HP policy on source files



#52

Does HP release the source and/or documentation files in electronic format of "ancient" calculators like HP41, HP71, HP71 and so on?

Best regards

Ray


#53

In general, no, but the museum has special permission. You canc getv it on the museum DVDs here:

http://www.hpmuseum.org/cd/cddesc.htm


#54

Is there an official "permission statement" from HP, somewhere on the CDs? Where would it be?

Joel


#55

Maybe Richard can help ...

#56

If you have the source files, we'd be interested in them! :-)

TW


#57

So this is why we have not seen a 15c+ yet.


#58

Same reason you will not see the 15D, 15E, 15F an so on ... because HP will not make these models. Same reason you will not see a TI59+ from TI. With limited resources, HP is most likely focusing on combining new (reads hip and exciting) and practical (that is stuff that sells well) technology. If HP will revive any vintage machine, I'd say it's the HP41CX because the 41C/CV/CX sold very well (check the number of these auctions in TAS).

As for me, I want my HP41C+ .. and I want it now ... with at least one or two USB ports to support USB printers and USB external drives!!! I hope Tim is listening!

Namir

Edited: 15 Feb 2011, 11:26 p.m.


#59

Quote:
As for me, I want my HP41C+ .. and I want it now ... with at least one or two USB ports to support USB printers and USB external drives!!! I hope Tim is listening!

Yeah. . . I'll get right on that. . .

Give a few days and I'll get that to you on the back of my flying unicorn.

TW

Edited: 15 Feb 2011, 11:47 p.m.


#60

Hi Tim,

Can you ask Scotty to beam me a HP41C+ too ?

Since I live over sea, I fear your unicorm will not do it.

Patrice


#61

It is a robust unicorn. I think it will make it. . .

TW

#62

There are 7 billion people on Earth and HP has only 1 calculator programmer. :-(


John


#63

It's not as bad as that! Unless something changed very recently, they have TWO programmers, Tim and Cyrille.


#64

Who wrote the firmware for the HP 30b? That's a very nicely done recent product that does *not* rely on an ARM emulating HP firmware from some ancient HP of two or three decades earlier.

My dream would be a small "scientific" machine with completely original new firmware native to some version of an ARM with good LCD controller capability and the ability to use lots of RAM. The implementation of HP 42S capabilities in an HP 30b-sized and -styled package would be an exciting product, especially if the firmware were optimized to the new machine's processor. The HP 30b is 100 times faster than the HP 42S for the types of non-finance problems that are not limited by the 30b's characterisitcs, and ten times faster for the same problems done on the ARM-based but Saturn-emulated HP 50g. Native code makes a big difference.

So...for new calculator development, the existence of legacy firmware from an old project should be of only historical (not production) value, IMO.

Edited: 18 Feb 2011, 3:02 p.m.


#65

Quote:
Who wrote the firmware for the HP 30b?

Most SDK files bare Cyrille's name.

#66

Quote:
My dream would be a small "scientific" machine with completely original new firmware native to some version of an ARM with good LCD controller capability and the ability to use lots of RAM. The implementation of HP 42S capabilities in an HP 30b-sized and -styled package would be an exciting product, especially if the firmware were optimized to the new machine's processor.

Mike, you nailed it there.
Quote:
So...for new calculator development, the existence of legacy firmware from an old project should be of only historical (not production) value, IMO.

In theory, yes. However, as we have seen in the case of the 33s and 35s, for whatever reason HP did not start from scratch, but apparently used the 32sii firmware as a starting point. Ideally, write new firmware implementing the functionality of both 42s and 27s, but if that is not possible, then start with 42s code and modify it.

#67

The 33s / 35s used the 32sii starting point for some obvious reasons.

Top of the list was the pretty solid market for the PE exams.
Along that same list was to match the functionality of the 32sii which is probably the most popular model ever made.

I( can't fault them for that approach. Too bad they didn't put more rigor into testing or being involved more directly or something so as to avoid the bug problems that had to be sorted, and the hardware (keyboard) issues that cropped up. They sorted it but got some bruises around the eyes.

#68

Quote:
in the case of the 33s and 35s, for whatever reason HP did not start from scratch, but apparently used the 32sii firmware as a starting point

I wasn't involved, but I don't think the 33s and 35s used ANY of the 32SII code. The 33s code appears to have been a from-scratch effort, most likely written in C, and intended to have functionality similar to the 32SII. Why? Because it sells!

Quote:
implementing the functionality of both 42s and 27s

Obviously a bunch of us here in the forum would buy that, but it's far from clear that the world at large would. HP has to invest their resources into designing what they think the market will actually buy.

Quote:
then start with 42s code and modify it.

I very much doubt that you'll see any calculators with code derived from any Pioneer series models, as those were all written in Saturn assembly and/or RPL. Any new models (other than 38/39/40/48/49/50 series) are likely to be derived from 20b/30b code base.

I don't work at HP, so this is all just my opinion.


#69

Quote:
I wasn't involved, but I don't think the 33s and 35s used ANY of the 32SII code. The 33s code appears to have been a from-scratch effort, most likely written in C, and intended to have functionality similar to the 32SII.

Not being a software guy, I did not use the correct terminology. Obviously, "code" was the wrong word. What is the proper term, a flow chart? What I'm trying to say, is HP, or whoever they source it to, used the 32sii functionality as a starting point.
Quote:
Why? Because it sells!

It sold because that's all HP offered at the time! It has been demonstrated here that the 42s was far superior as an RPN machine, and the 27s has a far superior Solver than any other pocketable calculator.

But even given that 42s sales faltered before it was canned, nonetheless if indeed written from scratch, the 33s/35s functionality could have been vastly improved over the 32sii.

Rather, I think Bill Platt hit on the real reason, passing muster with the NCEES folks with the least amount of effort.


#70

Quote:
What is the proper term, a flow chart?

I doubt that they even had a flow chart to work from. I suspect that the 33s code was developed entirely from scratch. (Well, probably with some code from the 10bii or something.)

Quote:
It [32SII] sold because that's all HP offered at the time!

No, that's not the reason. The 32SII sold far better than the 42S even when both were available.

Quote:
the 42s was far superior as an RPN machine, and the 27s has a far superior Solver than any other pocketable calculator.

That has nothing to do with it. Adding more features to a product does NOT necessarily make it sell better.


#71

Quote:
That has nothing to do with it. Adding more features to a product does NOT necessarily make it sell better.

Not more, a better feature set. Both 33s and 35s have received mixed reviews, even on consumer websites, let alone this one. I don't think you mean what customers say has no bearing on sales.
#72

Quote:
That has nothing to do with it. Adding more features to a product does NOT necessarily make it sell better.

In any event, not when the marketing is so incompetently prosecuted or purposefully self-defeating as it was with the HP 42S.

I'd used my favorite, an HP-15C, for three years before and eight years after the HP 42S appeared. Had I known the tremendous advance that the HP 42S represented over the HP-15C, I'd have purchased a 42S many years earlier.

When I finally retired my HP-15C, the HP 32SII was the only thing still in the RPN product line that seemed to offer similar capability. I found out after buying a 32SII that it wasn't even *remotely* as capable. Other than having the advantageous portrait layout, a limited alpha capability, and Saturn processor (for greater speed and accuracy), it was incredibly inferior to the old HP-15C. I tossed it in the clunker box (with other junk like the 28C and 38G) and found a source for the HP 42S. (In retrospective fairness, the 32SII is not a bad machine...unless one is comparing it to a 15C or 42S.)

Those of us who had used HP machines for decades made the mistake of assuming that the best models of one year would be replaced by even better models in later years. Sadly, HP broke this tradition in 1995. We're still waiting for a worthy 42S successor today, 23 years after introduction and 16 years after discontinuance.

The reason that the HP 42S did not sell as well as it might have is HP's piss-poor, shoddy, self-defeating marketing. I saw many HP advertisements in US engineering trade magazines such as the IEEE Spectrum for the HP 48SX. I recall none for the HP 42S, even though it is an electrical engineer's dream machine (and still is, even in 2011).

The reasons that the HP 32S/32SII sold adequately before the HP 42S was discontinued was penny-wise-dollar-foolish buyer attitudes (the 42S was 70 percent higher in cost than the 32SII), ignorance of 42S capabilities (due to HP's poor marketing of the 42S), and lack of buyer sophistication (who could imagine no value to 42S capabilities). After the discontinuance of the 42S in 1995, the 32sii sold well simply because it was the *only* small RPN machine available, as Martin points out. This was aggravated by the prices of yet available HP 42S stock rising very sharply after 1995 as HP's strategy was recognized: HP deliberately killed the grossly superior (to 32SII) 42S in order to kill competition to 48-series machines. There would be no successor. Ever.


Edited: 19 Feb 2011, 11:24 p.m.


#73

I concur with almost everything you wrote. The habitat of the 42S was given up deliberately to promote the more powerful but clunky RPL machines. Their margin must have been significantly better.

Now we have the toys at the low end (up to the 300s), the 35s for US tests (and as a basic scientific calculator with some well known shortcomings), and the 50g for the nerds at the universities etc. That's it. A powerful compact RPN tool is missing since 1995.

One more reason for pushing community projects ... d8-)


#74

It would be interesting to know the number of downloads for Free42 and other 42s implementations to be able to judge the interest in such a product.


John


#75

While the number of downloads of Free42 would certainly be interesting, it would tell almost nothing about how a physical calculator would sell.


#76

Eric, I don't follow your logic. Let's say that Free42 and 42s have less than 100 downloads from iTunes then I wouldn't be interested in further investigation. However, if those products have had 10,000 or more downloads then I would be interested in looking into the potential for a market for a hardware calculator.

It really doesn't matter. HP is going to continue to build financial calculators and do very little with scientific / graphing calculators. With only 2E0 programmers on a planet with 7E9 people it's my opinion that they aren't interested any more.

Regards,

John


#77

What I'm saying is that if Free42 has 10,000 downloads, that does NOT mean that a large number of people would buy a physical calculator, and conversely, that if Free42 has 100 downloads, it doesn't mean that a physical calculator would sell poorly. The markets for samrtphone calculator apps and physical calculators are MUCH different. It's certainly true that there is a non-zero intersection, but the size of one set tells you almost nothing about the size of the other.

I don't have numbers for Free42 downloads, but it is definitely the case that far more than 10,000 people have purchased various scientific calculator apps for the iPhone. If that makes you "interested in looking into the potential for a market for a hardware calculator", more power to you, as I'm all in favor of that and would love to see it happen. Heck, I've been working on that myself for years, as anyone who has attended HHC conferences is undoubtedly aware.


#78

I'm guessing here but I believe that Free42 users are either 42s owners or are those that are RPN calculator users that want to be 42s owners. There should be significant overlap between the two sets.


John

#79

... and by clearly stating there are 7e9 people on this planet you may even improve the relation by a factor of 1000 (pun intended ;)

#80

As much as I too want the same dream-machine Namir refers to, I'm not such an ambitious dreamer... but what about the 42S++??

Or was this the machine that allegedly got its source code trashed whilst still being in production, and therefore LOST-FOR-EVER?

Oh well, the 41CL is around the corner (fingers crossed), rejoice!


#81

Quote:
As much as I too want the same dream-machine Namir refers to, I'm not such an ambitious dreamer... but what about the 42S++??

What about the 34s??? A bit light on RAM but otherwise very well featured.

We've got the option to take out many inbuilt functions and replace them with user code alternatives. Less accurate but way smaller leaving plenty of space for other things. I've toyed with a matrix library in the xrom space :-) Likewise, I'm considering TVM support.


Then again, a 32k RAM upgrade for the 42s makes it extremely impressive.


Quote:
Oh well, the 41CL is around the corner (fingers crossed), rejoice!

Most definitely! I've got a pair of broken 41c's to make whole and convert into a CL. Failing that, I'll happily sacrifice a 41cv for the same purpose.


- Pauli


#82

Quote:
Then again, a 32k RAM upgrade for the 42s makes it extremely impressive.

... if some way of reliable I/O will be supported for this - else it will be impressively outdated.

Ceterum censeo: HP, launch a 43S (read the archives to collect the features necessary)!

Walter

#83

Angel,

Yes you can add an HP-43S++ .. I am with you.

I do have fun with emulators on Windows and the Apple iPad/iPod. I have iPad emulators for the 21, 35, 45, 55, 41, 42S, 48GX, 11C, and 15C. I enjoy the convenience of these machine. On the PC, I enjoy the HP-15C, HP-41CX, and Tom Okken's Free 42s (I run it with an HP41 skin!!!). Using custom skins is an awesome part of customizing the appearance, as well as the keyboard color and orientation of these emulators.

Namir

#84

That would be nice. I would also be happy with a Voyager with USB and HP41 style LCD.

Quote:
As for me, I want my HP41C+ .. and I want it now ... with at least one or two USB ports to support USB printers and USB external drives!!! I hope Tim is listening!

Namir


#85

I am listening (but I am not Tim). I do have a breadboard with the new MLDL2000 V2 running, and it will save and retrieve data on a USB stick (and provide ethernet connectivity as well) .... plus much more

Meindert


#86

jaw-dropping and eye popping don't come close to accurately reflect my feelings :)

Edited: 17 Feb 2011, 1:45 a.m.

#87

Here is a recipe to get your favourite Series 10 model with a + on tail.

Since all the Series 10 models are based on the same hardware, and the HP12C+ is a Series 10 hardware emulator

Then :

1) put your hand on an HP12C+.
2) find an image of the HP16C ROM (my choice)
3) go to Fort Collins, break into Tim's office and thief the tools and the ROM image of the actual 12C+ (since this image is not publicly available, as far as I know)
4) swap the 16C image with the one of the 12C inside the 12C+ image
5) load the new image the the 12C+
6) do some painting

and tada... you have a magnificent HP16C+ on your desk.

Just a dream :(

Patrice

#88

In my experience, many computer programmers "accidentally" leave copies of their work on paper tape, mag tape, floppy drive, CD ROM, DVD or thumb drive (depending on the era) at home. These copies have a curious way of remaining intact.

So the best way to find lost source code is to track down the original developers. For the cost of a release of liability and maybe a free plane ticket, you might get the lost source code.

Dave

#89

HP published the source code for the HP-71B mainframe ROM and the ROM of the related HP 82401A HP-IL interface as volumes of the Internal Design Specifications which were available for purchase by customers. These were the first Corvallis Division products for which internals documentation was made officially publicly available.

I would guess that this official relese of HP-71 internals documentation was done at least partly in response to the amazing amount of third-party development of microcoded ROM code and peripheral hardware for the HP-41 family. That third-party support came about as a result partially of informal release of technical information on HP-41 internals to the user community via the PPC club, and partially of a great deal of reverse-engineering effort by members of the user community. HP likely believed based on the experience with the HP-41 that providing formal support to third parties would result in even more community and commercial support for the HP-71B.

It seems evident that there was a significant expense incurred by HP in preparing the 71 IDS documents for public release. While HP normally did prepare internal documentation for computing platforms, the HP-71B IDS appears to have been prepared in a much more formal and comprehensive manner.

It would be interesting to hear from the individuals involved in the HP-71 IDS as to how successful the program was perceived to be.

In the case of the HP-41, the "VASM" source listings of the mainframe ROM and several plug-in ROMS were made available to the PPC club to be distributed to it's members as Non-Manufacturer Supported ("NOMAS"). I have heard an interesting story about how that came to pass, but not having been party to it, I'll leave it up to others to tell the story or not.

The HP-75 mainframe ROM listing was also provided as NOMAS.

In some cases HP provided other source code and internal documentation to third parties, but under NDA and thus those documents are not publicly available.

Unfortunately, in the handoffs of responsibility for calculator products to a series of other HP divisions after Corvallis Division, much (perhaps most) of the engineering data and documentation for many calculator models was destroyed. This actually included the source code for at least one product still in production at the time! While this sounds inconceivable to people who have not been involved in engineering departments at large companies, I can state from personal experiences at multiple companies (other than HP) that this is actually not uncommon.


#90

Source files (I mean not scanned pdf's but the real sources) on a stick would be a very nice goodie from HP at the next HHC .....


#91

Quote:
Source files ... would be a very nice goodie from HP at the next HHC .....

But Tim has already implied that HP doesn't have them.

#92

I should have put the FROM HP in the foreground.
Creating source files should not be such a problem (reverse engineering, OCR scanning of IDS docs and so on).
With the "FROM HP", I mean: usage without any limitation ....


The HP71B IDS docs have clear (c) marks on it,
prohibiting above techniques to convert them into "legal" copies of source files, ROM images and so on.

A lot of countries have a copyright by default even when nothing is claimed by the author and in these countries an explicit waiver is required.

A (legally binding) statement from HP concerning "ancient" calculator engineering work,
e.g. putting it under GPL, would be of great help.



#93

Did HP put any source-code listing in their handheld calculator patents?

From my limited knowledge of legal matters, I understand that anything that is written in a patent automatically moves to the public domain when the patent expires. Thus, any source code which was patented before, say, 1990 is no longer protected.

Could anyone in this forum confirm -- or not?

Joel


#94

AFAIK

(I am NOT a legal advisor!!)

US: roughly for work created since 1978 the copyright protection is
based on the lifetime of the author+ x? years, with multiple
authors the longest living author with a maximum of 120 years



Germany / Netherlands (Europe ??)

Like US up to 70 years after death (inheritable).



Rights on Patents is something completely different!



Since most employers write in the contracts with their employees
something like: "all rights concerning inventions, patents,
copyrights and so on shall be deemed to be transferred to the
employer at the time of creation of it", I assume that HP holds
all the rights.

HP can check this and act if they want ... :)



#95

Berne convention.

US was not a signatory back then, so anything prior to that in the US had to include "copywrite" somewhere within that or it is not protected. Thus there are plenty of 12c software versions around that HP can't do anything about. The ROM didn't include any sort of copywrite notice within it.

TW


#96

What about the copyright notice molded on the back of the calculator itself?


#97

Once the firmware is extracted, there is no notice. . .

TW


#98

TMO copyright issues have nothing to do with the visibility of a notice on a part of the product.



I would like to hire your care for a while. For sure, you get your chassis-identification plate back (but thats all!)



:)



PS: I Hope you have a at least a 540i ....

#99

Quote:
Once the firmware is extracted, there is no notice. . .

So to use this logic, I could "extract" the pages from a book (the physical kind), leaving copyright pages in the cover, then proceed to do what I want with the "extracted" portion? I certainly don't claim any special knowledge of copyright law, but this makes no sense. Then what was the purpose of molding the copyright notice on the back of the device to begin with?

Edited: 17 Feb 2011, 10:27 a.m.

Quote:
Creating source files should not be such a problem (reverse engineering, OCR scanning of IDS docs and so on)

Creating useful source code by reverse engineering is surprisingly difficult.
  • You can't just disassemble the entire ROM because some bytes are code and some bytes are data. Figuring out which is which requires finding and analyzing the code.
  • Even if you manage to get a disassembly, the lack of labels or comemnts means that it's hard to figure out what is doing what.
  • Within the disassembly, you'll find numbers being used, but what do the numbers represent? Addresses? Offsets from addresses? Binary numbers? The distance between two addresses?

Dave

I believe it is common. There is no money to be earned from keeping obsolete data. Eventually it needs to be migrated to another computer system or department, and then it just vanishes.

Quote:
Unfortunately, in the handoffs of responsibility for calculator products to a series of other HP divisions after Corvallis Division, much (perhaps most) of the engineering data and documentation for many calculator models was destroyed. This actually included the source code for at least one product still in production at the time! While this sounds inconceivable to people who have not been involved in engineering departments at large companies, I can state from personal experiences at multiple companies (other than HP) that this is actually not uncommon.


About 1985 HP made a pile of used HP-110 Portable computers from their sales force available to employees at an auction for $250 each. Sam Chau, a young HP PPC / CHHU / etc. enthusiast and I bought one each. Sam then figured out how and modified his to have 512k of RAM instead of the stock 288k. The natural next step was to modify the firmware ("PAM", the Personal Application Manager) to allow use of the added 224k.

Being HP employees, we contacted Corvallis and asked if the source was available NOMAS (No Manufacture Support - i.e. once we get it we wouldn't bother them with questions) and sent them the details of how it was done. They enthusiastically agreed to provide us the source as many there had Portables and wanted the same capability for theirs.

A month later we were called back. An extensive search in Corvallis had revealed that the source hadn't been backed up and the 8 inch floppies with it had been accidentally reused for the Portable Plus. So they were shipping computers without having the source code for what they were shipping.

Disassembling PAM revealed that the RAM size was hard coded into it in many places. Sam finally just wrote a driver to treat the extra 224k as a second RAM disk. Still works fine...

The one benefit of this is that Corvallis promptly backed up what they had then and put it away. Whether those backups survived to this day is altogether another story!


Possibly Related Threads...
Thread Author Replies Views Last Post
  How to copy files from 9114B to hard drive via PIL-box? Joe Horn 23 3,824 12-14-2013, 11:57 PM
Last Post: Geoff Quickfall
  How to update PRIME Firmware using Files on PC Harold A Climer 2 693 12-04-2013, 12:05 PM
Last Post: Erwin Ried
  Loading .prgm files to the Prime Thomas Chrapkiewicz 8 977 10-25-2013, 09:34 AM
Last Post: Eddie W. Shore
  Proper location for files on the PC for Connectivity Kit ,etc. Harold A Climer 8 1,051 10-23-2013, 02:43 AM
Last Post: Marcus von Cube, Germany
  Updated PPC DVD Version 2.10: HP-41 Searchable Program Files and Scannable Barcode Jake Schwartz 3 603 09-27-2013, 09:51 PM
Last Post: Olivier (Wa)
  HHC 2013 speakers' files Joe Horn 2 559 09-26-2013, 02:51 AM
Last Post: Geoff Quickfall
  Nonpareil Rom files davorin 2 454 09-20-2013, 06:20 PM
Last Post: Didier Lachieze
  any open source HP 10BII emulators? John 15 1,528 06-12-2013, 09:58 AM
Last Post: Kimberly Thompson
  HP48 source Han 0 279 05-26-2013, 03:34 PM
Last Post: Han
  Source for N-Cell NiMh Batteries and Charging adapters TnC_Products 14 1,315 12-16-2012, 10:26 AM
Last Post: Dieter

Forum Jump: