Print

[blusnowkitty]

Gonna spin this off from Ray's sales speculation thread...

Took a couple of hours here and scoured the internet for Lisa serial numbers and AppleNet numbers. Found a good few results thanks to old forum posts and eBay listings. Have a look-see, and if you want to contribute your S/N and AppleNet, post 'em here.

https://docs.google.com/spreadsheets/d/1iQ5O4xBpRkras6bkaH2r2nRMOAeK_AHTc1fEspSzGoo/edit?usp=sharing

[blusnowkitty]

https://68kmla.org/forums/topic/26659-lisa-2-serial-number-how-to-decode/
https://forums.macrumors.com/threads/apple-lisa-serial-and-date-codes.147497/

So here's some observations... You all probably knew most of these already but for future reference. This list doesn't account for Macintosh XL units.

• There's two distinct formats for the serial number, one being prefixed with B08B while the other is simply prefixed as A. The first five numbers of the B serial seem to correlate to year/day of manufacture (i.e., 82100 would be 100th day of 1982) and the remaining 3 is the Nth machine made that day. The A prefix has a 4-digit date of manufacture in an unknown format, the remaining numbers are the Nth machine made that day.
• The 68kmla link above states that any machine prefixed with a B in the serial means it was a Lisa 1 that got upgraded to a 2/5. Where did this info come from? Seems like there's a number of Lisa 1 units, still un-upgraded that are prefixed with a B.
• As far as I can tell, there seem to be no difference in AppleNet continuity between B and A units.
• Was Apple making 2/5 and 2/10 units side-by-side up until the end? Or at least, using up surplus parts until the end... See how there's 2/5s sitting right alongside 2/10s? Weird.
• Highest number I've found so far is Applenet #10858 which correlates with wherever Wikipedia got their 10,000 units number. But does it account for the 2,700 dumped in Utah, or the units converted to Mac XLs?
• This one might be one for Al if he's at liberty to talk about it - B08B820360010, or Applenet #65 is the second known oldest Lisa in existence currently, not counting things like the corroded proto from Facebook, or Digibarn's naked Lisa prototype or the GLM. If you search for the serial number, you'll find it listed both on the CHM's website and on Christie's - so did someone at the CHM have really deep pockets, or did the CHM offload a machine out of its collection?
• What's with this weird serial in the attachments?

Judging by the current data, it seems that there were at most 12,000 - 15,000 Lisas sold as Lisas.

[rayarachelian]

Yup.

These are mine from the chassis:

SN: A3334150, AppleNet: 00108643, MFG: 3334 - recent buy, last year off ebay, this had Jazz on the Widget drive and is a 2/10.

SN: B08B831460139, AppleNet: 00104391, MFG: 83146
Missing sticker, I think this was a lease, has two stickers, top one has a red border and black numbers, no bar code, in a white box says: "Property of Apple Computer Inc." A1081 and then a 2nd one, black box with a barcode: 106336 - not sure what these mean, but the 2nd one looks like an AppleNet and possibly the top one is the S/N.
SN: A3245610, AppleNet: 00104218, MFG: 3245 - this was also a recent buy off ebay, severely damaged due to battery leak, had a 30MB SunRem hard drive.
Interestingly the apple net IDs of two of these both start with 104, but one is the new B SN and the other an older A SN - this might be a clue.

And here's a couple off recent ebay listings:

SN: A33344145 AppleNet: 00108638 MFG: 3334

Here's one from a chassis on ebay that's being chopped up and sold as panels:
SN: A3197327 AppleNet: 00103676 Mfg: 3197

Another on ebay today:

SN: A3181120 AppleNet: 00102120 Mfg: 3181 Interestingly in this one the AppleNet last 3 digits match the SN's 3 digits. Possibly coincidence, but perhaps it means something (obv. what we're looking for is a unique count, if what you said was correct about AppleNet IDs, perhaps the first 3 digits have to do with a plant number.(as an aside, I wouldn't buy this Lisa, it has severe battery leak damage - also that's an ADB mouse from something like a PPC Mac. Current price on it is $300 but it will cost a lot of time and effort to repair.)

One of the ones on ebay match the MFG date with one of mine:  3334 so this could be useful as an estimate of how many were made on that day.Mine has SN 150, the ebay one has 145 (and this is already in that spreadsheet). So I'd guess there were at least 150 made that day (assuming MFG is by day and not week, and we don't know if that was continuous, day by day, or bursty.)  (Edit: that chassis bottom plate has S/N 327, so ~300/day/factory is the high end so far?)

Another in that spreadsheet has A3341046 - so one day later, but SN 046. So it looks like each SN is by day made, and not sequential. But this one's applenet is  00110381 and mine A3334150, AppleNet: 00108643, so yeah, perhaps the AppleNet ID is the way to go. The delta between the AppleNets is 1738 - that seems a bit high. Were 1738+ machines made between two days across all factories? Or did they skip numbers? If they skipped the actual number of Lisae would be far less than even 10K!

I'm starting to think that the date code is not very useful as a way to get a count as they're all "low" day by day - in the low 100s, but then all the AppleNets seem to start with 001; in that spreadsheet the highest one is around 8000, and one of mine is 0010:8643. So far it's all under 10K.

What would be useful would be to find serial numbers that are sequential from the same day and note what their AppleNet numbers look like, are they also sequential? And then find two that span two days, i.e. MFG: 3333 and MFG: 3334 and see the delta in the AppleNet. That would give us an important data point: are AppleNet IDs truly sequential? And then if so, what's the highest we've ever seen.

It would also be very useful to know when we come across 2/10s - i.e. what's the lowest SN for a 2/10, what's the highest SN for a 2/10, since presumably at some point before deciding to switch to "Lisa is now Mac XL", they switched to making 2/10s. But then again, did all factories immediately switch? or did some pump out 2's and 2/5's for a while from old stock?

The big difference is the I/O board, wiring harness, motherboard and the Widget ofc, but this would mean the chassis for a 2/10 is also different and you couldn't slide in a 2/10 mobo+I/O board in a 2/5 without taking most of the chassis apart to change the wiring harness. So most likely this was only done at a factory.


I also wonder if the keyboard SN's could also yield some clue - if we can figure out what part is unique and whether they're sequential.

At some point, I'll come back to this thread and post all the CPU board SNs from Service ROM, this will take a while due to being busy with other things, so it will be after a long while, most likely once I recap my video boards an power supplies.

@bluesnowkitty - are you able to make use of that German Tank Problem math in some way with this?

[blusnowkitty]

@bluesnowkitty - are you able to make use of that German Tank Problem math in some way with this?

So the math for the German tank problem, at least the Frequentist approach, is actually very, very simple. You take the highest serial number recorded, divide it by the number of units in captivity, then add the highest serial number to that result and subtract 1. The more samples you include in your estimation, the lower the estimate will be but you'll have a more accurate idea of how many units there are that you're trying to number.

What would be useful would be to find serial numbers that are sequential from the same day and note what their AppleNet numbers look like, are they also sequential? And then find two that span two days, i.e. MFG: 3333 and MFG: 3334 and see the delta in the AppleNet. That would give us an important data point: are AppleNet IDs truly sequential? And then if so, what's the highest we've ever seen.

There's a number of serial numbers in the sheet that share the same manufacturing date, and both serial numbers and AppleNet numbers are exactly sequential between both B serials and A serials. So far the only time we've had a manufacturing date difference is Date 4158 and 4159, and there's a difference of 349 AppleNet numbers between the last recorded 4158 and the first (and so far, only) 4159 number.

I'm starting to think that the date code is not very useful as a way to get a count as they're all "low" day by day - in the low 100s, but then all the AppleNets seem to start with 001; in that spreadsheet the highest one is around 8000, and one of mine is 0010:8643. So far it's all under 10K.

There is a machine with AppleNet #10898 floating around out there.

[rayarachelian]

There's a number of serial numbers in the sheet that share the same manufacturing date, and both serial numbers and AppleNet numbers are exactly sequential between both B serials and A serials. So far the only time we've had a manufacturing date difference is Date 4158 and 4159, and there's a difference of 349 AppleNet numbers between the last recorded 4158 and the first (and so far, only) 4159 number.

Right so that's the highest delta we saw so far, so that's useful as a data point in saying that they might have built about 350/day at the high end. We don't know if that was sustained or bursty. We do see a bunch more deltas around 150 between days.

The Lisa was introduced January 19, 1983 and discontinued mid 1985. That gives us 2 years. I'd assume only a few units were available at launch time and likely weren't all ready to ship yet, perhaps the software had bugs and needed patching, etc.

Presumably the factories were shared making Apple ]['s, Lisas, and Macs, but it's possible they were dedicated due to the need to reconfigure the machinery. So if we figure the factory worked M-F year around for ~3 years, at maximum they would have made 52 weeks * 5 days * 350 machines/day: 91000 machines/year, easily over 273K in a maximum of 3 years. Even at half of that, 52*5*150=39,000 machines/year, or 117K over 3 years.

So there's the a potential that 100K Lisas might have been made. Possibly this is where that number came from. But it doesn't take a lot into account. What was the lead time to build all the parts before putting them together, they'd obviously need enough stock for a complete Lisa to assemble one and then test it for a few hours before packing it up into a box. Lots of PCB making, drilling, pick and place, plastic molding, and then assembly. Most of that was done by humans. I would guess a single Lisa would take several days from parts to ready to ship.

Likely most were made in the first few weeks, and they'd keep making them until they needed to build some other machine. At some point, they decided to retrofit from Lisa 1 to Lisa 2, this likely occurred sometime in Aug-Dec 1983, so very early on. https://archive.org/details/bitsavers_applelisahKit1983_473392 - sadly undated. Not sure if the 29-0505 means it happened in May or this is just some coincidental document number. So around there, they would have made enough upgrade kits to match all the Lisa 1s they made, and then continued to change their own stock of I/O boards and faceplates, and then started to make new ones.
https://www.folklore.org/StoryView.py?project=Macintosh&story=Hide_Under_This_Desk.txt says it's August 1983.

The Lisa 2/10 was introduced with the Mac in Jan 1984, so it was designed and built at the mid-end of 1983. Possibly right after the Twiggy->Sony upgrade.
Looking at the ROM source code, it contains useful dates of various events:

• "Rich Castro  7/30/81 - Version 0.0 released to manufacturing" - so we know around this time they likely started building prototypes.
• 4/8/82  - Release version 1.00 - likely this was the "A" ROM, so most likely now is when Lisa production is starting up
• 5/21/82 - Release version 1.02
• 6/14/82 - Release version 1.04
• 7/16/82 - Create version 1.05
• 7/19/82 - Create version 1.06
• 7/20/82 - Release version 1.07
• 8/30/82 - Add auto boot from Applenet <- interesting data point - which means they had AppleNet working well enough to boot from at this point
• 9/10/82 - Create and do internal release of new ROM versions 0.25, 1.16 and 2.01
• 9/14/82 - Create and release ROM versions 0.26, 1.17 and 2.02. - looks like they were working on parallel Lisae at this point
• 10/9/82 - Create version 2.03 - this is version "C" of the ROM
• 10/12/82 - Create and release version 2.04. -
• 11/9/82  - Create version 2.07
• 11/19/82 -  Release versions 2.08 (internal) and 2.09 (for manufacturing)
• 5/9/83  - 1)Change ROM id display to rev # (D) (CHG001) -  this is the "D" ROM - pretty common with Lisa 1s and also 2's - can boot from both twiggy and Sony but shows both drives in the menu with twiggy icons, so likely no Sony support yet.
• 5/12/83 -  Create and release rev D of boot ROM.
• 8/8/83  -  Add changes for Pepsi system:   (CHG009) - so here we go, a bit before this is when they decided to switch to 400K Sony drives.

Code: [Select]

0000|                       ;                                           1) New icons.0000|                       ;                                           2) Display of icons with id #'s.
0000|                       ;                                8/9/83  -  Add save of disk ROM id in low memory. (CHG010)0000|                       ;                                           Add fixes for:
0000|                       ;                                           1) SCC init for Applebus.  (CHG011) - interesting - AppleTalk?
0000|                       ;                                           2) Test card boot search.  (CHG012)

• 10/10/83 - 1)Make Pepsi icon changes.  (CHG024) -
• 10/20/83 - Release as rev E for Lisa and Pepsi systems. - this was never released as far as I know
• 12/21/83 - Release as official rev 'F' for all systems - this is the first release for Sony drives and pretty common. The final one will be "H" (before the 3A)

• 2/7/84   - 1)Extend hard disk default read timeout to 16 - so at this point the 2/10s are being readied.

Code: [Select]

0000|                       ;                                             seconds for Widget systems. (CHG037)0000|                       ;                                           2)Add delay after hard disk reset for Widget
0000|                       ;                                             systems. (CHG038)

• 2/8/84   - Release as rev G for testing
• 2/24/84  - Release as official rev H - this is right after the Jan 24 "1984" announcement, so they would have shipped this later.

So most likely sometime around 2/7/84 I'd expect the factories to have switched to making 2/10s and using the left over stock for the remaining 2/5s.

This means they weren't sure that the sales would have failed or not as they didn't cancel the 2/10s. So that decision came after mid 1984. This NYT article describes overall sales/profit, but no breakdown by machine type, only says the consumers lost interest in the Apple ][s:

https://www.nytimes.com/1985/06/15/business/apple-cuts-1200-loss-seen.html

Apple Computer Inc., battling a fierce downturn in personal computers as well as problems in its own ranks, said yesterday that it was laying off 1,200 employees, more than a fifth of its work force.

In a terse statement from John Sculley, president and chief executive, who consolidated his position atop the company in a reorganization two weeks ago, Apple also said that it expected a loss from write-offs in its third quarter, ending in June.

Apple spokesmen said they could not estimate the size of the loss, but sources close to the company predicted that the pretax write-offs could total as much as $30 million, giving the company an after-tax loss of $12 million to $18 million. Last quarter, the company earned $10 million, on sales of $435.3 million. ''We expect these steps to significantly reduce the break-even point of the company,'' Mr. Sculley said in the statement.

And this one: https://www.nytimes.com/1985/06/01/business/the-restructuring-of-apple-computer.html

In a move aimed both at streamlining Apple and moving it out of the shadow of its co-founder, Steven P. Jobs, the company announced a corporate restructuring that does away with its two separate product divisions, in favor of manufacturing and marketing divisions responsible for all Apple products.

The restructuring eliminates the day-to-day operating responsibilities of Mr. Jobs, who had been general manager of the Macintosh division. Mr. Jobs will remain only as chairman of the board, and will take what the company called a ''more global role'' in product development and strategy formulation.

Apple insiders and analysts saw the move as an urgent attempt by John Sculley, president and chief executive, to assert leadership at a time when the company is beset with numerous problems. There is a growing feeling on Wall Street and in Silicon Valley that Mr. Sculley, who was recruited two years ago from Pepsico Inc., must act soon to turn the company around.

'It's Serious Now'

''It's serious now; it's not the time for fun and games,'' said Michele Preston, an analyst with L.F. Rothschild, Unterberg, Towbin.

Apple is reeling both from the general slowdown in personal computer sales and problems within the company. Analysts now say it will barely break even in the quarter ending June 28, and might even lose money for the first time in its history. Because the summer quarter is traditionally slow, Apple's business is not expected to improve before Christmas.

The reorganization is just the latest step Apple has taken to try to reduce costs and improve its management. It has scaled back advertising, closed factories for a week, eliminated its Lisa computer and some development efforts, and laid off more than 1,600 people. Today's move is expected to produce more layoffs. #2 Product Units Eliminated The reorganization eliminates the two separate product divisions - one for the Macintosh and the other for the Apple II - and replaces them with an operations division and a sales and marketing division. As important as the structure, however, are the executives Mr. Sculley has chosen to head the new divisions.

Delbert W. Yocam, executive vice president, an operations-oriented manager who formerly headed the Apple II division, will become the group executive in charge of product manufacturing, development and distribution for all Apple products worldwide.

William V. Campbell, formerly executive vice president of sales, will head United States sales and marketing. Previously, marketing was split between the two product divisions and sales. Michael Spindler will remain vice president in charge of international marketing and sales.

Even before today's reorganization Mr. Sculley had been moving in a new team. He brought Jean-Louis Gassee, head of Apple's French operation, to replace Michael Murray, who, like Mr. Jobs, was young and enthusiastic but relatively inexperienced. In the new reorganization, Mr. Gassee will become head of product development, reporting to Mr. Yocam.

I guess now we're on the quest to find the highest AppleNet ID for a 2/10, but so far it looks like 10K is the right number, not 100K.

[jamesdenton]

I have 3 Lisa's at home, but only two are easily accessible at the moment, so I'll return with information on the 3rd later.

SN: A3195247, AppleNet: 00102880, MFG: 3195 - Bought in 2018 off eBay from a seller in the US that brought it back from Germany years ago. This is a 2/5 with a 1A 220V power supply and German keyboard.

SN: A3165051, AppleNet: 00102513, MFG: 3165 - Bought in 2017 off eBay from a seller in the US, likely a member of this forum (or at least Google Groups). This is a 2/5 with a 1A 110V power supply and US English keyboard.

[jamesdenton]

From the Lisa recently mentioned here (by mjposner) - not mine:

SN: A4283098, AppleNet: 00106226, MFG: 4283 - This is the 2/10 recently sold by mjposner on eBay (https://www.ebay.com/itm/393245595464)

[jamesdenton]

Another from eBay:

SN: A3334145, AppleNet: 00108638, MFG: 3334 - This is a 2/5 currently listed on eBay (https://www.ebay.com/itm/194102391362)

[jamesdenton]

SN: B08B820200058, AppleNet: 00100067, MFG: 82020 - This is not mine, but one I was tracking a few years ago. From 68kmla.

[stepleton]

Here's my Lisa 1:

    Serial number: B08B831590504
    Applenet number: 00102905
    Manufactured: 83159

And my Lisa 2/10:

    Serial number: A3257425
    Applenet number: 00106191
    Manufactured: 3257

In the spreadsheet, I suspect the serial number for a CHM machine listed as "808B831990176" is probably "B08B831990176". The B looks a lot like an 8 on my Lisa 1; maybe an artifact of the machine they used to make those labels.

[jamesdenton]

One more for you:

SN: A3321293, AppleNet: 00107035, MFG: 3321 - This is a 2/5 currently listed on eBay (https://www.ebay.com/itm/393292386413).

[jamesdenton]

Alright, I can't find the sticker for my 2/10 and my attempt to reverse engineer the serial has turned up a match in the spreadsheet already

yy  ddd  uuuu  ppp  nnnnn
83  162  0256  001  04204

I think this comes out to:

SN: A31620256, Applenet: 00104204, Date: 3162

Here's what I see in the spreadsheet:

02831620256   00104204         blusnowkitty   chassis and CPU don't match; S/N is in ROM format

So, either I bought this 2/10 from you, or the vsrom that's installed came from a shared source (VintageMicros, perhaps)? It's possible this thing came to me with a 3A ROM and that's what I had on-hand, I don't recall!

[blusnowkitty]

02831620256   00104204         blusnowkitty   chassis and CPU don't match; S/N is in ROM format

So, either I bought this 2/10 from you, or the vsrom that's installed came from a shared source (VintageMicros, perhaps)? It's possible this thing came to me with a 3A ROM and that's what I had on-hand, I don't recall!

That's probably the VintageMicros ROM - I know I've never had a 2/10 (but I need one for a project!) and I legit could not remember if I installed the VintageMicros ROM in mine or not. Now to figure out what happened to my original ROM...

[compu_85]

Here's 3 more:

2/5
SN: A4186075
Applenet: 00106435
Manf'd: 4186
Notes: This computer was bought by Virginia Tech in 1985, then sold to a student in the CS program. Speaking with him, he said VT bought a bunch of Lisas and sold them to students, with UniPlus Unix. It was sold with a 10m Profile.

2/10
SN: A4293063
Applenet:00106665
Manf'd: 4293
Notes: Came from a meat packer's called Rock River provision in Indiana, they stopped using it around 1993. Had Macworks on the Widget.

2/10
SN: A4192153
Applenet: 00109415
Manf'd: 4192
Notes: Accelerated square pixel machine. Unsure on history.

[conmega]

Creating a list like this is a great idea. Hopefully it'll provide some good insight.

Here are 4 Lisa's I picked up recently.
A Lisa 1?, 2/5, 2/10, 2/10
I passed the later looking 2/10 to a friend since I don't need two. I kept the earlier one as it found it interesting historically, I need to write up the differences I found in another thread at some point. But for example the newer stylized Lisa logo that is normally on the bottom backplane of the 2/10 is not present of the earlier example I have. Its also an earlier rev of the backplane. But it seems that one has a later Applenet No? So that's odd.
And yes I did check, the video rom's SNs do match on both 2/10s. I have yet to check on the Lisa 1 or 2/5.

Lisa 1? Not fully upgraded to a 2/5? Twiggy rom (40), R47 snipped, drive cage and face plate missing.
Has Aluminum Apple Asset tag.
SN: B08B823550042
Applenet: 00100049
Mfg: 82355

Lisa 2/5:
SN: A4159392
Applenet: 00105235
Mfg: 4159

Lisa 2/10 with earlier rev boards inside?, but later AppleNet No?:
SN: A4197089
Applenet: 00111264
Mfg: 4197

Lisa 2/10 with later rev boards, but earlier AppleNet No? (one sold to friend, no longer in my possession):
SN: A4295043
Applenet: 00107051
Mfg: 4295

Here is an album with photos of the machines and their stickers incase I typo'd or you want to verify:
https://photos.app.goo.gl/oEZVfLgCHd4mkSdm9

EDIT:
I have been poking around google images as-well and found another data point:
https://www.worthpoint.com/worthopedia/vintage-apple-computer-a6sb100-screen-1867087741
Lisa 2/5 upgraded from Lisa 1?
SN: B08B820200262
Applenet: 00100271
Mfg: 82020