been challenged or is likely to be challenged, must include an inline citation to a reliable source that directly supports the material
 * I now challenge the phrase one of the most influential computers ever made from the first article paragraph. Therefore, an inline citation must be provided.
 * Regarding that same point, I also estimate that I have been deceived by Paradoctor's and VQuakr's interpretation of the situation so far, but it was probably an honest mistake on their part, so at this moment I'm willing to just forget it. Z80Spectrum (talk) 13:34, 28 March 2024 (UTC)
 * I now challenge the phrase That was never an option in this particular case for you. are the one who copied the statement to the lead, and  copied the ref there. It is obvious that your challenge is not about disbelieving the claim, but about you trying to get your way. What you're doing here right now is called WP:POINTY behavior, and puts you on a short track back to ANI. The only option for you that is in line with policy is to attempt to change, through WP:dispute resolution, the current consensus not to duplicate that ref in the lead.
 * I should point out that the momentary consensus in your ANI case is to indef block you. Which means you be blocked soon, unless something substantial happens to change that consensus. Being pointy now will only hurt your chances at appealing that block later.
 * Which brings us to I have been deceived. Stow it. Accusing us of deceiving you in this way is unacceptable. If you believe we have said something false, you point to evidence, quoting and linking to the relevant WP:DIFFs. That is something we can discuss. Making unsubstantiated accusations, OTOH, is WP:UNCIVIL, and will put you on a short track to being not merely WP:BLOCKed, but WP:BANned.
 * Now would be a great time to consider what you want to achieve on Wikipedia, because you're setting yourself up to making that impossible. Paradoctor (talk) 15:48, 28 March 2024 (UTC)
 * I now challenge the phrase... objection, your honor! Just no. I liked the phrase "computer model" better than just "computer". I disagree; it's a word that adds no value. VQuakr (talk) 16:24, 28 March 2024 (UTC)

Z80Spectrum, you additionally need to back off with variations of the phrase  I've seen you use this and similar phrases to imply that you're correct and the person you're interacting with is wrong, but you'll be magnanimous about it. As with practically everything else you're doing right now - this is borderline insulting, and is just one more spade digging your hole.

Staying germane to the point (although it's getting harder and harder to see the point) - I am against the term "computer models", because it's inaccurate. "model(s)" implies a specific version - or "model", if you will - of the ZX Spectrum range, rather than the entire range, and the 5 million refers to all models bunched together. Also - as per VQuakr, it's duplication and unnecessary duplication which is unnecessary. Chaheel Riens (talk) 17:19, 28 March 2024 (UTC)


 * imply that you're correct and the person you're interacting with is wrong, but you'll be magnanimous about it I think the word you're looking for is condescending. ;) Paradoctor (talk) 17:46, 28 March 2024 (UTC)
 * User:Paradoctor said: You are the one who copied the statement to the lead.
 * It was not me who added the disputed sentence to the lead. I just searched, and as far as I can tell, the disputed sentence was added by @Elliott-AtomicInfinity, in this diff:. Z80Spectrum (talk) 18:17, 28 March 2024 (UTC)
 * This is Paradoctor's statement that I find deceptive (Paradoctor is implying that the "consensus" is the same as "local consensus", which it is not): So the decisive point is what the consensus here is. We have you in favor, and three naysayers, all experienced editors. This situation permits only two interpretations: either there is consensus, then it is clearly against. Or if, for the sake of discussion, one reads this as WP:NOCONSENSUS.
 * This is another Paradoctor's statement that I find deceptive (where he implies that WP:V can be subverted by status-quo): the one reads this as WP:NOCONSENSUS. In that case, we retain the version of the article as it was prior to the proposal or bold edit. Which is the one without ref.
 * This is the VQuakr 's statement that I find slightly deceptive (when looking at other articles, I see many inline citations in the leading section, which is somewhat contrary to what VQuakr says here) : As already discussed above, we generally avoid in-line citation clutter in the lead except for cases where WP:BLP is a factor or for direct quotes, things like that.
 * As I have said, I still consider those to be honest mistakes, provided in good faith. Z80Spectrum (talk) 19:18, 28 March 2024 (UTC)
 * This is an WP:OTHERSTUFF argument that you should know better than to propose by now. Yes, many articles have unnecessary citations in the lead. Many articles have no citations at all, or are written in a non-neutral point of view, or violate guidelines or policies in many other ways. It is not in the slightest deceptive to point out a guideline, even if some articles do not adhere to it. VQuakr did not say that no articles have citations in the lead, they said that we generally avoid in-line citation clutter in the lead, which is exactly what the MOS says. CodeTalker (talk) 20:01, 29 March 2024 (UTC)
 * To reply to Chaheel Riens: I won't insist on "computer models" even the slightest. I was just stating my opinion, because Paradoctor asked me whether I agree. Also, I have a right to point out problems in other people's arguments and behaviors (before those get out-of-hand). It the same as VQuakr did here:.
 * (end-of-my-replies) Z80Spectrum (talk) 19:28, 28 March 2024 (UTC)
 * ( yet another edit conflict / old and slow I am )
 * I told you I'm watching here.
 * It was not me who added the disputed sentence to the lead.
 * First off, the sentence is not disputed by anyone. We're disputing inclusion / omission of a redundant citation.
 * As regards adding the sentence, I stand corrected. My bad. But that doesn't really change the situation.
 * Remember, I said Or, one reads this as WP:NOCONSENSUS. [emphasis added]
 * To avoid any misunderstandings: My position is that there is a clear consensus against inclusion of the cite. As I said above, if you want to challenge that, you'll have to start WP:dispute resolution. BTW, is also against inclusion, per this edit. And I wager, Elliott-Atomicinfinity wouldn't have added the ref had he known about WP:LEADCITE, and that the same claim was already cited to three sources in the article body.
 * It's hard to keep track of all the objections Can't help you there. Par for the course when you argue against clear consensus. Paradoctor (talk) 19:51, 28 March 2024 (UTC)
 * See WP:OTHERCONTENT for why we generally don't give much credence to reasonings based on comparisons with other articles. But enough of this distraction; let's move on to something else since consensus is clearly against you here and endless churning argumentation is going to go nowhere. No more pings on this topic, please. VQuakr (talk) 19:55, 28 March 2024 (UTC)
 * +1
 * Paradoctor is implying that the "consensus" is the same as "local consensus" I'm not implying, I'm saying. That is because, and I have already told you this, MOS:LEADCITE says the presence of citations in the lead [...] should be determined on a case-by-case basis by editorial consensus. That means first and foremost local consensus. This does in no way exclude wider consensus arrived at in the course of dispute resolution. If you desire that, go ahead. Until such time, local consensus decides. Paradoctor (talk) 20:02, 28 March 2024 (UTC)
 * I obiously don't understand the "pings", so, feel free to help me (here ).
 * I will seek the wider consensus on the disputed inline citation (probably at RfC). However, I needed to post at least some of my arguments here first. Z80Spectrum (talk) 20:06, 28 March 2024 (UTC)
 * he implies that WP:V can be subverted by status-quo Uh, you left the track there. I have no idea what that is even supposed to mean. As regards your quote, you seem to ignore the links to applicable policy I amply provide you with.
 * WP:NOCONSENSUS: When discussions of proposals to add, modify, or remove material in articles end without consensus, the common result is to retain the version of the article as it was prior to the proposal or bold edit.
 * So whatever you read into that, that's between you and Wikipedia policy, and has nothing to do with me. Paradoctor (talk) 20:10, 28 March 2024 (UTC)
 * don't understand the "pings" Then don't use them. Just write "Paradoctor:". There is no need to link to my page, it's found in every single of my signatures. Paradoctor (talk) 20:14, 28 March 2024 (UTC)
 * As I have said, I still consider those to be honest mistakes, provided in good faith.
 * Stop this shit. You have been told by two editors already. Saying this when no one has cast doubt on that casting doubt on it, whether you understand that or not. I point you towards
 * WP:FOC: comment on content, not the contributor
 * This is not optional or a recommendation, this is policy. Abide by it. Paradoctor (talk) 20:32, 28 March 2024 (UTC)
 * This is not optional or a recommendation, this is policy. Abide by it. Paradoctor (talk) 20:32, 28 March 2024 (UTC)

Inconsistency of "models"?
Two statements in the article seem inconsistent.

"Ultimately, the Spectrum was released as six different models"

"Amstrad's acquisition of the brand name saw the release of three ZX Spectrum models throughout the late 1980s"

I think the first statement possibly refers to the 16k, 48k, 48k+ 128k, +2 and +3. I think the second statement possibly refers to the +2, +3 and +2A.

If we consider the +2A as a distinct model, should the first statement say that there were seven models altogether? 147.161.166.255 (talk) 09:54, 9 April 2024 (UTC)
 * I think you are right, there is nothing to refute the +2A being a model in its own right. I've amended the statement in the lead. ♦ JAG  UAR   12:27, 9 April 2024 (UTC)

Look on these edjits, ye Mighty, and despair!
It's fair to say the current state of the Unofficial clones section is a bit of a cluster. It's more difficult to say exactly when that part of the article went off the rails. Perhaps one arbitrarily chosen derailment point might be this edit, which might have "broken the window" by someone adding an example clone – and doubtlessly many people have fond memories of their clone in particular, along with a home bias towards clones in their respective country. Over time, this subsection has devolved, with repeated additions of parenthetical inclusions inside parenthetical inclusions, and when attempts were made to resolve some of these stylistic problems, the cleanup made things worse, by disassociating enumerated example clones from the country in question. Many of these edits very distinctly appear to have been introduced by ordinary people from those respective countries, people with apparently somewhat limited proficiency when it comes to English grammar and style – no offence. I'm not saying this to knock any particular editor, country, clone, or country's Speccy clone market. I would agree, if these kinds of machines were "big in Japan" (or whatever the country), that's relevant. But I don't quite dare to tread, don't quite dare to come in and cut ad-hoc norms and conventions from whole cloth here – I'm just the editor who points at the silly section and says, "That's starkers." Somehow people are going to have to form some consensus as to how to refer to the possibly transnational phenomenon of Speccy clones in certain parts of the world, whether and how to reference individual countries, and whether and how to reference individual clone models – and how to do all that while maintaining some sort of coherency and consistency in style, grammar and general readability. If you can tackle that and not make things worse, where over time so many others have done just that, then kudos. —ReadOnlyAccount (talk) 18:15, 14 April 2024 (UTC)

"it is also one of the best selling computers ever"
This was sort of discussed in the DYK nomination above, but the sales of this computer are somewhat inflated in the prose. The only source for it being "one of the best selling computers ever" in the article is this article, which seems to have arbitrarily picked nine popular computers from the '80s (or thereabouts) and designated them "9 of the Best Selling Computers of All Time". I'm also not sure that this is a reliable source. For two of them, the number of units sold is "Unknown"! I find the claim that the ZX Spectrum is "one of the best selling computers ever" hard to believe, even though it's already an exceptionally vague statement.

Counterexamples: The modern Mac sells several million units a quarter. ThinkPads (arguably of many different models) have sold over 200 million units. Compared to this, is it really warranted to call the Spectrum "one of the best selling computers ever"?

Suggestion: Reword lead to "one of the best selling computers of the 1980s", or the DYK hook of "one of the best selling British computers". The current version is a stretch. Toadspike  [Talk]  11:19, 12 June 2024 (UTC)


 * Pinging @Elliott-AtomicInfinity and @Paradoctor, who added and sourced that sentence, as well as @DigitalIceAge, @Panamitsu, and @Jaguar for their work on the GA and DYK reviews. Sorry for pinging so many folks, but since it's a matter of opinion I don't want to go changing things without proper discussion. Toadspike   [Talk]  11:27, 12 June 2024 (UTC)
 * The claim is directly sourced to a reliable source. I previously explained that counterexamples are WP:SYN. And I also explained previously why the existence of counterexamples does nothing to actually invalidate the claim, and what would be needed for that to happen.
 * I'll recap:
 * The claim is directly supported by a reliable source. You can either that the source isn't RS, as opposed to making unsubstantiated claims about it.
 * Failing that, the best you can do is provide one or more RSes that contradict the one we have, and we then report on the lack of consensus in the relevant literature as guided by WP:DUE. I doubt you will find any source directly stating that the Spectrum isn't an all-time bestseller, so what would be needed are other such lists that a) claim to be comprehensive and b) do not include the Spectrum. B) alone is insufficient, because a Top Ten not including of the present list would at most relegate the Spectrum to the Top Twenty, which is still easily consistent with th claim. We cannot just pick nilly-willy some cut-off pleasing us. Defining that cut-off is the job of the literature. Paradoctor (talk) 13:16, 12 June 2024 (UTC)
 * The reliability of Interesting Engineering was discussed only once as far as I can tell, and they should be reliable. However, this specific piece is a listicle of 80s computers that claims to list "9 of the best selling computers of all time", while making no mention of how these were selected (aside from "nostalgia"). There's an extremely similar listicle from HowStuffWorks, which has not fared well at RSN, but hasn't actually been designated as unreliable. I'm concerned that we're treating this list cobbled together from goodness knows what (speculation? rumor? over 1/5 of the stats are "Unknown"!) as completely reliable – I think it isn't.
 * My attempt at finding more reliable sources turned up this Ars Technica article, which does not mention the ZX Spectrum. (Its source data is here and also doesn't list the Spectrum, so perhaps the data is flawed.) A Google Scholar search shows that this book calls the Spectrum "best selling" (page 186), and that's it.  Toadspike   [Talk]  15:48, 12 June 2024 (UTC)
 * The article is flawed not the data. If you look at the spreadsheet it shows that the best selling computers up until 1983 are actually "Others", and not the TRS, Apple, Altair or PET as the article implies. Stooob (talk) 19:20, 12 June 2024 (UTC)


 * I would excise that sentence in the Legacy section and replace it the sentence in the lede that starts, "The Spectrum is one of the best-selling British computers of all time ..." It is true that other models lines of computer have eclipsed the Spectrum many times over, but as far as British computers it is still one of the top sellers. DigitalIceAge (talk) 15:59, 12 June 2024 (UTC)
 * I agree that it should specifically be described as one of the top-selling British computers. In the DYK review process I was under the impression that the entire article said "British computers", so I really should have read the whole thing. ―Panamitsu (talk) 22:12, 12 June 2024 (UTC)
 * I remember reading that section to give it a light copyedit, but somehow my brain must've put "British" in between "best selling" and "computers". I've edited the sentence accordingly now. DigitalIceAge (talk) 22:25, 12 June 2024 (UTC)