PCMCIA History: From Laptops to CableCards | Hacker News Reader

PCMCIA History: From Laptops to CableCards

tedium.co

51 pointszdw19 comments

This took me a while to read because I ended up going down the rabbit hole of reading about the PoqetPC - a tiny subnotebook that could last for weeks or months on two AAs: http://www.bmason.com/PoqetPC/

I had a Fujitsu Lifebook P-2046 in college (2002-2007). It was frustratingly slow, but could manage 15 hours of battery life. Even five years after purchasing it, people would remark on how compact it is.

I had not one, but two given to me by a family friend in the mid ninties! I was young and it was epic. I could send faxes via a 14.4k serial modem with a PoqetPC to parallel, parallel to serial adaptor. I could play stunts on it, and write BASIC programs, what more would you want at age 7? Yes the battery lasted weeks, and the memory cards held a small watch battery in them. I've still got one somewhere, which I now want to dig out.

I would pay lots for a modern thing with that form factor.

The Pandora [1] is a few years old at this point. I had one for a short while and wasn't too impressed. There will be an update called the Pyra [2] coming later this year.

[1] https://www.openpandora.org/index.html

[2] https://pyra-handheld.com/boards/pages/pyra/

Yeah they look beautiful. I went down that rabbit hole too.

I'm not sure why the article laments CableCard. It allows me to watch most of my cable channels using my HD HomeRun ethernet CableCard tuner using just VLC. The premium (think HBO) channels are protected by DTCP DRM but lets just say that isn't the most secure DRM ever. The proposed VidiPath replacement is just going to lock down the TV experience even more. It will allow cable companies to have a DRMed path to your display where they control every pixel, including the guide and DVR interface. It is directly aimed at taking out the likes of Tivo and Silicon Dust.

The primary lament of CableCard is that it's so poorly implemented that the vast majority of people aren't even aware that it exists. Moreover, things like "On Demand" content are baked into the cost of my subscription and I can't take advantage of them through a CableCard.

Cable Television, as we perceive it today, is in a steady decline (I'd cite something; but even the most cursory Google search will turn up dozens of articles confirming). While I recognize your concern about the cable company ownership and DRM, I believe that over-the-top services like Sling or DirecTV Now in conjunction with Netflix and Amazon are far bigger threat to Tivo and the like than anything the cable companies themselves can do.

> The primary lament of CableCard is that it's so poorly implemented that the vast majority of people aren't even aware that it exists.

You're inferring that they're related, but I see them as orthogonal.

I'll take your word that it's poorly implemented. From my POV it just works—my two experiences are with a TiVo and an HDHomeRun (first with Media Center, now with Plex's beta DVR feature), and in both cases use has been uneventful.

That people don't know it exists doesn't surprise me, since the cable industry has always fought this FCC-required concession to consumer choice. If anything, it strikes me that cable was highly motivated to ensure CableCard's failure.

I don't have any love for CableCard, but I'll be sad when it goes because it'll mean that kind of interop is gone forever.

But DirecTV Now is, while not a traditional cable service, something a cable company is doing.

And that makes it the biggest threat to Netflix when net neutrality (and particularly the specific prohibition on ISPs favoring their own video services) is killed by the new leadership at the FCC.

I agree that things like DirecTV now is the future. As long as those services are owned by companies that have traditional cable/satellite/fiber services, they are going to try to upsell or limit these services.

The problem with CableCard is that it was difficult to get working. The cable companies were against and it did not put too much effort into make it work.

I recall someone saying that technicians would bring a basket of CableCards and hope one of them would work.

I've done a couple of implementations, it's complicated but not that hard .... I think there were a couple of issues - the cable companies liked the revenue stream from renting you boxes.

Mostly though cable crypto is dominated by two companies who provide the head end equipment and who for a long time provided all the settop boxes - they didn't want to lose that lockin so while they were mandated to provide cable cards they weren't all that happy about it

Based on this I doubt that will happen: https://www.dslreports.com/shownews/Senators-Urge-New-FCC-Bo...

It should be noted that the modern "expansion bus" is not PCI or other such busses... but USB3.1.

Today, the best audio drivers are external USB DACs and amplifiers. You can extend your storage with GB of USB flash storage. USB->Serial has kept up as a legacy option (although its latency issues and Win7 transition has made things difficult to work with in practice... gotta keep that old WinXP box to work with a lot of Serial stuff. But that's not USB's fault per se and more about Windows Vista's new security model).

Audio (DAC, Headphones, Synthesizers, USB->MIDI, Instruments), Storage (USB Flash Drives, USB Hard Drives), video game peripherals (XBox Controllers, PS4 Controllers, Joysticks, Driving Wheels), even legacy (USB->Serial Modems if you really need it) are all supported from the 5Gbps USB3.0 port.

--------

PCMCIA was needed in the 90s because computing power wasn't good enough. DMA transfers, directly to the computer's RAM, was the only way to get latency / bandwidth requirements of practical devices.

Furthermore, hardware interfaces were obscure. Your "Jazz drive" external storage was SCSI based and required a specific device. USB internet modems were years away from being invented (and thus you used PCMCIA modems on those laptops).

In essence... PCMCIA was the way you attached external devices to laptops back then. More so than USB or Firewire. Modern interfaces are more secure, more inter-operable, and faster.

I mean, those flash-sticks require a rather beefy CPU to handle all of the USB-packets that go back and forth between the OS and the USB stick. That level of miniature processor power just didn't exist in the 90s, so "dumb" hardware attachments had to use simple direct-to-memory transfers through a relatively simple bus. I wouldn't be surprised if a modern USB-stick's processor had more computing power than 90s-era laptops.

> I wouldn't be surprised if a modern USB-stick's processor had more computing power than 90s-era laptops.

Data transfers are handled in hardware; the MCU in these is not really powerful. Many are just 8051s. USB 3.0 controllers are a bit beefier, though.

Exactly. And for even faster expansion, DMA, etc. there's Thunderbolt. You can easily (relatively) hook up external graphics card now.

You forgot the most prominent feature of PCMCIA/PC Card: You can stuff them inside of the laptop instead of carrying bunch of USB dongles and cables.

For storage expansion this was easy way to get more space but of course network cards required additional dongle for the RJ45 connector etc. But still the idea was awesome :)

I actually had one of the two Amiga models (A600) that traded the generic properitary expansion slot for a PCMCIA one. However, living in the country I live in I never actually encountered a PCMCIA _card_. Catching on the topic years later it turned out I did not miss on much, PCMCIA being not very well defined at the time of A600's design phase, the slot wasn't really useful for anything more than a slow flash based disk drive or very sluggish memory expansion (there were some really neat gizmos you could put in there, but those came about YEARS later)

Reply on news.ycombinator.com