Talk:Modem

Talk page cleanup

Several dozen discussions going back to 2006 were moved to Talk:Modem/Archive_1 after I read each one and concluded that they were irrelevant or already resolved. The ones remaining on this page are in my opinion still open issues which should be addressed.Gravislizard (talk) 15:55, 14 August 2020 (UTC)

Bonding modems

I started a section on bonding modems (aka inverse multiplexing modem). Their is more info I did not add in the link I provided in the references section if anyone wants to add to or improve my wording in the section. --Cab88 13:34, 6 March 2007 (UTC)

Rewrite

This article is hard to read and certain sections repeat eg 'V90' and 'Using digital lines and PCM'. The article gives greater coverage to voiceband modems than other modem types such as DSL modems, due to the many developments over the last 40 years. The descriptions of the variants of voiceband modems is now largely of historic interest. The article does not descibe the relationship between modems and networks, such as todays internet. For example a voiceband modem network typically used to be two identical modems over an analogue voice circuit but nowadays is typically a wide range of different 56k modems connecting over a digital PSTN to a shared central modem server. The section on 'Popularity' refers to using 56k voiceband modems to access the internet, obviously DSL modems are gaining in popularity. A better structure for this article is needed.John a s (talk) 23:23, 12 January 2008 (UTC)

I agree the article is repetitive. I have thought about eliminating most of the first section and merging it with the next section (about breaking the 9.6k barrier, data compression, et cetera). The only reason I did not do it sooner was because of time; maybe I will tackle it this weekend. ----- As for newer modems like DSL, it's probably worth adding a section immediately following the 56k section, but I would not eliminate the earlier stuff. POTS modems dominated for over 30 years; DSL is still just a baby in comparison. (And POTS modems still have uses; like in my hotel. A 56k is my only option.)

"For example a voiceband modem network typically used to be two identical modems over an analogue voice circuit but nowadays is typically a wide range of different 56k modems connecting over a digital PSTN" <---- The article does discuss this in regards to the transition from 33.8k to 56k. —Preceding unsigned comment added by Theaveng (talk • contribs) 20:43, 14 January 2008 (UTC)

Traditional TTY/TDD modem frequencies and format? Also, v.18

This topic somehow seems to have been omitted from the article. Some time ago, I read about "Wybrecht" and "Weinbrecht" modems; both were misspellings of the name of a born-deaf engineer, Robert Weitbrecht, who worked to establish TTY/TDD technology. (By all means, Google on his name!)

Apparently, TTY/TDDs were developed quite early compared to modems for use with computers, and their technical specifications differ in many respects from those of computer modems, although they operate on the same principles (FSK). There's more at {typical header}911broadcast.com/tech-emergency_technology_54.htm (a blacklisted Web site, unfortunately; do be careful!) for the details. There's a table there, although it doesn't specify start and stop bit timings.

Fortunately, modern technology makes compatibility between TTY/TDD traditional protocols and modern modem protocols affordable to many; the v.18 standard [1] specifies this. This standard also permits successful international use of TTY/TDDs.

I hope somebody else will pick up this thread and incorporate it into the article; I'm out of energy, for now. Regards, Nikevich (talk) 09:28, 1 February 2011 (UTC) Best regards,

I could not find support for the idea that they predate computer modems; by 1964 the computer modem had been around for several years. I did however add a section linking to the TDD page.Gravislizard (talk) 16:27, 14 August 2020 (UTC)

Multiple issues

This article has multiple ambiguities, errors and omissions.

Broadband

The terms "broadband" and "wideband" are time dependent; at one time they meant 56Kbps with a V.35 DCE/DTE connection. Shmuel (Seymour J.) Metz Username:Chatul (talk) 13:05, 12 April 2020 (UTC)

Agree here. Wideband, ultra-wideband, broadband are not entirely stable, they haven't stabilized in the same comical sense as Very high frequency, Ultra high frequency, Super high frequency.--Mvqr (talk) 11:44, 27 April 2020 (UTC)

DAA

The acoustic coupler was not the only option for connecting your own modem; you could use a conventional modem connected via a DAA. Shmuel (Seymour J.) Metz Username:Chatul (talk) 13:05, 12 April 2020 (UTC)

Am working on finding sources for this. Found reference in computerworld 1977, Vadic claiming theirs was the first part 68 device not requiring a DAA, so that caps the era, now I only need a reliable reference defining the function of the DAA, possibly its design, and hopefully when they began using them; otherwise can propose as early as Carterfone?Gravislizard (talk) 19:14, 13 August 2020 (UTC)

Found thorough, excellent references. Updated Data access arrangement, linked and ref'd in Modem. Thanks for calling this out, I admit I had not known about this whole business.Gravislizard (talk) 22:10, 13 August 2020 (UTC)

Dialing

Hayes was not the first to provide a modem with the ability to place calls; they were, however the first to do so on an inexpensive modem. A dial interface was part of, e.g., X.21. Also, automatic calling units were available much earlier. Shmuel (Seymour J.) Metz Username:Chatul (talk) 13:05, 12 April 2020 (UTC)

@Gravislizard:The Data Auxiliary Set 801A (Automatic Calling Unit)] and Data Auxiliary Set 801C (Automatic Calling Unit) were not third party. The linked documents are from 1965; I don't know whether that is when Western Electric (AT&T) started offering them or whether they are older.

By 1988 there was X.21, which did not require an ACU. Prior to that, I recall there being modems with autodial capability, but I don't recall whether they used the RS-232 secondary channel, a second RS-232 connection or something else. Shmuel (Seymour J.) Metz Username:Chatul (talk) 00:48, 14 August 2020 (UTC)

Integrated info about the above; adjusted assertions about the Smartmodem to ref the new section on ACUs, cited a book on X.21 & articles on Smartmodem appeal and uptake.Gravislizard (talk) 16:51, 14 August 2020 (UTC)

History

The History section discusses both dial-up and leased line modems, but it is subsidiary to Dial-up modem. Either it should be split or it should be promoted to a section of its own. Shmuel (Seymour J.) Metz Username:Chatul (talk) 00:17, 11 August 2020 (UTC)

A ref that may be of use for this topic: https://books.google.com/books?id=_z4EAAAAMBAJ&lpg=PA20&dq=hercules%20network%20card%20parallel&pg=PA20#v=onepage&q&f=false claims the first modem that can switch from leased to dial mode. Not sure that's true but it's a datum.Gravislizard (talk) 18:51, 17 August 2020 (UTC)

Synchronous modems

The article does not explain the difference between synchronous and asynchronous modems and it doesn't mention the ubiquitous 408A and 408B. Shmuel (Seymour J.) Metz Username:Chatul (talk) 13:05, 12 April 2020 (UTC)

I'm afraid I'm utterly defeated on the 408. I can find no mention of it whatsoever in any of my usual sources (google/google books, internet archive) - this isn't a typo, right?

Regarding sync/async, I am trying to find a citeable reference defining it.Gravislizard (talk) 16:56, 14 August 2020 (UTC)

Sorry, typo. That should be 208A and 208B.There are a few hits, but technical data are sparse. I was surprised to see that there are still people selling them. Shmuel (Seymour J.) Metz Username:Chatul (talk) 05:43, 16 August 2020 (UTC)

Oh, that's a relief - I actually ran this one by a friend who has extremely deep connections in Bell history and the only thing they could find was a modem that used *plain DTMF* signaling, and we didn't think that was it... haha! Yeah, the 208's were clearly ubiquitous and we should cover them.

Now, I admit that I personally do not know anything about the sync/async topic, and I am having trouble finding something to cite that explains it. The earliest publications that discuss modems that *I* can find are periodicals which assume the reader already knows this distinction; none of them explain it, and I can't find it in any books online. Do you know of any literature we can reference that clearly explains what synchronous and asynchronous modems are?Gravislizard (talk) 05:52, 16 August 2020 (UTC)

Sync modems have a clock signal (either on the 4-wire line, or recovered from the signal) that rely on the clock signal to decode the incoming stream. The synchronous modem is designed for applications where you have a constant bit-rate (e.g. audio encoding without fancy compression). In the more modern standards (late 80s? 90s for sure), async modems anyway recover a clock from the underlying data so the technical difference (of having an external clock or clock on leased line wires) are a bit moot, but there is a difference in terms of input/output to the modem (whether you get a constant bit stream, or are encoding start/stop over the serial (or equivalent ISA/PCI interface)).--Mvqr (talk) 10:26, 17 August 2020 (UTC)

That's pretty much what I figured, thanks for confirmation. Now we just have to find somewhere that anyone clearly explained this.Gravislizard (talk) 15:09, 17 August 2020 (UTC)

Synchronous dial modems like the 208B were 2-wire. Shmuel (Seymour J.) Metz Username:Chatul (talk) 12:59, 1 February 2021 (UTC)

There's a brief mention of synchronous communication over a leased line in the Binary Synchronous Communications article. Perhaps some of the references from the synchronous serial communication article have more details?

a modem that used *plain DTMF* signaling --

I feel this article *should* mention such "touch-tone modems".

The "TV Typewriter Cookbook" reference in the article (on p. 178) mentions "touch-tone modem" and also mentions "the Bell 401L or 402C systems that offer 99 or 256 characters", but my online searches are having trouble finding any more details for "Bell 401L" and "Bell 402C" modems or "touch-tone modem". Is there a better reference? --DavidCary (talk) 06:37, 1 February 2021 (UTC)

56k section rewrite

I found the 56k / v.90/v.92 section completely incomprehensible and rewrote it. The first paragraph appeared to be passages on several different topics that had been pasted in between each other, rendering the whole thing impossible to follow.

I split out the early 56k and standards-based 56k modems into separate subsections, organized the info and clarified dates, and most notably *removed* the passage on telco digital line restrictions. It's perhaps relevant, but is also perhaps a bit too specific to mean much in this context.

I originally stated here that the other page, "56kbit/s line," was a "main article" on the topic of 56k modems since it was linked as such, but after reading it it appears to be about an entirely different subject (essentially a stub that should probably be rolled into the DS0 article), so I've removed that link from this article as irrelevant, and am currently composing a summary of what the digital signaling paragraph was trying to say, based on references. Gravislizard (talk) 21:21, 29 June 2020 (UTC)

User:Gravislizard, it is an improvement but not entirely accurate. You still have loss from D/A conversion on the downstream, as the digital PCM stream is 8bits at 8000 times a second, 64,000 bit/second while the modem is 56k (even with the best local loop, often less), contrast to Integrated Services Digital Network which gives you the whole 64k (or 64k *2 or more channels). However, you have better control over the D/A encoding side, as while this is non-linear resampling you can control the output wave better than on the A/D side (because you always have varying gain on the sampler). A 56k modem, on the half-digital down stream side, does take into account the a/mu-law encoding in a manner a 33.6/28.8K modem couldn't, but it still loses information from the conversion. Also I am not sure there is much of a point in discussing modem prices. Every new modem generation cost double, triple, and sometimes even x5 from the previous generation when just launched. Typically, however, 2-3 years after launch prices would fall well below where the previous generation was. So, sure, 56k was 200$ a card when it came out. But so was V.34, V.32, V.32bis, and even the 2,400 baud generation vs the 1,200 baud.--Mvqr (talk) 12:52, 30 June 2020 (UTC)

User:Mvqr, that's valid. My primary goal was to reorganize the info that was there and add the bare minimum needed to get the core concept of the "56k problem" across, while removing as little as possible - but these issues are common to all modem designs, really; basically every one has to deal with these same hurdles in some way, even if it's not as important as it is in the 56k technologies, so I felt the most important point to highlight was why the digital connection from the ISP to the telco was critical. I had found that information repeated in many places but never really explained. The referenced page from US Robotics, explaining the X2 technology, is the first time I've found anything that more or less states "it's because A/D is functionally lossier than D/A" - and even then it's only hinted at, otherwise I'd have made a clearer statement.

Regarding the prices, I was leaving those in (they were there in the previous version) because it felt relevant to have a notion of how much more expensive it would have been to "early adopt" 56k (and then, perhaps, find out you'd bought in to the wrong technology) - but that context isn't valid unless the later standards-compliant modems were cheaper when they first hit the market, and now that you mention it I don't know if that's true. If I can find evidence to that point I may add it. Gravislizard (talk) 15:31, 30 June 2020 (UTC)

The digital connection is critical for the downstream of 56k and sets it apart from all previous standards, yes. I was just commenting there was still a loss. As for price, I don't know if there is a good source for historical modem prices, but I do recall pricing myself. Generally, pre-standard modems (something that also happened with 9600, 14.4, 28.8 (V.Fast prior to V.34, there were also a bunch of 19.2K and 24K modems), etc.) were expensive. Modems at standard roll out (and even the first year) would generally be expensive. But then prices would fall. One-two years after rollout, the new standard modems usually were the same or lower cost than what "old" modems used to be. The "old ones", at that point, were usually out of production. At least this was the sequence in the 90s. The 1200/2400/9600 baud transition may have been different pricing wise, I recall those less.--Mvqr (talk) 12:03, 1 July 2020 (UTC)

Automatic calling units

Shouldn't the article mention automatic calling units such as the 801A and 801C? Does anybody have the dates for the 801A and 801C? Shmuel (Seymour J.) Metz Username:Chatul (talk) 12:34, 10 August 2020 (UTC)

"Early" 56k was decades after the fact

56k modems were in use for decades before the modems mentioned in Modem#Early 56k. The section should be called something like "Early 56k dial modems".

There really ought to be a section on the 4-wire modems in at least the range 1200 bps to 56k bps, e.g., 202C, 408A. Shmuel (Seymour J.) Metz Username:Chatul (talk) 12:50, 10 August 2020 (UTC)

If you look at the table of contents, that section is in fact within another section called "Dial-up modems." I agree with your complaint, that this info isn't represented here as it should be, but I think we should take a broader approach to solving it.

This page, as-written, is almost entirely about dial modems, yet the page is not called "Dial modem" - that's simply the bulk of the content that has been written. There are other sections for broadband and radio, for instance, and I see no reason we shouldn't begin a new section on leased-line modems (I believe that's the technology you're describing) and maybe another on short-haul serial modems, and so on.

This work may justify moving "dial modem" onto its own page, since it's such a large topic and will be disproportionate to the rest of the page eventually. I encourage you to start the leased line modem section if you have the chops to write a reasonable summary of it - I'm vaguely aware of the technology but would need to do substantial research to kickstart this.Gravislizard (talk) 23:04, 10 August 2020 (UTC)

My experience with 4-wire modems is strictly from a software perspective; the telephone companies handled everything from the RS-232 and V.35 connectors out. I can't address modulation techniques and such, and I don't even have spec sheets on such common modems as the 208A, 208B and 212A, although I've used all three. I could put together a summary of leased line modems, but it would just be a stub.

In addition to not saying much about leased line modems, the article doesn't say much about synchronous modems. Every shop that I worked in during the 1970s and 1980s had those, including synchronous dial modems. Shmuel (Seymour J.) Metz Username:Chatul (talk) 23:34, 10 August 2020 (UTC)

Notes

Sure thing, I'll do some reading and see if I can put together a safe, conservative summary of the technology so we'll have some coverage, and then if you have some details to submit once the framework is there you can do so. I'm already trying to clean up the overall layout of the article and have made mention of leased line modems at the beginning of the dial-up section.Gravislizard (talk) 23:56, 10 August 2020 (UTC)

Thanks. What are your thoughts on #Broadband and #Automatic calling units? Shmuel (Seymour J.) Metz Username:Chatul (talk) 02:21, 11 August 2020 (UTC)

I rewrote the Broadband section earlier because I felt it didn't make the necessary points clear, I don't know if you've seen it since then but I think it's improved. I'm actually not sure what an automatic calling unit is and I can't find any info on google, what's the synopsis?Gravislizard (talk) 03:00, 11 August 2020 (UTC)

An ACU allows a computer to place a call. I believe that the 801A and 801C connect over an RS232A interface.

There's a table of old modems at http://bitsavers.org/communications/westernElectric/modems/Bell_Modem_Types_Oct69.pdf that you may find of interest. I just found out about http://bitsavers.org/communications/, and it's been helpful, but it's missing the modems I'm most concerned about. Shmuel (Seymour J.) Metz Username:Chatul (talk) 03:48, 11 August 2020 (UTC)

Modem versus router

Modem#Terminology states The term "modem" is often partially inaccurate when applied to broadband.. However, a modem that performs other functions is still a modem. Further separate modems and routers are common in, e.g., cable modems, FIOS. I certainly agree that there are a lot of combined units in use, but I don't believe that the term inaccurate is justified. Shmuel (Seymour J.) Metz Username:Chatul (talk) 19:04, 11 August 2020 (UTC)

Well, would it be accurate to call a car's entire drivetrain "the engine" just because it does also include that component? I believe that a modem by definition is a "media converter," something to produce an analog signal from a digital one while making no alterations to the content, while broadband gateways almost universally do that, at a minimum by encapsulating the traffic.

I can say that from the perspective of many years as an ISP employee, this usage has a reductive implication: a person who thinks of their internet gateway as a "modem" frequently perceives it as a simple "converter" with no intelligence of its own, when in fact that's very rarely true, and this negatively impacts their understanding of how the device affects their internet traffic.

We could say that this is a de facto usage and that Wikipedia isn't the place to editorialize about how language should develop, but I think it's reasonable to comment on a vernacular term __if__ that term is specifically misleading, and if its use has not become formally "correct." After many years of occasional usage in this way, Websters still does not define 'modem' as anything other than a device for modulation/demodulation. I think as an encyclopedia we should uphold the de jure definition, especially when the de facto one can cause specific misunderstandings.

All of this of course is somewhat moot if we can't find a reliable source to back this (or much of anything on this page) up, though I don't know how we could even do that for this specific subject.Gravislizard (talk) 20:14, 11 August 2020 (UTC)

Thinking about it, I feel I should concede that this terminology is nearly universally used by residential ISPs. "Cable modem" and "DSL modem" are almost never simple modems, yet I think the overwhelming majority of "plastic boxes that sit at a subscribers house to provide internet" are referred to by those terms by the ISP. I think this is wrong, but calling it out in the article is maybe inappropriate because the de facto usage is not only by inexpert consumers, but by an enormous number of organizations. I think this is just one of the perils of trying to formally document terminology that is largely defined by marketing decisions.Gravislizard (talk) 20:14, 11 August 2020 (UTC)

Would you say that X.21 DCE units are not modems? They certainly do encapsulation. For that matter, is an RS-232 DCE unit a modem if it also does dialing? Are they not both modems from an ISP perspective?

The statement For several reasons, broadband modems are almost always full-fledged network routers. is also dubious even for cable and incorrect for FIOS. Verizon installs an ONT and you need to either provide a router or obtain one from Verizon.

I think if there's any dispute over the neutrality of a position then we should go with a safe option. What do you think of a milder approach of the subject, like "The term 'modem' is used interchangeably with 'router' and 'gateway," though these terms are often more specific when used in a technical context. 'Router/gateway' is often used to emphasize that a device contains higher level network functions like Wi-Fi and security in addition to the basic modem function."

The reason I think it should be included at all is simply that there *are* confusions that occur, and a person who reads this article should be aware that all three words can mean the same thing or distinct things in different contexts. Do you think this is neutral enough, with the assertion that the usage is *incorrect* now removed?

I think my 'almost always' assertion adequately describes the *majority* of broadband modems. I've worked with thousands of cable modems in my career and I'd say less than a percent of them were pure modems, with no presence on layers above the PHY. They exist, certainly - devices with an ethernet port and a cable jack and absolutely no configurability other than perhaps line parameters, but how many of those have been deployed versus the enormous number of cable gateways? That said, I doubt any statistics on this are available, so should I change this to something less definite like a simple "many devices"?

Finally, I agree with you on the fiber point. I missed on that one. It's true that basically all fiber service uses a discrete ONT which is inarguably a modem; I have never seen a fiber drop from a provider connected directly into a network device unless it was a port in a multiservice router, which I think is outside the scope of this largely consumer focused article. Thanks for catching my mistake!Gravislizard (talk) 23:18, 11 August 2020 (UTC)

I made some adjustments, let me know if that seems more fair.Gravislizard (talk) 02:03, 12 August 2020 (UTC)

It looks much better. Would it be TMI to mention that combined ONT/router units are not currently available? What about a brief explanation that the usage of the term brodband has changed over the years and that the article uses the current FTC definition? Shmuel (Seymour J.) Metz Username:Chatul (talk) 16:08, 12 August 2020 (UTC)

Glad to hear it. Regarding the ONT/router, if I google that phrase I find consumer-grade devices that have direct optical ports on them as well as multiple LAN ports and wifi hardware. While I can't say whether any ISP deploys them, they certainly appear to be available, so we can't state that they don't exist. I think we should stay mute on that issue in the article unless we can find some kind of cite-able data stating how many direct fiber subscribers are receiving plain ONTs versus full-package gateways, even if my personal experience suggests it's heavily skewed towards plain ONTs. Regarding 'broadband', the word itself is linked at the beginning of the section to the page Broadband, where the topic of the terminology is addressed, so I think we can let it speak for itself if a reader wants clarifying details.Gravislizard (talk) 16:42, 12 August 2020 (UTC)

• A modem is a modulator/demodulator converting digital data to the underlying analog medium - and doing the reverse (demodulating) on the analog input. Many broadband modems (e.g. those with ethernet interfaces) also do basic routing, and you can buy routers with VDSL modems included - but the functions are separate even if on one device.--Mvqr (talk) 13:38, 14 August 2020 (UTC)

Encoding for Western Electric (Bell) Modems

I'd like to add some synchronous modems to the table in #Evolution of dial-up speeds, but don't know what modulation techniques they use, or even the baud rates. Also, I know that there was a 2400 bps DataPhone circa 1962 but have no idea what the model or modulation techiques were. Can anybody fill in the gaps? Thanks.

Vendor	Model	S/L/W	bps	baud	Modulation
Western Electric	202C1	Switched	1200	?	?
Western Electric	202C2	Switched	1200	?	?
Western Electric	202D1	Leased Line	1200	?	?
Western Electric	202D2	Leased Line	1200	?	?
Western Electric	201A3	Leased Line Switched	2000	1000	PSK
Western Electric	201A4	Leased Line Switched	2000	1000	PSK
Western Electric	201B1	Leased Line	2400	1200	?
Western Electric	201B2	Leased Line	2400	1200	?
Western Union	2241A	Wideband	2400	?	?
Western Electric	208A	Leased Line	4800	1600	DPSK 8 phase
Western Electric	208B	Switched	4800	1600	DPSK 8 pase
Western Electric	TELPAK A	Wideband	19,200 40,800 50,000 230,400	?	?

Shmuel (Seymour J.) Metz Username:Chatul (talk) 15:50, 12 August 2020 (UTC)

@Gravislizard and Mvqr:I found some data on Jack L. Douglass (1984), Applied Data Communications Handbook (PDF), Universal Data Systems, Motorola Inc.; bitsavers.org/communications/universalDataSystems/ looks like a useful resource. Shmuel (Seymour J.) Metz Username:Chatul (talk) 21:56, 20 August 2020 (UTC)

This looks like a fantastic resource, I'll try to get it digested as soon as I have time. Great work!Gravislizard (talk) 16:31, 21 August 2020 (UTC)

Fax modems can't do efax

Fax modems can certainly transmit and receive documents that never were and never will be on paper, but [[efax]] redirects to Internet fax. Either the text here should be changed or Internet fax should be changed to define the term efax in a general fashion. Shmuel (Seymour J.) Metz Username:Chatul (talk) 01:13, 13 August 2020 (UTC)

Concur, I hadn't checked that page before I linked it. Having thought about it, I'm not sure efax is ever taken to mean the non-internet variant. I'll remove the link.Gravislizard (talk) 02:06, 13 August 2020 (UTC)

I recall efax used for paperless fax, i.e. fax reception via software and dialup modem. There actually was a software package called... EFax for use with modems: [2][3][4]. The internet fax probably appropriate the term (instead of using IFAX).--Mvqr (talk) 13:43, 14 August 2020 (UTC)

Proposed split

Given the size of the dial-up modem subsection, I suggest we split it off to Dial-up modem and remove the redirect from that page. It's certainly enough content for its own article, and adding even more information (possible, given the enormous history of dial-up) could become more practical if we did this. I believe this article would retain enough content to stand on its own, would be more accessible without the enormous preface on dial-up, and would make room for more specific information on the general topic. I think this qualifies on its own merits, but I'm a relatively new editor and would like some consensus. Thanks!Gravislizard (talk) 02:16, 13 August 2020 (UTC)

I agree, but please read WP:SPLIT for the procedure to follow. . Assuming that there is a consensus on splitting it, there is an issue of granularity. My first take is to retain modem as a summary, with {{Main}} for each type of modem in the appropriate section. One possible breakdown is:

• modem (Switched)
• Modem (Leased line)
• Modem (Wideband)
• Modem (Internet)

There is a case for either more granularity or less, so this is definitely something that needs to be hashed out as part of arriving at a consensus. Of course, if there is a consensus to not split then the question of granularity can be ignored. Shmuel (Seymour J.) Metz Username:Chatul (talk) 02:39, 13 August 2020 (UTC)

Dial modem history section

I've been working on cleaning up the history of dial modems but it's an enormous task. It appears to have been largely unchanged since 2007, but has always been a huge and somewhat unencyclopedic essay. I don't doubt that much of it is valid, but I feel the tone is a bit too informal, and much of it is I think very hard to source.

Assertions like "The 103A2 gave an important boost to the use of remote low-speed terminals such as the Teletype Model 33 ASR and KSR" or "Three U.S. companies became famous for high-speed versions of the same concept" seem very hard to prove (and indeed are uncited) but I'm reluctant to go through with a machete and rip them out wholesale.

Consequently, I've been doing a lot of medium-size edits that don't necessarily improve the quality of the article vis a vis the unverified facts, but attempt to clean up the information that's already here. For instance, I moved the subsection on acoustic couplers (and the one I created from it, on direct-connected modems) out of the history section because these are not specific points in history (as per the existing text which states that both were used concurrently,) and now those sections can be cited without depending on precise positions in the timeline, which is far easier.

I'm not sure that "history" is the best way to present all of this information. It may be more practical to make this article conform to Wikipedia standards by cutting back the history section to only cover the early era of modems (sage > bell 103 > development of third party products) and perhaps their rise and fall (increasing use by microcomputer users, then replacement by broadband) and reduce the rest of the history section to a simple breakdown of the various standards in a basic chronological order.Gravislizard (talk) 16:10, 13 August 2020 (UTC)

I rewrote the entire history and simply split it into decades, since things like "moving past 9.6 kb/s" were basically impossible to prove, given the enormous number of proprietary modems doing things that did not become standardized. I removed some information that I felt was irrelevant, as well as everything that seemed to be about leased-line modems, which are their own section where that info should now be reinstated and expanded. The goal is that what remains (which still needs trimming and sourcing) should now be about particularly notable events, and plenty of other interesting info (that was removed or has yet to be added) does not really need to be in a historical timeline format.Gravislizard (talk) 01:59, 14 August 2020 (UTC)

The 4800 bps 208B would seem to contradict The limited frequency range of telephone lines meant the symbol rate of modems was still only 600 baud.. That would have required 256 symbols, which would have been unheard of at the time.

BTW, there was a thriving market in modems interoperable with the 208A and 208B; often a single product allowed user selection of 208A and 208B mode. Would discussing that be TMI? Shmuel (Seymour J.) Metz Username:Chatul (talk) 02:28, 14 August 2020 (UTC)

I didn't scrutinize that line closely. I feel we should nix the line in question completely since I doubt we'll find a source for it.

I think mentioning the market in intercompatible modems fits well. I feel, basically, that the events that qualify for the history section are standards (and de-facto standards) that saw widespread use, as opposed to ones that did not provably become used outside of one manufacturer or model. If the 208's were cloned then yeah, I think it totally qualifies. As you may have seen I'm using a lot of references to Computerworld and I expect I'll be able to find some citeable examples there.Gravislizard (talk) 03:02, 14 August 2020 (UTC)

Where is the archive for CW and is it indexed? I need to find the dates for the 208A and 208B. Thanks. Shmuel (Seymour J.) Metz Username:Chatul (talk) 10:55, 17 August 2020 (UTC)

You can either use books.google.com (query like "computerworld 208a," although it will show non-CW results) or go here https://archive.org/details/computerworld and set the search field on the left side to "Text contents." Good luck, either one takes a ton of digging.Gravislizard (talk) 15:04, 17 August 2020 (UTC)

Null modem nomenclature

@Gravislizard:I've seen null modems listed as RS-232 crossover cables more often than as null modems; shouldn't the article at least mention the term?

BTW, after a search I was shocked to see that they are still being sold. Shmuel (Seymour J.) Metz Username:Chatul (talk) 02:47, 14 August 2020 (UTC)

Not sure I've ever personally seen it, but the terminology is valid on its face. Is there a citable reference handy? By the way, they're still heavily used in IT work - serial ports on routers et al are still very important and you never know when something will be DCE instead of DTE because it's expected to be plugged into a terminal server or modem instead of directly into a terminal for maintenance.Gravislizard (talk) 03:14, 14 August 2020 (UTC)

RS-232 is going to survive us all. It is the easiest interface to slap on a chip. Do they sill use the null modem term? Back in the day it was used because you were doing RS-232 without the modem attached to it, but RS-232 has soldiered on while RS-232 connected modems are a thing of the past mostly.--Mvqr (talk) 13:46, 14 August 2020 (UTC)

Conditioned lines

Some dial modems required a conditioned line. Shouldn't there be a discussion of that in the article? I'd suggest that the details be held to a short paragraph plus a list of modems. Shmuel (Seymour J.) Metz Username:Chatul (talk) 10:02, 23 November 2020 (UTC)

TrailBlazer is missing

I think the Telebit TrailBlazer is an important part of dialup modem history and I'm surprised that it is missing. There is a good description of the technology on its own page: https://wiki.riteme.site/wiki/Telebit Could someone knowledgable add some content here based on that? — Preceding unsigned comment added by 86.171.61.52 (talk) 19:19, 6 September 2022 (UTC)

Is ONT a modem?

This point is made in regards to the discussion regarding is an ONT a modem? Whilst they perform a similar function, modems are typically devices that transmit signals over copper based networks whilst an ONT is used in fiber optic networks. An ONT is not a modem because fiber has no carrier wave. This is not to be confused with a fibre optic modem (FOM).

A modem is a modulator/demodulator, converting between analog and digital. There is no requirement that the analog medium be copper. In fact, the document that you referenced says ONT is a fiber-optic modem. -- Shmuel (Seymour J.) Metz Username:Chatul (talk) 16:54, 16 April 2024 (UTC)

References