At least some companies listen (or check the auctions)



#60

I like listening to music throughout the day, and I've had my fair share of lousy headphones. I had a recommendation once that Koss makes some great sounding clip on earphones, the KSC-35. However, by that time, the KSC-35 had been discontinued and replaced by a futuristic looking KSC-75 (with the same speaker element). KSC-35s were going for over five times the original price on the auction site. I recently checked Koss' site, and saw that they had the KSC-35 back again, along side the -75. I clicked on the live chat and the person at customer service told me that they were back by popular demand, and were now made with higher quality materials for the target market. So, at least one company notices when a discontinued product is popular and selling for a premium and takes that as an opportunity to give the market exactly what it wants.

-Jonathan


#61

Jonathon, not to dispute what you said about Koss (I had one of their headphones years ago, and it was great until the material that fit over your ear started disintegrating), if you are suggesting that HP should reissue the 15c or 42s or even the 41, consider the size of the market. In this age of ipods, iphones and what have you, the market for a quality headset is huge, probably in the millions. The market for a 15c reissue, I would guess, is pretty tiny, probably in the thousands. I think if HP really thought they could make some bucks, they'd do it in a minute. The bottom line drives most companies, HP included.


#62

Quote:
Jonathon, not to dispute what you said about Koss (I had one of their headphones years ago, and it was great until the material that fit over your ear started disintegrating), if you are suggesting that HP should reissue the 15c or 42s or even the 41, consider the size of the market. In this age of ipods, iphones and what have you, the market for a quality headset is huge, probably in the millions. The market for a 15c reissue, I would guess, is pretty tiny, probably in the thousands. I think if HP really thought they could make some bucks, they'd do it in a minute. The bottom line drives most companies, HP included.

According to THIS website they can sell at least 53,870 of them. The changes required from the 12C are simply new key tops and badging and stick in the original 15C ROM, they have already done all the engineering. If they couldn't sell 100K of them then they would have to have seriously goofed it up, like chose the wrong price point or something.

I'll bet serious money Cryrille already has a prototype on his desk. I'm surprised it hasn't been released already, it's a complete no-brainer. Even marketing people know that niche retro products can sell very well.

And it actually makes sense from a marketing point of view given they have revived the retro look with the 35S and updated the 12C. The market almost expect this from HP now.

The 42S is massively harder to do, so forget that.

Dave.


#63

A prototype of an ARM-based 15c was seen at a previous HHC conference. There is no question that HP has the technical expertise to produce a new 15c if they choose to do so.

From a business perspective, though, if they only expected to sell 50-100K units, it would be somewhat questionable as to whether it would make sense to introduce it as a product.

Edited: 21 Oct 2009, 7:31 p.m.


#64

Quote:
From a business perspective, though, if they only expected to sell 50-100K units, it would be somewhat questionable as to whether it would make sense to introduce it as a product.

Perhaps, depends more on corporate attitude than economics. At the current 12C price point of $70, that's $3.5M to $7M in retail sales, not spectacular, but not bad either for a niche (struggling?) division of the company. If you couldn't crack a reasonable margin on 50-100K units then you probably shouldn't be in business. And if you are in the business, then STAY IN the business.

And considering that a good lot of the engineering has already been amortised on the 12C, the risk is fairly small and margins potentially wider.

And then there is the potential upside, if you market it right it could go through the roof. People love retro toys.

Dave.

#65

I'm not sure how accurate those statistics are today. 11,120 people buying 53,870 units? Sounds strange. I've signed that petition some years ago promising to buy at least 2. People change their minds though. If the suggested new edition of the 15C is identical to the original I won't buy it. A must-have things are I/O, mnemonics and more memory.

Cheers


#66

There is no serious proposal for a 15c with your "must-have" enhancements. Doing that would require HP to completely reengineer the 15c from scratch, which is extremely unlikely. The plausible scenario is for HP to introduce a new 15c based on the new 12c platform. It would be functionally identical to the original 15C but significantly faster.


#67

In this case they would sell 5,000 max - my guess. Needles to say, I'd withdraw my offer.


On a second thought I'll still buy a couple of those retro style machines - but that's me, ;)


Edited: 22 Oct 2009, 5:41 a.m.

#68

Making a version of the HP42S, perhaps a long-awaited HP42SII, would be the only winning strategy for re-issue of an older scientific calculator.

The HP-15C, as beloved as it has been, is a throw-back of more than a quarter century. The HP42S, OTOH, after 21 years is still the 2009 state-of-the-art in its RPN feature set and ease of use. But 2009 hardware could greatly enhance the speed, display, memory, and I/O that serve that outstanding feature set. (Let's choose a decent new color scheme too, something classic like that of the HP-15C.)

Edited: 22 Oct 2009, 4:28 p.m. after one or more responses were posted


#69

Quote:
Making a version of the HP42S, perhaps a long-awaited HP42SII, would be the only winning strategy for re-issue of an old scientific calculator.

The HP-15C, as beloved as it has been, is a throw-back of more than a quarter century.


If by winning strategy, you mean greatest appeal as a functional RPN calculator for the working professional, then you are right.

But this misses the point, IMO. A re-issue of the 15c using the new 12c+ platform, by comparison has such trivial developmental costs (easy for me to say!), that it is more likely to happen first. So as an armchair (office chair?) expert, the 15c appears to be a sounder business decision, even if the market is smaller.


#70

The question is: Why is the 12C so successful, despite of lacking all the bells and whistles of more modern constructs like the 20b or the 17bii+ ?

Because it hasn't changed for decades. It's a dependable tool!

This might work for a scientific calculator, too.


#71

The HP42S would have been, I suspect, a continuous success had it continued to be marketed after 1995, especially if the rather weak HP32S/HP32SII RPN scientific had been dropped instead. Few of us, myself included, ever dreamed that the best RPN scientific calculator ever made would be discontinued without replacement by something equivalent or better. Very shortly after the discontinuance of the HP42S was announced, it sold out everywhere and became very hard to get. It has remained so. It isn't because of collector appeal...it's because it is such a masterpiece of really really useful RPN features.

I also don't understand why a modernized HP-15C would require significantly less effort than would a modernized HP42S. The firmware of each of the originals would need to be translated to the new hardware...it's not as if a ROM image of the HP-15C could easily be translated, but that of the HP42S could not.


#72

Neither a modernized 15C nor a modernized 42S is easy. Either would need a large investment in new software development by HP. In the "good old days", HP had a fair number of software engineers to develop calculators. Today they have two engineers, who have many demands on their time. HP has to carefully decide what product development efforts are the most effective use of their available resources, including the time of those two engineers, and designing a modernized version of either the 15C or 42S is unlikely to make the cut, because neither is likely to be a huge hit. As much as we like the 42S, it is safe to assume that it was NOT selling nearly as well as the 32SII, or it wouldn't have been discontinued.

I don't have any special insight into what goes on inside HP; I'm just writing on the basis of how business and engineering resource allocation decisions are normally made.

#73

Quote:
The HP42S would have been, I suspect, a continuous success had it continued to be marketed after 1995, especially if the rather weak HP32S/HP32SII RPN scientific had been dropped instead. Few of us, myself included, ever dreamed that the best RPN scientific calculator ever made would be discontinued without replacement by something equivalent or better. Very shortly after the discontinuance of the HP42S was announced, it sold out everywhere and became very hard to get. It has remained so. It isn't because of collector appeal...it's because it is such a masterpiece of really really useful RPN features.

I also don't understand why a modernized HP-15C would require significantly less effort than would a modernized HP42S. The firmware of each of the originals would need to be translated to the new hardware...it's not as if a ROM image of the HP-15C could easily be translated, but that of the HP42S could not.


The 15C is much easier because they already have the platform in place with the 12C. It's simply a matter of using the 15C ROM instead of the 12C ROM, and probably some housekeeping code. The software side would have to be an order of magnitude less complicated than a new 42S, of which they have no basic hardware platform to work from. i.e. no LCD driver, no keyboard driver etc.
The 12C platform already has those drivers in place, as the LCD and keypad is the exactly same between the calcs.

The 42S would have been dropped for one reason, and one reason alone - it wasn't selling enough.

Remember, the fanboy and general purpose scientific markets are trivial, and HP know this, which why they market their calcs to target markets like education and finance.

If a new 15C or 42S comes out, it will because it's easy for them to do, has as little risk as possible, and can be marketed with retro appeal. The 15C is the only sensible choice here.

Dave.


#74

they also have a great platform for the 42S.

See 17BII+.

a different overlay, great KB and a 42SII can be achived.

regards


#75

Do you know how wonderfully amazing the 6502 chip inside that is? If you did, that statement would be a little different. . .

TW

Edited: 22 Oct 2009, 7:07 p.m.


#76

Isn't that what was in the Commodore Vic 20?


#77

Quote:
Isn't that what was in the Commodore Vic 20?

And the following:
  1. Commodore C64
  2. Commodore C128
  3. Commodore C16, C116 and Plus/4
  4. Commodore P500
  5. Commodore 600/700 family of computers
  6. Apple ][ family
  7. Atari 8bit machines (e.g. 2600)
  8. Oric Atmos
  9. Nintendo Entertainment System (NES)
  10. Supervision Game Console
  11. Atari Lynx Console
The 6502 is a very capable processor. Last weekend I computed pi to 1000 digits in just a bit over 25 minutes on my Apple //c (6502 @ 1.023 MHz). I used cc65 (cc65.org) to compile my C program.

Now emulating a different platform is a different story and I'll guess that the 17BII+ is not running at 1 MHz (or is it?).

However if the 17BII+ was a 17BII clean room edition, then a 42S clean room ed. should also be possible. And, all the work is already done in Free42. Ah, but then there is that large company with many lawyers "Not Invented Here" attitude.


#78

Quote:
Last weekend I computed pi to 1000 digits in just a bit over 25 minutes on my Apple //c (6502 @ 1.023 MHz).

Bob Bishop's Apple Pi (Integer BASIC) is about 100 times slower:

http://bob-bishop.awardspace.com/ApplePi/index.html

Listing in text format available at

http://groups.google.com/group/comp.sys.apple2/browse_thread/thread/110c648eaf9a4ebf/b5857de668f04adb?lnk=raot


#79

Quote:
Last weekend I computed pi to 1000 digits in just a bit over 25 minutes on my Apple //c (6502 @ 1.023 MHz).
6502 cores in custom ICs for embedded control have exceeded 200MHz. (Dig that sub-50ns interrupt latency!) They have more instructions and addressing modes, have all the old NMOS 6502 bugs fixed, and several other benefits. My 5MHz 65c02 workbench computer does a 2,048-point complex FFT in fixed-point in 5 seconds-- in high-level Forth, not even assembly.

#80

Sure, but these 6502-compatible cores don't achieve anywhere near the uA/MIPS power stinginess of low-power microcontollers, because people building ASICs generally don't have access to the extremely low power processes developed for those microcontrollers. And in general, a 6502 is going to have less than 1/10 of the equivalent MIPS of an ARM at the same clock frequency.

The 6502 core in the GeneralPlus (formerly SunPlus) microcontrollers being used in calculators only runs at a few MHz maximum, but is fairly low power.

Don't get me wrong, I actually like the 6502, and have written a huge amount of 6502 assembly language over the years, but it's NOT the best choice for an advanced calculator.


#81

Quote:
Sure, but these 6502-compatible cores don't achieve anywhere near the uA/MIPS power stinginess of low-power microcontollers, because people building ASICs generally don't have access to the extremely low power processes developed for those microcontrollers. And in general, a 6502 is going to have less than 1/10 of the equivalent MIPS of an ARM at the same clock frequency.
While it may not be the most suitable for a modern calculator (primarily because of its 16-bit data bus), it is nevertheless being sold at a rate of hundreds of millions of units per year today, mostly rather invisibly because it's going into custom ICs (ASICs) used in embedded control for products ranging from consumer to automotive to medical. Western Design Center, its owner, sells mostly IP licenses, not hardware. Of the little hardware they do sell however, their 65c02 takes a guaranteed maximum of 150µA per MHz at 1.8V for the core itself, and a guaranteed maximum of 500µA per MHz at 5V. 6502 MIPS is roughly 1/4 of the MHz speed, meaning the aforementioned 200MHz is worth about 50 MIPS. A PIC16, which I also have put in many products we've brought to market, would have to go about 400MHz to get comparable performance (regarding time to get a job done), if the PIC's poor instruction set doesn't keep it from doing the job altogether. I will estimate that a 65c02's performance gain over the 41, at comparable power, would be over two orders of magnitude. If I had to redesign the 41 and had to use a 65-family processor though, I would use the 65816 which is more powerful and even easier to program. In the Sieve benchmark, it performs approximately on par with the 68000, even though the 65816 is a much simpler processor.


Edited: 24 Oct 2009, 10:46 p.m.

#82

Quote:
Bob Bishop's Apple Pi (Integer BASIC) is about 100 times slower:

http://bob-bishop.awardspace.com/ApplePi/index.html


I was cleaning my office a while back and stumble onto an article by Ron Knapp about how he bested an Apple II with a 41C when computing Pi to 1000 digits. I think the times were 40 hours vs 8.5 hours.

Being an Apple II fan that didn't sit right with me, so I decided to see for myself. I benchmarked both and also got 40 and 8.5 hours. Ron was right.

So, I decided I'd figure out what the Apple could do. I wrote 3 programs to compute Pi to 1000 digits--Forth, K&R C (Aztec), and ANSI C (cc65). The Forth and K&R C were available at the time of the Apple and the 41CX. The ANSI C compiler is very modern.

Times:

                            Apple                     HP
------------------------------------- -------
INT MAF Aztec cc65 41CX
Benchmark BASIC Forth C 3.2 2.13 RPN
========= ====== ========= ========= ========= =======
pi (atan) ~40 hr 24.93 min 36.33 min 25.23 min ~8.5 hr

Ron's HP time (1981) wasn't bested until 1985 when someone posted a USENET C program for the Apple. The author claimed 40 minutes, but I timed it at closer to 2 hours. I contacted the author to see if he could remember the exact compiler and system used, but he could not. Our algorithms were similar, but he computed with base 10 whereas I used base 216. My code needed much fewer operations. My guess is that he created a base 216 version and benchmarked that, but posted an older base 10 version. He acknowledged that may be the case and said he'd search his backup tapes.

Not that it matters. 40 min or 1:52 hours. In any case Ron's HP was bested after a 4 year rein.

Challenge: Can a 41C MCODEr best the Apple II?

Edited: 24 Oct 2009, 5:56 p.m.

#83

I don't know about the other computers but the c64 used a 6510 which was an upgraded 6502 and subsequent commodores also used upgraded versions.


#84

I should have said, "6502 family". The 6502 has many permutations, e.g. the Atari 2600 used a slimmer 6507 that only had 8KB of addressable space.

#85

Quote:
And the following:

Don't forget the Commodore PET, the predecessor of all the Commodore machines in your list.
#86

I know it, look at my post with an hp17BII+ disassembled.

If you see my post you will see...changing overlay and keeping the keyboard, nothing about inside hardware is related in my post.

regards,

nacho

#87

Quote:
Making a version of the HP42S, perhaps a long-awaited HP42SII, would be the only winning strategy for re-issue of an older scientific calculator.

The HP-15C, as beloved as it has been, is a throw-back of more than a quarter century. The HP42S, OTOH, after 21 years is still the 2009 state-of-the-art in its RPN feature set and ease of use. But 2009 hardware could greatly enhance the speed, display, memory, and I/O that serve that outstanding feature set. (Let's choose a decent new color scheme too, something classic like that of the HP-15C.)


I think you are missing the point.

Either calculator will sell based mostly on it's retro appeal and perceived coolness, it's got little to do with its functionality.

The majority of buyers will NOT be HP fanboys.

It is quite possible (most likely IMO) that a retro 15C as-is would outsell the pants off a new super-duper 42S for a few reasons.

The first of which is the NRE which is extremely low for the 15C. The 42S would be a complete redesign from the group up using a new platform. This means the 42S would need a large development cost amortisation which would increase the price point and make it even less appealing to buyers. Not to mention hardware wise it's going to cost more than the 15C to begin with. It anything it would likely be based on the new 20B platform.

And a 42S would look like regular modern calcs, so it's not going to capture the imagination of retro geeks the same way a small landscape 15C with it's retro minimalistic 7 segment display would. You can bet your bottom dollar the 15C will make every geek blog on the planet, and that turns into serious sales if the price point is low enough. A new higher priced 42S would barely make a whimper outside of the HP fanboys.

And given that neither would find a niche educational market where you get big numbers, you have to market it based on other factors as I've mentioned. The 15C would be the most sensible choice in this respect.

Dave.


#88

Quote:
It anything it would likely be based on the new 20B platform.

No it wouldn't. Not enough RAM on the CPU to emulate the 42s without adding an extra chip.


- Pauli

#89

Quote:
I think you are missing the point.

Either calculator will sell based mostly on it's retro appeal and perceived coolness, it's got little to do with its functionality.

I'm not the one missing the point. The point is to sell calculators. The advantage that results from functionality is demonstrable. Any supposed advantage from "retro appeal" and "coolness" is in the realm of religion.

Calculators do not sell in any great numbers based upon any kind of retro appeal. I'd be curious to know what anchors that in your mind as the principal product characteristic that would dominate and dictate sales. If retro appeal is all that can be achieved in sales by some re-issue, I hope HP has sense enough to abort immediately.

Calculators sell in great numbers to a general audience based principally on price, functionality, and ease of use. The HP-15C fails the HP42S in the last two of those. I doubt that many of the people who for almost 15 years have paid high auction prices for the HP42S did so because of any retro or even any collector appeal. I suspect they do so because the HP42S is (somewhat unfortunately) **STILL** 2009 State-Of-The-Art in RPN calculator functionality. No other RPN calculator has been produced that can take its honored spot. Had there been such a later machine, it would currently be the most sought after, not the HP42S or the HP-15C. For functionality. Not for retro appeal.

Quote:
It is quite possible (most likely IMO) that a retro 15C as-is would outsell the pants off a new super-duper 42S for a few reasons.

That is MOST unlikely. The exact opposite will happen if an HP-15C is re-issued.

Quote:
The first of which is the NRE which is extremely low for the 15C. The 42S would be a complete redesign from the group up using a new platform.

It's my understanding that current HP-12C models are based on completely different processor hardware and firmware than were the units of the same vintage as the HP-15C. The original HP-15C firmware translation to the new hardware would be almost as much an effort, for very little return, as it would be to translate HP42S firmware to the new hardware, with potential for much greater return for the effort.

Quote:
And a 42S would look like regular modern calcs, so it's not going to capture the imagination of retro geeks the same way a small landscape 15C with it's retro minimalistic 7 segment display would. You can bet your bottom dollar the 15C will make every geek blog on the planet, and that turns into serious sales if the price point is low enough. A new higher priced 42S would barely make a whimper outside of the HP fanboys.

I don't understand this obsession with some mythological and purely hypothetical mass retro geek appeal. It doubtless does NOT exist in enough quantity to make it worth even a week's development costs.

OTOH, the HP42S design appeals to those seeking the highest development of an RPN calculator for actual day-to-day use. I've used mine in work capacity almost daily for 12 years, and it's held up far better than the HP-15C that I used for eleven years before that. I suspect most HP42S users do likewise. I suspect these people far far outnumber any who would purchase a cryptic cyber-cipher like the HP-15C, were they given the choice.

Quote:
And given that neither would find a niche educational market where you get big numbers, you have to market it based on other factors as I've mentioned.

Other factors yes, but not the *ONE* upon which you hang *ALL* your hopes. More reasonable would be the really radical approach of producing an instrument that provides a significant advance in capability over the proposed alternative, one that equals or advances what has been available, one that will see actual daily use in a work environment, by workers (not students, not retro geeks).


Edited: 22 Oct 2009, 8:34 p.m.


#90

It's true that the current 12C (but not the 12c Platinum) is running on an ARM processor. However, it is running almost exactly the same firmware as the original 12C from 1981. The same could be done for the 15C, and in fact a prototype of that has been demonstrated at a previous HHC conference.


#91

Quote:
It's true that the current 12C (but not the 12c Platinum) is running on an ARM processor. However, it is running almost exactly the same firmware as the original 12C from 1981.

Is there some sort of emulator of the original hardware between the almost original firmware and the ARM processor? If so, wouldn't something similar be possible for a "new" HP42S-type of machine?

I'm frankly tired of HP taking the half-assed approach to scientific RPN calculators since the HP42S was discontinued 14 years ago. Everything in RPN-land since then has been extremely disappointing: The HP32SII, the HP33S, the HP35S...all these are fundamentally and substantially inferior to the functionality of the 1988 HP42S. The HP49G+ and HP50G are highly capable, but much larger and RPL-based. If HP continues down the path of mediocrity and issues its next RPN scientific calculator without features that one of their products carried more than 21 years ago, it would be evidence of lack of capability and commitment to produce useful and innovative products that were routine two decades ago. Very sad.

If HP-China burps out a pointless and unnecessary HP-15C clone, expect poor sales after the initial load of HP-15C freaks buy one. I won't be a customer. "Landscape" calculators are testiments to eccentric style over logical functionality.


#92

Quote:
Is there some sort of emulator of the original hardware between the almost original firmware and the ARM processor? If so, wouldn't something similar be possible for a "new" HP42S-type of machine?

I'm sure Eric or others will correct me if any of this in inaccurate:

Yes, HP bought rights to use Eric's Nonpareil simulator/emulator software and ported it to their new ARM platform they use on the current production model 12C.

HP also reaffirmed(?) their right to the original 10 series (11C/12C/15C/16C) calc ROMs, that is why Eric is now not allowed to supply these ROM's with with Nonpareil software like he used to:

http://nonpareil.brouhaha.com/

Yes, in theory you can do the same with the 42S, but it requires more memory and screen resources than HP's current 20B/12C ARM hardware platform can provide. So that means HP have to develop not only the suitable hardware base, but also the emulator/driver base as well. This is a lot of work, good luck selling that internally when they have just spent their time and effort on the current ARM platform.

And then of course with ROM simulation you don't get any of the fanboy desired 42SII extras. Exact 42S functionality, take it or leave it.

Dave.

Edited: 23 Oct 2009, 2:14 a.m.


#93

There is no 3rd party code in the 12c emulator. Eric was very helpful, but it is all cyrille's code.

As for the ROMs, it is much simpler than that. There is no copyright notice built into the ROM at all.

TW


#94

Quote:
As for the ROMs, it is much simpler than that. There is no copyright notice built into the ROM at all.

So are you saying that if I have the ROM dumps I can do whatever I like with them? Redistribute them? Create my own 12C and complete with HP?

You say built into. Given how sparse the space is I'd wager that there would be no space for a copyright and that copyright is implied. Or is it? Can HP give permission to someone to post the ROMs for public use like the 48GX and 48SX ROMs were published?

This is a very gray area.

Thanks.

#95

As Tim said, the new 12C is NOT running my Nonpareil simulation code. I offered to license my code to HP under non-GPL terms, but they were not willing to do that unless I also ceased to make it available to others under the GPL.

Cyrille has done an excellent job on the simulation code he developed for the 12C (and which I assume is also used in HP's iPhone 12C app, etc.). His simulation code is much more focused on achieving maximum performance then mine was, although I've done some work on speeding up Nonpareil and shared some high-performance BCD arithmetic C code with him. Cyrille hand-translated that code to very highly optimized ARM assembly language, and posted that here in the forum some time back.

#96

You nailed it, Mike, that was my point too:

Quote:
If HP-China burps out a pointless and unnecessary HP-15C clone, expect poor sales after the initial load of HP-15C freaks buy one. I won't be a customer. "Landscape" calculators are testiments to eccentric style over logical functionality.

#97

Quote:
You nailed it, Mike, that was my point too:
Quote:
If HP-China burps out a pointless and unnecessary HP-15C clone, expect poor sales after the initial load of HP-15C freaks buy one. I won't be a customer. "Landscape" calculators are testiments to eccentric style over logical functionality.


But surely a re-issue of the 15C is better than HP bringing out nothing at all, or another fluoro colour $10 rebaged toy instead...?

The 15C was just as popular as the 42S in it's own way, so why any HP fan would boo-hoo over it's reintroduction is beyond me.

One could also argue that a re-issue of an all-mighty super-powerful programmable 42S you can't save or download programs into would be equally as silly...

Dave.

#98

Quote:
I'm not the one missing the point. The point is to sell calculators. The advantage that results from functionality is demonstrable. Any supposed advantage from "retro appeal" and "coolness" is in the realm of religion.

When comparing two products of the same type, yes. The 15C and 42S have always will be fairly different beasts.

Quote:
Calculators do not sell in any great numbers based upon any kind of retro appeal. I'd be curious to know what anchors that in your mind as the principal product characteristic that would dominate and dictate sales.

I'm not saying it's the principle appeal, but it's an additional appeal that will not be lost on modern forms of social marketing such as tech blogs.

Quote:
Calculators sell in great numbers to a general audience based principally on price, functionality, and ease of use. The HP-15C fails the HP42S in the last two of those. I doubt that many of the people who for almost 15 years have paid high auction prices for the HP42S did so because of any retro or even any collector appeal. I suspect they do so because the HP42S is (somewhat unfortunately) **STILL** 2009 State-Of-The-Art in RPN calculator functionality. No other RPN calculator has been produced that can take its honored spot. Had there been such a later machine, it would currently be the most sought after, not the HP42S or the HP-15C. For functionality. Not for retro appeal.

It's silly to compare the 15C and 42S in functionality, it's chalk and cheese.

Quote:
It's my understanding that current HP-12C models are based on completely different processor hardware and firmware than were the units of the same vintage as the HP-15C. The original HP-15C firmware translation to the new hardware would be almost as much an effort, for very little return, as it would be to translate HP42S firmware to the new hardware, with potential for much greater return for the effort.

Not so. The new 12C uses the original 12C ROM via an emulator, so all the real hard work has been done, HP recalled rights to the original 15C ROM for just this purpose. Just drop the 15C ROM into the current 12C emulator and you should be good to go.

Not so with the 42S where no current hardware/firmware platform exists to emulate that I am aware of.

As someone else said, a new 15C has already been demonstrated, so a 42S is nothing but pie in the sky. If HP are going to re-release either the 15C or 42S, smart money is on the 15C, by far.

Quote:
I don't understand this obsession with some mythological and purely hypothetical mass retro geek appeal. It doubtless does NOT exist in enough quantity to make it worth even a week's development costs.

Yet HP chose to push the retro angle when it marketed the 35S, and has also continued a retro angle with the 12C.

Quote:
OTOH, the HP42S design appeals to those seeking the highest development of an RPN calculator for actual day-to-day use.

Which is an audience bordering on almost ZERO in terms of the current market!

Do you really think Joe-Public gives a toss about having the most powerful RPN calculator ever made? Hint, it's down about the same level as your perceived level of retro appeal.

The 15C at least has a novel form factor compared to other scientific calcs currently on the market, which *might* give it an edge in marketability over the 42S in today's market. Especially when it's going to come in at a much lower price point.

The 42S has neither, all it has going for it is its (hidden) power.

But once again, we are not comparing apples to apples here, so the point is bit moot.

I'm not in marketing, but if I was I think I'd prefer the low price point limited functionality retro form factor 15C than the ordinary looking but powerful 42S at a higher price point.

Any marketing people here who want to share their views?

Quote:
I've used mine in work capacity almost daily for 12 years, and it's held up far better than the HP-15C that I used for eleven years before that. I suspect most HP42S users do likewise. I suspect these people far far outnumber any who would purchase a cryptic cyber-cipher like the HP-15C, were they given the choice.

The point is not about HP fanboys and what they want/like, it's about the general market and what will appeal to them more.

I wager the 15C will generate more market excitement than the 42S, at much lower development cost and risk to HP. Will that translate into good sales?, only time will tell.

Quote:
Other factors yes, but not the *ONE* upon which you hang *ALL* your hopes. More reasonable would be the really radical approach of producing an instrument that provides a significant advance in capability over the proposed alternative, one that equals or advances what has been available, one that will see actual daily use in a work environment, by workers (not students, not retro geeks).

It's all about what HP can actually bring to market with their limited resources.

A few almost certain facts are in order:

- A new 15C has already been demonstrated

- A new 15C requires comparatively little NRE

- A new 15C can leverage off existing 12C production to lower the cost (the hardware is identical to the current production 12C)

- A 15C will be easier to sell internally to management

- A new 42S requires much more development and NRE

- No current platform exists to emulate the 42S, it must be developed(?)
- A new 42S will be at a more expensive price point

- A new 42S will be tougher to sell internally to management.

Based on that the odds of HP bringing a new 42S to market are pretty darn low. Sorry to all you 42S fanboys!

Dave.


#99

... but never give up hope!

Ceterum censeo: HP, launch a 43S!

Walter


Quote:
... but never give up hope!

Of course, the odds of anything are never zero, quantum theory and all that!

Dave.


Quote:
Of course, the odds of anything are never zero, quantum theory and all that!
Well, I don't really want to wait 1E9 years ...

PART 1: Here are just a few other items.

Ease of Use - The HP42S is fully alphanumeric. It is far easier to program, edit, debug, and produce meaningful prompts and displays than it is on the HP-15C. A program step looks like a program step..."LBL B" instead of "42,21,12".

Accuracy/Precision of Results - The HP42S firmware/Saturn combination produces far better results than does the old HP-35 type of processor/firmware in the HP-15C. (I know the HP-15C is closer to the HP-34C, but still all pre-Saturn calculators harken back to 1971.)

I/O - Though limited, the HP42S swamps the HP-15C. It has IR printer output, and audible tone output. I find both to be great assets. (I only wish it had a clock with time/date function.)

Safety of Programs and Data - I've had a 370-step Runge-Kutta routine for solving systems of differential equations in my HP42S since I entered it in 1997 from a HP-41CX listing. I have carried and used the HP42S almost daily at a nuclear generating station where it must pass through high-performance security x-ray screening. In the entire time I've used the HP42S I have suffered no losses of any sort, no data corruption even with very routine x-ray exposure (not to mention the several battery changes and other routine threats like drops and other impacts).

PART 2: If HP re-issues an HP-15C clone would HP also re-issue the HP-15C manual set, spiral bound like the original? IMHO, the one area in which the HP-15C indisputably *still* sets the world standard is the quality of its basic user manual AND advanced user manual. It would be a shame to re-issue the calculator, then fail to provide its greatest asset.

PART 3: The following is for the next generation small pocket-sized successor, and not part of the HP-15C vs. HP42S re-issue argument.

a. Incorporate ALL the features now existing on thr HP42S, coded natively on the ARM processor for optimal run-time results (no emulation of firmware coding that was written decades ago for the Saturn processor). (I'd also like to see that on the successor to the HP50g, if there ever is one.)

b. Incorporate the features of the HP17Bii.

c. Add bigger, clearer display.

d. (Only if battery resources make it practical) Add micro-SD card I/O...maybe USB support could power the SD card write process when connected to a PC.

Edited: 24 Oct 2009, 9:13 a.m.


I don't disagree with any or your points.

But all that doesn't mean a new 42S would:

1) Sell as well as a 15C which would be at a lower price point

2) Be easy or cheap to bring to market

It's all about market/business/internal politics practicality, risk, development cost, and what HP can and are likely to actually do.

I'm sure we'd all like to see both released, plus Walters 43S etc, plus my much dreamed after true pocket calc or scientific calc watch etc. Out of all those, the 15C is the only likely winner.

Dave.

I for one would buy a HP-42S+ for approximately twice what I'd pay for a HP-15C+.

Of course, I'm probably in the fan-boy category.

I like the more flexible two line display with alpha capability and HP-41C compatibility of the HP-42S.

And my fond wish for either would be more memory, speed, and I/O possibilities.

All for a reasonable price, of course.


Quote:
And my fond wish for either would be more memory, speed, and I/O possibilities.

If the 42S ever comes out again and they decide to do a ROM simulator like the 12C/15C, that guarantees no extra features, just speed.

Dave.


Hallo,

I have some notes, and I wonder or it is useful for HP to write it down.
This are no complains, just compare notes.

In 1987 i bought the HP27s, and in 1989 i bought the HP42s.
That¡¦s 20 years ago, and for 20 years i am a unpaid salesman for HP calculators.
But I have never seen a follow up for both calculators.
I wonder, if you don¡¦t have a follow up, why don¡¦t you keep them in production?
(too expensive?).

I bought the HP35s this year, in theory must be better than HP27S.
I also know that I have to compare with today¡¦s calculator the finical calculator HP17II.
But as far I can remember, the HP17 in that time, and probably still, is not so powerful as the HP27s.

For example, comparing HP27S with HP35s :

„« if you want to solve a formula you have to tell first witch variable you want to solve.
Instead the menu bar, than you can solve all variable, and in any direction.
„« if you write a program you can only store it as LBL A or LBL B not a real functional name. When you have 30kb memory, you must have a good memory by your self.
„« You have to write all programs down each other. Not separate. If you press in the solving mode, (a equation) one time to much on R/S he start the next program.

Comparing HP42S with HP35s.

„« Much more functions then HP35s
„« Better program labels, you can assign a word to each program.
„« You can programming every lcd dot.
„« Store and recall a alpha string
„« Smaller and thinner than HP35s
„« HP42S CLST? 3 operations, max 4 to clear somthing. HP35 CLST? 6 operation. max 7x button downs for clearing something.

Strong points HP35s.

„« RPN and algebraic
„« Beautiful design and beautiful button shape.
„« Fractions.
„« Enough memory.
„« Making a HP35s program is much easer, than programming the HP27S.
„« Lucky HP, that they made the HP35s so beautiful ;-)

Of cause you can say why don¡¦t take the HP50g? I like the HP50g for the office, but it¡¦s to big to ware in you suite. And for students,.. the teachers don¡¦t accept a graphic calculator. In the time I went to university, there was no teacher that recognised my calculator as an programmable calculator..and still nobody recognised it today. That¡¦s a good, thing, everybody thinks that you can calculate very quick. And that¡¦s nice isn¡¦t is? Thank You!

Kind regards,

Nico Tak

Holland

The lucky owner of: HP27s HP35s HP42s HP48sx HP48gx HP50g„» (Enter button wrong place, isn¡¦t? ;-))


Quote:
I wonder, if you don't have a follow up, why don't you keep them in production?
Only reason: Because they don't sell anymore.

Quote:
Only reason: Because they don't sell anymore.
Once it's in production, I expect that the only reason they wouldn't be able to afford to keep it in production is if the production space is needed for a more-profitable model. If they don't have that, it's probably foolish to quit making the model they've already been making. HP has made some very foolish decisions though.

Sometimes, it's not just the space needed for a more profitable model. It's shareholder value business: What? The profit is below 10% while with product XYZ, produced far far away, we can get much more profit! Just lay off the much too expensive workers an ditch the whole division!

Quote:
Once it's in production, I expect that the only reason they wouldn't be able to afford to keep it in production is if the production space is needed for a more-profitable model.

Even if the initial costs have been recouped, there are still ongoing manufacturing, marketing, distribution, support, personnel, etc... costs that require a minimum number of units sold to justify continuation.

Quote:
Even if the initial costs have been recouped, there are still ongoing manufacturing, marketing, distribution, support, personnel, etc... costs that require a minimum number of units sold to justify continuation.
You guys act as if there were no such thing as a niche market, or any viable companies in niche markets. If the market it tiny, those post-development costs are too. I myself have been in a niche-market electronics company since 1992 that has never grossed more than a few hundred thousand dollars a year and yet we've done most of our electronic assembly by hand in the U.S., our most complex product having nearly 1500 parts, which makes it far more complex productionwise than a calculator. If you look in the tiny ads in the back of hobby electronics magazines you'll find advertisements for products made by individuals in their garages, sometimes even moonlightling. One person or family does everything-- design, purchasing, assembling, advertising, shipping, technical support, sweeping, the works. A one-man company can still get a custom LCD made, custom keyboard, etc.. If I personally had what it takes to run a business, I'd like to make the calculator.

Quote:
You guys act as if there were no such thing as a niche market, or any viable companies in niche markets.

That's completely different. I thought we were talking about HP.

The setup for a manufacturing capacity of 1,000/year is different than 1,000,000/year. If you initially setup for 1,000,000/year and then reduced that to 1000/year with the same facility you would most likely be losing money.

Its moot anyway since HP outsources all consumer electronics. Like any other large consumer electronics company you send out a spec with an estimated quantity and then you get back a price. Smaller quantity, higher/unit price.

As for niche calculator market some recent examples of this would be the PIL-Box, NoVRAM/Clonix, uWatch, and MLDL2000. I paid 100 Euros for my PIL-Box that contains a single digit Euros of parts. But I am sure the cost just get started including custom circuit boards makes up the rest of the cost and I expect the volumes of all four to never get to a point where the cost will drop. And that's OK. I am fine with $100 Euros. Now I imagine that a calculator would be significantly more complicated with more upfront cost. However the volumes should also increase, but if we are talking about bringing back the 42S or something like it, I fear the volumes will be just the people on message board. Furthermore the price would have to be less than a used 42S on eBay.

Now I know that this changes everyday. Its possible to print your own circuit boards, there is an abundance of cheap LCDs, and even the cost to use automated 3D mills is getting cheaper. So who knows perhaps it is possible today to pull it off. But I'd defer to DaveJ and Eric Smith for the final word on that since they have actually done it.

Quote:
If I personally had what it takes to run a business, I'd like to make the calculator.

I'll be your first customer. :-)

Ok guys, I know you need a market to sell these things.
But there is a market for a more complex calculator the HP50g too..
1000 pages user’s guide… (HP27s 256 pages. HP42S 350 pages)

But if you launce a new calculator like the HP35s..
- And there is no menu bar down the screen, a thing they developed 20 years ago…it sounds to me like Back to the future.
- If you like to solve V= HxRxBxTxP and your calculator ask: What x to solve?
Uh sorry,.. I forgot the names of the variables..

If you use a HP17II, HP27S or HP42S you can see the variables |V|H|R|B|T|P| down the menu bar. And enter them any direction, and solve them any direction. Much easer I guess.

A HP35S should be better than a HP27S isn’t it?
One of HP strongest points is the Solve menu bar.
HP invent it, use it!!

- HP35S program name you can only mark it with LBL A,B,C,D..Z.
Do you think you can remember the program LBL K after three weeks?

Uak,.. I am boring,.. think positive… ebay.
Kind regards,

Nico Tak


Nico,

the 35s has a feature the 50g misses: Ease of use. Despite its memory, it is not intended to keep large programs or sets of equations. Instead, it hepls you to automate some calculations when you need it. And you don't have to mess with an alpha key, typing long variable names. The basic concepts of the 32S and all its successors are just brilliant, there's nothing to improve.

Thomas


Hi Thomas,

Your right there is nothing to improve, I forgot the word “did”, HP did invent it, use it.
I took up the 35S to see or I am wrong. But now two days later I still not convinced, that they make it very much easer. If it’s developed for easy use, why do you give the calculator so much memory? Use the memory for the menu bar,in the solve menu. Isn’t it? Programming is ok. Did you ever used a HP27S of HP42S? Because you must have the same problems, with the HP35S.
And if you don’t have the same problems, you may be solved these problems. Please tell me your solutions..

Kind regards,

Nico


Hi Nico,

you're right: one of the most annoying features of the 35S (and the 20b, BTW) is the outdated display and the missing row of soft keys. Of recent calculator models, we have this row in the 17bII+ Silver, however, so HP seems to know it still ;) OTOH, the display of this calc has the same visible specs as the old 27S and 42S, so no progress in 20 years.

With a state-of-the-art display, a lot of opportunities open.

Ceterum censeo: HP, launch a 43S.

Walter

Quote:
If it’s developed for easy use, why do you give the calculator so much memory?
To never run out of it :-).

Quote:
Its possible to print your own circuit boards, there is an abundance of cheap LCDs, and even the cost to use automated 3D mills is getting cheaper. So who knows perhaps it is possible today to pull it off. But I'd defer to DaveJ and Eric Smith for the final word on that since they have actually done it.

It's easily possible to pull it off, you just need to:

a) want to do it

b) have the time and some money to do it

c) do it reasonably well enough so people want it

My uWatch is an interesting example.

I've now discontinued the original Mk1 watch, not because it wasn't selling, but because I've lost interest in it. I still have 300+ people on the email waiting list, so I could probably sell another 50 or 100 at least at the current $99 price, that's practically guaranteed. And the profit is rather large now that the NRE has been absorbed. So I could be continuing to make money, but I just couldn't be bothered any more, it's just too much hassle.

Perhaps I need to outsource the hassle to someone!

Dave.


Possibly Related Threads...
Thread Author Replies Views Last Post
  Prime Version Check Thomas Chrapkiewicz 1 369 10-21-2013, 10:12 AM
Last Post: Han
  How to check / repair a HP 82106A module for HP41C calculator ? Eduardo Mingo 13 1,246 01-07-2013, 11:05 AM
Last Post: Diego Diaz
  Check on mathematical identity Namir 21 1,637 08-26-2012, 11:21 AM
Last Post: Les Koller
  [WP 34S] Final documentation check Walter B 18 1,557 07-06-2012, 02:10 AM
Last Post: Walter B
  The most abused phrase "Check Back Soon" Michael de Estrada 8 866 10-11-2011, 10:25 PM
Last Post: hpnut
  Auctions Vintage HP-15C Jim Johnson 5 651 09-13-2011, 08:46 PM
Last Post: bill platt
  question about calculator auctions Gonzalo Fernandez 1 335 06-03-2011, 07:08 PM
Last Post: Massimo Gnerucci (Italy)
  Do the Ebay sellers check the merchandise? Gonzalo Fernandez 35 2,680 02-05-2011, 01:21 PM
Last Post: Michael de Estrada
  41C Memory Module 82106A Check / Test AlbertTocarLoseggs 6 705 10-11-2010, 02:41 PM
Last Post: Diego Diaz
  WAS: HP vs. TI on eBay auctions Joerg Woerner 21 1,808 05-02-2010, 02:23 AM
Last Post: db (martinez, ca.)

Forum Jump: