Teardown Of A 50 Year Old Modem

A few years ago, I was out at the W6TRW swap meet at the parking lot of Northrop Grumman in Redondo Beach, California. Tucked away between TVs shaped like polar bears and an infinite variety of cell phone chargers and wall warts was a small wooden box. There was a latch, a wooden handle, and on the side a DB-25 port. There was a switch for half duplex and full duplex. I knew what this was. This was a modem. A wooden modem. Specifically, a Livermore Data Systems acoustically coupled modem from 1965 or thereabouts.

The Livermore Data Systems Modem, where I found it. It cost me $20

The probability of knowing what an acoustically coupled modem looks like is inversely proportional to knowing what Fortnite is, so for anyone reading this who has no idea what I’m talking about, I’ll spell it out. Before there was WiFi and Ethernet and cable modems and fiber everywhere, you connected to the Internet and BBSes via phone lines. A modem turns digital data, in this case a serial connection, into analog data or sound. Oh yeah, we had phone lines, too. The phone lines and the phones in your house were owned by AT&T. Yes, you rented a phone from the phone company.

90s kids might remember plugging in a US Robotics modem into your computer, then plugging an RJ-11 jack into the modem. When this wooden modem was built, that would have been illegal. Starting with the communications act of 1934, it was illegal to attach anything to the phone in your house. This changed in 1956 with Hush-A-Phone Corp v. United States, which ruled you could mechanically attach something to a phone’s headset. (In Hush-A-Phone’s case, it was a small box that fit over a candlestick phone to give you more privacy.)

The right to attach something to AT&T’s equipment changed again in 1968 with Carterphone decision that allowed anyone to connect something electronically to AT&T’s network. This opened the door for plugging an RJ-11 phone jack directly into your computer, but it wasn’t until 1978 that the tariffs, specifications, and certifications were worked out. The acoustically coupled modem was the solution to sending data through the phone lines from 1956 until 1978. It was a hack of the legal system.

This leaves an ancient modem like the one sitting on my desk in an odd position in history. It was designed, marketed and sold before the Carterphone decision, and thus could not connect directly to AT&T’s network. It was engineered before many of the integrated chips we take for granted were rendered in silicon. The first version of this modem was introduced only a year or so after the Bell 103 modem, the first commercially available modem, and is an excellent example of what can be done with thirteen or so transistors. It’s time for the teardown, so let’s dig in.

Previous works and references

The modern history of this modem goes back almost a decade, to a single video from [phreakmonkey] who was given a Livermore Data Systems Model A modem by the widow of an IBM engineer. This eventually became a fantastically popular video on YouTube, and a talk at Defcon 17.

Items of note from [preakmonkey]’s demo of his modem is that his was an extremely early model, with a serial number in the 200s. So far as I can see, his wooden box is made out of walnut, with dovetail joints. According to [preakmonkey], the type of joint on the wooden box will tell you how old it is; dovetails are labor-intensive, and the first production units were simply not optimized for production. Later units, around serial number 850, used teak with box joints that were much more suited for mass production than dovetails. Still later units, like mine, used a rabbet on all the joints. This is a clear optimization in the production process that allowed Livermore Data Systems to crank out more modems faster.

This modem has been dissected before; in 2007, [Brent Hilpert] found one of these modems in surplus, and documented the internals. The schematic is especially interesting. There are thirteen individual transistors in this modem, all a mix of relatively standard, contemporary PNP and NPN general purpose transistors. There is one Germanium PNP transistor which is there for reasons I cannot understand, but for the most part these components are standard, off-the-shelf items that are still available as New Old Stock. You can buy all the transistors to make this modem today for under twenty dollars. If you want to build your own acoustic modem, you’re going to spend more on the tiny transformers.

The Actual Teardown

Apart from the lack of an acoustic coupler, this modem is more or less exactly what the documentation and references told me. There is a backplane of sorts, holding three cards. One card holds the power supply (less transformer), another card holds the modulator, and another holds the demodulator. The transistor date codes, specifically for the 2n5138 transistors are from the 37th week of 1969. Nice. This puts the date of manufacture for the transistors in September of 1969, and the date of manufacture sometime after that. Unfortunately, there’s no way of narrowing down the vintage of this modem any further, but it’s entirely possible it was built and shipped before 1970.


Is This Thing Worth Repairing?

My initial draw to this modem finding it at a ham swap meet was simply to own it. It’s a wooden modem, and I don’t think anyone I know respects exactly how cool that is. It’s on objet d’art, and it’s useful to store various sundries. I have considered repairing or refurbishing this modem, however there are a few things that make this impractical.

First, there is no acoustic coupler. This, I believe, is the reason why no one at the swap meet realized what it was. The Livermore Data Systems emblem and serial number was also adhered to the acoustic coupler, and without that, you would really have no idea what you were looking at unless you had an intimate knowledge of the garbage pile of a warehouse in a vintage computer museum. But I could easily 3D print an acoustic coupler, and that might be tempting.

Secondly, this modem is completely untested and half a century old. I guarantee the caps will need to be replaced, and the carbon composition caps are out of spec. I would need to reengineer the entire device to repair it, and there are better things I could do if I wanted to make my own modem.

Finally, if I wanted to engineer my own modem, I could just build a data toilet. This is a German / Chaos Computer Club project from 1985, built in response to modems being highly regulated by the Deutsche Bundespost. From a design standpoint, the data toilet is much simpler and much more capable. It’s built around an AM7910 chip that supports up to 1200 baud, and these chips are somehow still available from the usual online retailers. If I wanted to build a modem from scratch, this is how I would do it, not with discrete transistors.

Unfortunately, it’s just not worth the time to repair this ancient wooden modem, but a good teardown does give some great insight into how things were built before there were integrated circuits in everything. It’s an object of curiosity, or at least it is until I can find an acoustic coupler for it.

82 thoughts on “Teardown Of A 50 Year Old Modem

  1. Brian, sorry I missed you at the W6TRW swap meet.
    Perhaps I’ll see you there tomorrow.
    Yes, I live just a modem squeal away fro the W6TRW swap meet, and rarely miss one.
    As you already know, unexpected treasures are available there every month.
    Some updates on running the W6TRW swap meet. The city changed the access due to 9/11 tragedy.
    these days, vendors are allowed in at 6:00 am to 7:30 am, no buyers may enter until 7:00 am.
    No in or out vehicle traffic after 7:30, and the driveway is now blocked with a truck across the entry.
    See ya..

    1. “The city changed the access due to 9/11 tragedy.
      these days, vendors are allowed in at 6:00 am to 7:30 am, no buyers may enter until 7:00 am.
      No in or out vehicle traffic after 7:30, and the driveway is now blocked with a truck across the entry”


    2. Unexpected treasure today at the swap meet..
      A cute Dell small form factor desktop, sporting an Intel i7 running at 3.30 gHz, with 8 cores.(Yes, Eight Cores)
      Ram is 8 gig, hard drive is 500 gig, and a Catalyst digital video card.
      I was forced to pay $125.00 usd.
      I have high hopes for a SSD transplant.

      1. You think you bought a Dell Intel Desktop with 8 cores. Lol.

        Sorry, no. It’s a 4 core, 8 thread CPU. Intel JUST THIS YEAR started making 8 core mainstream CPU’s.

  2. Of course, before modems there was the current loop. Teletypes were sending and receiving data over the phone line starting in the 1930’s. They required a direct connection to the phone line, so could only be used with the phone company’s blessing.

      1. Before even… depending on whether you’re counting the first commercial use (1865) or the inception and testing (nearly 20 years before the commercial product) I always feel a sense of wonder when I try to wrap my head around that when paired with my perception of technology during the civil war. Everyone I tell has the same reaction I did at first, which of course is to scoff at such a preposterous notion and then attempt to prove them wrong…. followed by incredulously attempting to make sure I wasn’t being fooled by some sort of internet-wide hoax. heck of a thing.

    1. I believe current loop was run over “dry pairs” which are lines, owned by the telephone company, but not “phone lines” in the usual sense which means POTS unless otherwise specified.

      1. The dry circuits I worked on were usually “bananas” or something like a 69.BANA.123456 circuit ID. later on the circuit ID’s were 69.LADA.123456 etc. Fun to find up to a few years ago. Some were even still working in the Central Office I was in. Everything from SCADA to trellis modems and current loops could be found on them.

    2. Teletype over telco lines was either by acoustic coupler (Omnitec 700 and 701 had both 20mA current loop and RS232 connectors) for dialup POTS lines, or TWX or Telex service which was current loop. TWX was area code 810 and 710, Telex I’m not sure about. TWX was Bell System, but isolated from the POTS system, calls made from a standard telephone line to area code 810 or 710 got an intercept and wouldn’t go through, but Teletypes on TWX could dial other TWX numbers just like you’d dial a phone call.

      I maintained ASR33s for my University Computing Center. We had 100 or so dialup incoming lines, and the Teletypes scattered around campus had Omitec 700 and 701 couplers. All done over the school’s Centrex exchange.
      So you powered up the TTY, picked up the phone handset, dialed “5-1500” (Centrex let you skip the first two digits if you were on the exchange) and waited for the answer tome, then jammed the handset into the acoustic cups and hit RETURN until you got a response from the mainframe.

      Thanks for the memories :-)

  3. Would you mind not bashing everyone younger than you in your writing? Piss poor attitudes from the over 30 crowd are why I personally left the manufacturing and engineering sector, and you aren’t helping

      1. I don’t think some90skid is complaining about the ‘young kid bashing’ in this specific article as much as the attitude that Benchoff consistently demonstrates toward the younger generation. On a website where the statements he makes are most likely to be false of all places.

    1. Holy crap, kid, did they really drive you out of the industry?

      I have been in technical fields since about 2000, I’m just outside of the “millennial” generation, depending on how it is calculated. However, having been around a while and having struggled to climb the ladder, I can tell you that managers still love hiring young, enthusiastic engineers and technicians. The new whippersnappers will always be given a hard time by the old guard, but the old timers will always genuinely appreciate the young blood that is serious about the technology and the art.

      Yes, it’s true that the young apprentice must prove themselves, a degree is not enough (sometimes not even necessary). But this is where the “generation gap” often rears its ugly head. There are many young graduates who don’t understand that it is completely normal to have to prove yourself in a field before you get the respect and the big pay day.

      When I hire, I want the most brilliant folks I can afford on my team. I want them to be smarter than I am and have more potential for greatness. And so I end up with a broad distribution on the age scale. The young pups that come on board get put through their paces to prove their mettle, and that will never change. But the only time it becomes toxic or negative is when someone feels they shouldn’t have to put in that time and effort to gain the respect of their peers. You can’t just skip over the hard part, the entry level. That’s just life!

      If you are being abused and discriminated against for your age (in this case, youth), that is truly an awful thing and I don’t want to make light of it. But you need to know that the greatest bias among hiring managers tends to be that they don’t want to hire folks who whine and complain, regardless of age. Make sure that isn’t you, and you will find someone who will take a chance on you. It may take a long time! There are billions of folks competing for resources these days. But it is not hopeless, and any trouble you may have finding a job in your field is certainly not because no one wants young, energetic, and typically lower cost employees.

      1. “But the only time it becomes toxic or negative is when someone feels they shouldn’t have to put in that time and effort to gain the respect of their peers.”

        I’ve definitely dealt with ‘old guard’ being convinced that anyone younger than them was useless by design and extremely toxic. Once with my suggestion (and PoC) for how to solve a bit of a complex problem being thrown completely out as unusable, while they introduced more and more sketchy ways to attempt to fix it and ridiculed my every suggestion as impossible for over a year until it finally circled back around to exactly what I had done without their help in the very beginning, where they promptly ‘forgot’ that was how I originally designed it. When I bought this up the response was ‘well I guess you were right’ with no further discussion, followed by still being excluded. That was an expensive waste of a year of work and thousands of man hours. It’s definitely one of the factors that made me leave engineering as a career (or at least working on teams). Hopefully I can do some stuff solo because I really like it. The younger kids I deal in engineering with are universally excited to do it and universally smarter than I am (and extremely fast learners!) it’s always a pleasure to work with them, even if I don’t connect on a cultural level.

        And in my case there was definitely no ‘thinking I didn’t need to put in my time’ because I was new to the field with no formal training and know I’m still not very good at it. But just because I don’t have the technical skills to execute on the first try doesn’t mean I’m incapable of thinking about a problem and narrowing down solutions, as the older engineers believed.

          1. The trick is in explaining the idea.

            If you’re quick to draw a really nice picture about it, people lock on to it regardless. If you can illustrate it and present it well, it takes over their imagination and suddenly nobody can think otherwise. If you’re trying to explain it verbally or in text, people just ignore it because they won’t bother assimilating the idea.

            Avoid any technical details, numbers and charts – anything that needs interpretation. An animation is the best.

        1. Jacob, I can totally commiserate with your story and I know that is not uncommon at all. My previous employers have always had pockets that were resistant to change, reality, and reason. Sometimes this infects a wide swath of corporate culture, and sometimes it even starts with the C level.

          That widespread resistance to innovation is the recipe for a failing organization. I’m glad you got out with your dignity and moved on. It may take decades for that place to fall apart, depending on how well funded and entrenched in the market they are, but trust that they will fall. If they are that rotten from within, the cracks will show and the whole thing will begin to crumble at some point.

          My point is, if you find yourself in a mess like that as a young techie, take the time to assess yourself and make sure you are taking the valid criticism for what it is worth. Then, as you keep improving yourself to be your best self, start quietly looking for a place that isn’t rotten and smelling of death.

        2. What I’ve run into many times is coming fresh into a situation and quickly spotting a solution by combining knowledge I have from something else with what I’m looking at. Some specialists tend to dislike it very much when a generalist takes a ‘peek over their shoulder’ and solves in a minute something they’ve been struggling with for a long time.

          Well, too bad. Could’ve stared at that for another week and never found the problem because yer too stubborn to ask other people to take a look.

          Even worse is when a person who doesn’t know anything at all about the gizmo looks at it for five seconds, finds a loose wire or fastener and fixes it after the people who know every last detail had missed it for hours. Some times you just have to pick a thing up and flip it upside down and shake it… *clack* “Is this thing that fell out important?”

          1. I once saw someone who wasn’t even being paid (some bosses do that in the DCPHP market) look at a DB problem two “high level” people” were staring at on his first day and say “You need an XOR there.” Saved them hours based on how long they’d been staring at it. He wasn’t even a quirky genius, just a fresh set of eyes. I know because he is me.

          2. I’ve seen it numerous times in the Navy when the person assisting didn’t even really have to say anything about the problem, and I’ve been on both sides.

            Tech: I can’t figure out what the problem is with this equipment.
            Assist: what’s the problem?
            Tech: Well, it won’t do this, this or this. It will do that, that and that. I checked here and got this reading. I checked there and got that reading… oh, never mind, I figured out what the problem is. Thanks.
            Assist: No problem, glad to help.

            I was sent over to a ship once and told to replace a converter on one of four antennas on a satellite receiver. The tech over there said he didn’t think that was the problem. I said I was going to replace the converter because the old one had a crack in the case (and since that is what I had been told to do). I said if that didn’t fix it, then I would see if I could figure it out. The converter didn’t fix the problem. What was the problem? The receiver had four ‘idiot lights’ on it, one for each antenna, and one was lit. I told the tech that the first thing to do was to determine if the problem was with the receiver or the antenna. If we swap the
            antenna that has the lit ‘idiot light’ with one of the other three antennas, the problem will either follow the problem antenna or stay with problem port. I swapped the two antennas and none of the lights lit up. OK, let’s swap them back and see if that antenna is somehow incompatible with that port. I swapped them back and none of the lights came on. We chalked it up to obviously just a dirty connection. What should have been a no-brainer check, but nobody checked it.

    2. Please point out the age-bashing to which you’re referring. All I could see is one mention of Fortnite in a neutral comparison. Merely comparing a vintage tech term to a modern game seems benign.

    3. Sometimes a writer’s style might appear demeaning, when he’s just intending dry humor:

      I happen to be mentoring a recent grad new hire right now in the intricacies of Verilog. He has to put up with my reminiscences of old technology (anyone else remember ABEL and DataIO programmers with their stupid 4 digit codes?) and my constant reminders that he doesn’t know how good he has it.

      He’s a bright kid. He’ll make a good engineer once I increase his level of skepticism by a couple of orders of magnitude. :-)

  4. I have a really hard time believing no one else knew what this was. The combination of the full/half duplex switch and the DB25 connector simply scream ‘modem’ to anyone with a comms background, surely?

    1. Well, without the acoustic coupler (and attached label) I’d guess it was modem by the Full/Half Duplex switch.
      But, I wouldn’t be certain it was one.
      And, please, don’t call me Shirley!

  5. The one thing I would add is the definition of modem: you say “modulator” and “demodulator” referirng to the sub-boards without the clue that modem is a portmanteau of the two.

      1. Typical Socialist Professor, eh?
        Modem = Modulate, De-modulate = modem
        I’ll wager he also wanted free state funded housing, cars, and much more.
        Socialism only works until you run out of other peoples money.

  6. so imagine if corporatism didn’t slow down the adoption of more modern communication techniques by decades just to maintain one companies’ legal monopoly.
    imagine if Exxon didn’t suppress climate change data for decades for the same purposes.
    sorry, of topic, but if we don’t learn from history, blah, blah, blah….

    1. I’m not disagreeing, but consider that when those more modern communication techniques did arrive, they were built on and spread on the proceeds of that companies legal monopoly.

      Don’t forget what funded R&D that brought us the silicon transistor, information theory, and more. Those more modern communication techniques built on the huge investments in the telephone network.

      Would those things have come to pass if The Phone Company Monopoly had never come to pass? Probably, but when?

      1. Saturday Night Live Transcripts

        Season 2: Episode 1

        76a: Lily Tomlin / James Taylor

        The Phone Company

        Ernestine…..Lily Tomlin
        Technician in background…..Al Franken

        Ernestine: A gracious hello. Here at the Phone Company, we handle eighty-four billion calls a year. Serving everyone from presidents and kings to the scum of the earth. So, we realize that, every so often, you can’t get an operator, or for no apparent reason your phone goes out of order, or perhaps you get charged for a call you didn’t make. We don’t care!

        Watch this… [ she hits buttons maniacally ] We just lost Peoria.

        You see, this phone system consists of a multibillion-dollar matrix of space age technology that is so sophisticated — [ she hits buttons with her elbows ] even we can’t handle it. But that’s your problem, isn’t it? So, the next time you complain about your phone service, why don’t you try using two Dixie cups with a string? We don’t care. We don’t have to. We’re the Phone Company.

      2. well, i think my point is (FWIW) that it could have been faster.
        i’m trying to figure out how to bring something to market, i’m not going up against amazon or FB.
        back in Ma Bell days, you would be crushed and no one would even hear the sound of it. so no one even tries. except maybe some other oligarch, who’s not really risking anything.
        if we had universal health care, more people could take that chance to launch something risky.
        so we don’t have universal health care. (not to mention everyone would be registered to vote, so, no again)
        these are direct relationships to the development of technology, in my opinion.

      3. The AT&T monopoly is why we have Sprint. Sprint used to be SPRINT or Southern Pacific Railroad Internal Network Telephony.
        SP took advantage of the new fiber optic technology to install communications and signal control along all their railroad tracks. Since it was optical it was immune to EMI and RFI.
        But there was a ‘problem’ with the fiber. It had far more bandwidth than SP would ever use, so they got the idea they could make $$$ selling that bandwidth for telephone calls originating and terminating outside the SPRINT system.
        When approached about connecting SPRINT to the PSTN (Public Switched Telephone Network) AT&T said NO.
        Southern Pacific sued AT&T and won.
        Eventually the railroad spun off SPRINT as the separate company, Sprint.

        So if you’ve ever felt railroaded by Sprint, now you know why. ;)

        1. Microwave Communication International was a big player too, they later got absorbed by … eh you know the story. Ma Bell’s divestiture was because of Sprint and MCI. Good? Bad? We’re here now I suppose.

        2. And WilTel, founded in 1985, (which became VYVX, which became Century Link) and its parent company, The Williams Cos. Inc., ran fiber-optic cables through decommissioned natural gas pipelines. Beats digging up roads.

  7. “The probability of knowing what an acoustically coupled modem looks like is inversely proportional to knowing what Fortnite is…”

    I first read that and stopped, thinking… I used to know how many days in a fortnight. Then I thought… why would that be inversely proportional? Then I looked back and realized the article was talk Fortnite the game, not a fortnight. Oh well.

    1. I suspect it’s true both ways. ;)

      I don’t usually notice the byline in an article, until I run into a Benchoff sentence like that and scroll back up to see if I was right!

  8. My first modem was an acoustically coupled 300 baud modem on a TRS80 model 100. I still have it and the TRS100. I should check if they still work. After that I moved on to selling 1200 baud plug in modems for pre-Internet services like AOL. Then I got lucky and the company I worked for supplied me with a blazing fast 56kb frame relay service! Then to DSL and so on. Although we have come a long way in terms of communications speed, the services in the USA seem to be hitting a wall at around 50mbps or less (actual speed not maximum speed) in most places. Gigabit Internet speeds are still rare. Maybe 5G will change that. The volume of data just for a simple web page is huge nowadays and it seems like load times never get better. More bandwidth just seems to result in more more garbage being added to sites.

    1. I swear one of the best browsing experiences I ever had, on text only with images turned off, was on this unicorn of a 14.4k modem I got, it was a cheapy unit built just before the lempel-ziv patent war shit hit the fan and had multiple compression technologies implemented, and was only matched by the ISP multi-protocol modems at the POPs… Anyhoo, text webpage throughput was upwards of 10 kilobytes a second, seeing 15 sometimes and it just flew around the web, until I needed images or downloads, then of course it was chugging at 1.5-2KBs it made both the 28.8 and the 33.3 modems I got after it rather disappointing since their compression only managed 2x on text, though binaries got 3-4KB… finally a later 56k after all the standards were mature started to match it, but in 2 or 3 years the bloat had caught up, and you couldn’t see half of what was what unless you turned on images.

  9. I took one of these apart as a kid in the 1970s and used the parts to build a rudimentary intercom. It was a very nice wooden box. My dad brought it home from work and it must have been completely obsolete if he gave it to me to play with.

  10. “It’s an objet d’art, and it’s useful to store various sundries.”

    Shopkeeper: Hey , may I help you?
    Customer #1: Yeah. This is kind of neat, what is it?
    Shopkeeper: Oh, that’s from New Guinea. It’s a ceremonial spirit box.
    Customer #1: Wow, that’s cool. What do you do with it?
    Shopkeeper: You put your weed in there!

  11. Strange that the made they made the PCBs symmetrical. I can’t see anything that would prevent putting in the PCBs upside-down and the single-sideded PCB would still make contact with the other side of the connector, but in reverse pin order….

    1. These weren’t meant to be user-serviceable, but in any case, what prevented putting the PCBs in upside-down was training and inspection.
      Believe it or not, they used to make symmetrical integrated circuits, too (dual-inline packages), and those really WERE a hazard, especially when socketed, since the user COULD do something stupid.

      The obvious question would be, why did the sockets have contacts on both sides?

  12. My first experience with a real computer was through a Model 33 Teletype plugged into one of these. My high school had four of these pairs tucked away in a corner of a classroom. Only, ours were in the dovetailed boxes. Which seemed anachronistic, even in 1970. I mean, computers then weren’t even plastic, yet. They were METAL. I knew instantly from the picture, what this must be, although I never saw one with the box opened up. This was in 1970, and the computer we called (using a rotary-dial phone) was about five miles away at the University of California at Irvine, where a monstrosity of telephone relays and stepping switches called the “line finder” connected phone lines through similar modems to available serial ports on a Xerox Data Systems Sigma 7 mainframe.

  13. God I feel old. My wife used a 300baud accoustic modem for a while years ago from home for program development. I think it had an option to switch to 110 baud as well. But that was in the 1970’s. It was included with a typewriter in a portable case. Well semi-portable with a ghastly thermal printer for output.

  14. Datamation Sept 26, 1977 issue page 35, there is a brief blurb about a friend of mine, Roger Modeen who had just presented a schematic for a home-brew modem based on a Motorolla 6860 modem on a chip. He ended his paper with the line “No attempt is made to defraud, defang, defrock, or discredit the Bell Telephone System . . .” In my memory, the word deflower is in there somewhere”.

  15. Back in 1976 I was trained on 24 channel telegraphy system working with mechanically vibrating tongues to decode the dual tone signaling used for a single 50bps channel. Imagine having 24 of them riding a single 2.1 GHz beam working with Klystrons emitting 15W of microwave power. This was also the time when we repaired 6bit A/D D/A discretely built used in digitally scrambling voice. At the time I was trained on that system is was already pretty old, it was installed initially just post war in 1948. Some things seemingly never change :) our laws of physics still standing firm.

Leave a Reply

Please be kind and respectful to help make the comments section excellent. (Comment Policy)

This site uses Akismet to reduce spam. Learn how your comment data is processed.