Talk:Apple Desktop Bus

Powerbooks
"No machines being built today currently use ADB." Actually, at least the Titanium Powerbooks (2002-3) still used ADB for the internal trackpad, and I'm pretty sure the Aluminum Powerbooks and iBooks still do as well. Anyone have one of these that can do a verbose boot to check? -- Kaszeta 15:02, 20 Oct 2004 (UTC)
 * ADB trackpads were used until 2005. The entire line switched to USB with the introduction of the new two-finger scrolling trackpads. Binary 20:26, 14 April 2006 (UTC)

Keyboards' default address
"For instance, all keyboards were set to $2, and all mice to $3."

"Given that all keyboards were on $3..."

Clarification, please? --Ntg 02:56, 16 Jan 2005 (UTC)-ntg--4.35.128.173 02:44, 16 Jan 2005 (UTC)

The old Apple documentation used the dollar sign ($) to indicate hexadecimal (base 16) numbers (see the entry on Hexadecimal). I believe the current programming convention is to use "0x" as the prefix for hex numbers. The phrase in question is indicating the default addresses ADB assigned to different devices. Hope this helps. -- K Cundiff —Preceding unsigned comment added by 65.197.74.25 (talk) 21:34, 28 February 2006

The above comment is more than four years old by now, and the article text has changed; but just to avoid potential confusion, a keyboard's default address would be $2 and nothing else. This document from Apple states that the default device ID for "encoded devices" (i.e. keyboards) is $2. ID $3 is "relative-position devices" (mice, trackballs). -- magetoo 10:48, 15 June 2009 (UTC)

Speculation
To quote the article "As a result, the system was not hot swappable, although this could have been implemented very easily using the PSW had they thought about this case" - what exactly is the evidence that this possibility was not considered? ThomasHarte 15:35, 19 November 2005 (UTC)

Adding to this speculation the Apple Developer Note for the Macintosh PowerBook G3 Series computer (commonly called the Wallstreet) states:
 * "Unlike earlier ADB-equipped computers, the Wall Street computers allow the user to unplug and replace ADB devices while the computer is operating" (p 49)

suggesting that in the latter implimentations of ADB this issue was resolved. Does anyone have more information on this? --scott 05:24, 17 June 2006 (UTC)


 * It's not clear from this quote if the change was to allow this to be done safely, or whether the machine actually noticed these changes and did a reset. I was hoping there would be more information in the dox, but you pretty much quoted the entire section! Maury 23:52, 22 December 2006 (UTC)

USB is also still not all that safe to put in when a system is active. —The preceding unsigned comment was added by 82.171.59.199 (talk • contribs) 16:10, 22 December 2006 (UTC-7)

no plug is safe to plug-in while a system is running technically speaking, as it will cause static. (there is always a potential of frying a controller chip) Markthemac (talk) 22:25, 4 March 2010 (UTC)

Similarity to S-Video connections

 * Should the similar appearance of s-video cables be mentioned in this article? --DangApricot 00:55, 27 November 2006 (UTC)
 * Well the ADB is an s-video connector, so... yeah. Maury 23:49, 22 December 2006 (UTC)
 * I actually needed an S-Video Cable one time when i didn't have one, and it worked very near perfectly (image was a tad blurry)173.48.62.136 (talk) 11:24, 20 August 2009 (UTC)

that's right they are S-Video compatible, it always was a cheap and easy plug to produce on a mass scale. Markthemac (talk) 22:30, 4 March 2010 (UTC)

Clarification of ADB status
From the article:
 * but up to February 2005, PowerBooks and iBooks still used the simple ADB protocol in the internal interface with the built-in keyboard and touchpad. The internal connection for the trackpads has now been changed to USB.

This seems to imply that internal keyboards still use the ADB protocol. Is that actually the case, or did both keyboard and trackpads switch over to USB? —lensovet–talk – 21:06, 8 June 2008 (UTC)
 * The keyboard became USB in October 2005. . Really, calling any internal PowerBook devices ADB is a little misleading. The keyboard and trackpad would connect to a small microcontroller, which in turn connected to the southbridge (in earlier models, that controller essentially was the southbridge). That microcontroller emulated an ADB interface using a few of its interface registers, allowing the existing mouse driver to be reused. So there was no ADB signaling or ADB bus, just ADB emulation. Potatoswatter (talk) 22:11, 8 June 2008 (UTC)

that way apple saved cost to write a new driver for every single keyboard type, USB was a pretty bad standard for keyboards when Apple first started using it (USB was Aka Unused Serial Bus, apple hacked around an unused standard to get a unified keyboard driver functional which also worked on older non-usb hardware) Markthemac (talk) 22:35, 4 March 2010 (UTC)

Confusing paragraph
Quoting a paragraph that I find confusing:
 * Most serial digital interfaces use a separate clock pin to signal the arrival of individual bits of data. As ADB was designed to be low-cost, Wozniak recognized that a single wire had enough bandwidth to carry both signals at the required data rate. Moreover it was economical to decode the clock and data and use cheaper cables.

What does this even mean? "a single wire had enough bandwidth"? What does bandwidth mean in the context of a clock signal? What bandwidth does a wire have? (What is the airspeed of an unladen swallow?) What the heck does "[decoding] the clock and data" mean here?

I suggest it should read something like "Wozniak decided that a separate wire for a clock signal was not necessary, and it made economical sense to leave it out". And if someone knows the actual details it could be added to the "communication" section. -- magetoo 11:05, 15 June 2009 (UTC)


 * The more I read the paragraph, the more it looked like gibberish, so I rewrote it. Does it need more details or is it fine as it is now?  You could argue that one would want discuss why other busses have a clock signal when ADB could do without, or when it is appropriate to leave it out, but that can quickly get pretty technical and hard to read... -- magetoo 11:20, 15 June 2009 (UTC)

Bandwidth (signal processing) is the range of frequencies found in a signal. Bandwidth of a wire (or other signal carrier) may be considered the range of frequencies that pass though it. The process of decoding signals from a carrier involves separating the frequencies. You could read the Wiki article on the confusing term… but I think your rewrite may be a better baseline for adding details. Potatoswatter (talk) 16:37, 16 June 2009 (UTC)


 * I get what you're saying (no, really!) but it just seems like a really weird way to think about it, like trying to describe Morse code in terms of separating out a "clock" signal and a "data" signal when dits and dahs is such an obvious way of thinking about it... Well, too much talk page space already spent on this tiny detail, it'll get lost in a rewrite sooner or later anyway. -- magetoo 20:54, 16 June 2009 (UTC)

I removed this paragraph and just referred to the article on self-clocking signals. Intellec7 (talk) 16:42, 29 June 2020 (UTC)

The inventor of ADB
The article claims ("as the story goes") that Steve Wozniak "went away for a month, and came back with ADB". Even without the stuff I'm going to talk about below, this story seems a bit too pat.

The patent section of the article lists a number of patents associated with ADB, none of which have Steve's name on them.

What brought this to my attention was the fact that I was at a product demonstration at an apple office in the 90s in which the presenter pointed out a member of the audience and said "we are honored to have the inventor of ADB here today." and it wasn't Steve Wozniak. I went and spoke to the gentleman (I believe Michael Clark) afterwards, who confirmed that he had invented the protocol and mentioned that apple only had the license to use it for personal computers. When I asked what else you would use it for, he said that he had licensed it to the navy for use on submarines where communications links needed to be made without breaching a hull or internal bulkhead.

I'm not going to update the article based on my personal anecdote from 20 years ago, but I've been looking for more confirmation and found a couple links that I may use to edit the article later.

A mailing list archive from 1999 has Michael's CV making the claim that he invented ADB. http://www.eyetap.org/wearables/wear-hard/19996632.html

The apple fritter wiki has an article where someone quotes an email allegedly coming from Michael where he also makes the claim to have invented ADB along with some more specific details. http://www.applefritter.com/?q=node/294

I can find nothing about the submarine thing, maybe he wasn't supposed to tell that story. Mkanoap (talk) 14:45, 17 March 2017 (UTC)

Noting that ADB is proprietary
I clarified that ADB was controlled by Apple. This is important context to understand, both about the bus itself, and the resulting marketplace for peripherals. It also helps explain why a good standard was ignored by everyone but NeXT. Please do not softball this aspect. (nop (talk) 17:27, 27 April 2017 (UTC)).

ADB vs mini-din serial
This article seems to heavily imply that ADB and the Mac's serial ports are one in the same. I'm not expert, but I think clarification (especially from someone much more knowledgeable than I) needs to be made. Either ADB and the mini-din serial /are/ different physical implementations of ADB, or they're completely separate. I know the one example posted as an exception to input devices, the Global Village modem, at least the two I had, used the Mac serial, NOT ADB.

There are adapters to use DB-9 serial devices on Mac serial ports - I've never once seen an adapter to use a PC serial mouse on an ADB bus, however. I grew up with these Macs, ADB, and mini-din serial ports.

I believe there is a serious fundamental misunderstanding of the two technologies (either by the authoring community, or myself, though I'm pretty sure I'm right here) and that the entire wiki article needs to be completely re-authored as many of the suppositions in this article are based on ADB and Mac serial being one in the same. 24.173.150.130 (talk) 13:37, 3 August 2023 (UTC)