▼
Posts: 764
Threads: 118
Joined: Aug 2007
Having just got it from HP on 3/16:
* This is a nice, lightweight calculator. I so far do not have any keyboard issues. The keys feel better on the 30b than on the 20b.
* The calculator comes with a cover pouch, which I have to say it is better than the pouch for the 17bII+. I cannot accidently turn the 30b on while it is in its pouch. Plus the 30b contains declining depreciation w/crossover, inverse and quadratic regressions, and trigonmetric functions.
* Programming is similar to the HP 12c, with every number being literally being a step. For example, entering 12000 in a program takes five steps. The comparison mechanisms are nice too (test operators work similar to HP50g).
* While there are only 290 steps, one could fit some nice programs, after all most HP 12c owners have been working with only 99 steps since 1982.
* Like the 20b, the 30b is super fast.
* Ther learning manuals on HP's web site are well written. I only wish that they put more concrete examples in the getting started guide.
* The truly annoying thing about programming in the 30b is that if you have to access something in the Math menu, you will use a lot of steps to get there.
Sample Program: Blueprint for the Sum Function - RPN Mode
M2
Sigma f(M0)
M0 = M1
store values in M1 (beginning) and M2 (end) before excuting the program (4 in this example)
0
STO 3
EEX
3
+/-
RCLx 2
RCL+ 1
STO 4
Lbl 41 (hold Shift+%)
RCL 4
Math
Up (hold Shift+Up arrow)
Up
Input (Math, Up, Up, Input gives IP [integer part])
STO 0
CALL 42 (like XEQ, hold Shift+7)
STO+ 3
ISG 4 (for ISG, hold Shift+ +/-)
GOTO 41 (hold Shift+CshFl)
RCL 3 (Sum)
Stop (hold Shift+RCL)
Lbl 42
[input f(MO) here]
RTN (hold Shift+8)
Eddie
Edited: 18 Mar 2010, 12:02 a.m.
▼
Posts: 1,278
Threads: 44
Joined: Jul 2007
Have you had a chance to try the solver yet?
TW
▼
Posts: 764
Threads: 118
Joined: Aug 2007
Yes. The solver is like the Sovler on the 15c. The equation is entered as a program. Then Sovler is activated in calculator mode.
▼
Posts: 72
Threads: 25
Joined: Nov 2007
The one thing I really don't like about the 20b and 30b is the number of steps it takes to execute some simple functions thru the MENU. For example a simple % change in the 12C (and most hp calcs) it takes 4 steps: type number, <ENTER>, second number, <%change> and you have the answer.
To do the same operation with the 20b and 30b it takes 5 key press to get to the <% change function> and then 8 additional steps to do the actual % change calculation ==> 5 + 8 = 13 steps. It is a very inneficient way to perfom this operation.
What used to be done in 4 stpes, it takes 13 steps now.... I know the benefit of the menus is to add more function capabilities, but a second color function (let's say an orange)would have provided more of the common functions without having to go through the complicated menus.
▼
Posts: 1,278
Threads: 44
Joined: Jul 2007
And on the 17bII? (just wondering as I am not familiar at all with those units)
Unfortunately, the 20/30 do so much you couldn't fit it on the keyboard even if you tried. You'd end up with a lot of keys that would activate menus, but you'd still have menu operations.
TW
Edited: 18 Mar 2010, 9:00 p.m.
▼
Posts: 42
Threads: 4
Joined: Sep 2008
It is isolated as an example, but the 17Bii uses a menu (under Business) for % change. Once you go to the menu, you enter the before number, then the after number, and click on the soft menu for % change). Basically, it functions as though you had coded up a solver equation.
--Sancerre
▼
Posts: 1,545
Threads: 168
Joined: Jul 2005
That's part of the design approach chosen. Has pros and cons.
Good news is that you can easily write a program that does the functions you use often and assign them to the keyboard so you don't spend time finding them in menus.
My 30b has several small programs loaded for the inverse trig functions. These are "assigned" to the shift-hold locations of the regular trig functions. Very nice.
I've also assigned the normal distribution and inverse normal distribution functions in a similar way.
If you keep in mind that the programming approach is **really** much more a macro ability than a programming environment, then the keystrokes in a program required to access functions in menus makes sense.
That said, some VERY good programs, some very good FAST programs can be written on this little machine.
And, financial functions never before found on a business calculator too, such as FMRR.
I think this may grow on you. I for one don't pick up my 12c nearly as often as I used to...and I've used that 12c for 25 years.
Posts: 528
Threads: 40
Joined: Dec 2008
Overlays overlays overlays. I really believe that this problem and all sorts of others could be avoided by providing a user-assignable keyboard and keyboard overlays. 99% of the "my favorite function is buried in the menus" complaints would go away.
I know that some (most? all??) of the calculators allow user-assignable keys already, but with no convenient way to tell what function has been assigned to a key, this extremely useful feature is effectively crippled, IMHO.
Dave
▼
Posts: 1,392
Threads: 142
Joined: Jun 2007
It would have been nice if the 30b package contained an overlay that you could write on. The overlay that is included is very good, but it shows only the programming commands, which is necessary for program development, obviously, but not useful to the end-user. Perhaps someone will offer for sale a good quality overlay (like the enclosed one) that you can write on (or computer print on).
Posts: 1,619
Threads: 147
Joined: May 2006
Quote:
but with no convenient way to tell what function has been assigned to a key, this extremely useful feature is effectively crippled, IMHO.
On the 41 you press and hold. The function will be displayed, and if you keep holding 'NULL' will be printed indicating no action taken. I think its a perfect solution and it works with user defined keys as well, i.e. it prints what you defined, not the base function of the key.
Posts: 553
Threads: 57
Joined: Sep 2006
On the 30b, you can achieve this by creating up to 10 user-defined keys. Just create a program with the functionality you want (even as simple as just one keystroke) and then assign it to another key on the keyboard. Shifted or unshifted.
In the program catalog, you can also see which key it is assigned to, or review all user-assigned keys on the calculator.
thanks,
bruce
▼
Posts: 1,392
Threads: 142
Joined: Jun 2007
Neither pressing and holding the key (a la 41) nor bringing up the program catalog, as Bruce suggested, is the ideal solution to this problem; both require work a user shouldn't have to do to run a custom program assigned to a key. Anyone who wants to use the 30b as a platform for users (or anyone, really) to run custom-coded solutions really needs a custom overlay.
I believe that HP should step up here and offer some such service. The 30b will never reach its fullest potential without this capability.
Of course, the "best" solution would be programmable keytops!
▼
Posts: 528
Threads: 40
Joined: Dec 2008
I agree with Don that keys with programmable key top displays would be ideal, but I don't think the technology exists today, and if it does I have to believe that it isn't cost effective.
I agree with Egan that the 41's solution was wonderful and it's practical. Holding the key down showed you what function it performed. In addition, the calculator came with overlays that had the normal functions printed on them, along with a set of small stickers that could be placed on the overlay to indicate new functions. Some of the stickers had existing functions pre-printed on them and others were blank, so if you reassign existing functions or assign keys to execute new programs. You put the stickers on an overlay where the reassigned keys are, slap it on on the calculator and all the keys would be labeled. Indeed, this is how my calculator went through life for about 20 years.
I don't know how hard it would be to display the function of a key on a 50g, but it might not be trivial. At the very least, it means that a key's function isn't executed when the key is pressed, but when it is released.
But overlays seem pretty easy to me. All it requires is a small lip around the top of the keyboard and a spring latch to hold the overlay in place. It looks like the keyboards for many calcs in the current lineup are the same shape, so one size would fit all, so to speak.
There's an alternative that would work well with the existing calculators if someone has time on their hands to do it. Craft stores sell programmable paper cutters. One could probably be programmed to cut out an overlay. Now if the paper is really a big peel-off label with post-it adhesive, you've got an overlay that you can firmly attach to the calc. Add a word-processor template that lets you print the key labels on the sheet before you feed it into the cutter and you have your overlays.
Dave
▼
Posts: 1,392
Threads: 142
Joined: Jun 2007
Quote: But overlays seem pretty easy to me. All it requires is a small lip around the top of the keyboard and a spring latch to hold the overlay in place.
Actually, on the 30b, you don't need any spring latch. The overlay for the programming commands, which is included with the calculator, is clear plastic with the Shift+hold programming functions listed in black type, and it just clings to the surface of the calculator very well. It is great because it does not peel off when you don't want it to, but you can remove it easily if you do want to do that.
I'd say the ideal solution is this exact type of overlay that can be printed on by your PC printer and it covers the entire keyboard (as opposed to only the top 4 rows). If HP would furnish that as an accessory, at a reasonable cost, that would make this calculator truly revolutionary.
▼
Posts: 764
Threads: 118
Joined: Aug 2007
The overlay would be nice. I think the overlay provided with the HP 30b is necessary for programming.
Posts: 74
Threads: 2
Joined: Jul 2005
Maybe this is a good time to ask a simple question about the 30b. Does the 30b have degree-minute-second math functions? On the 41/42, those would be HMS+, HMS-, HR, and HMS. I'm just trying to figure if I could reassign keys and make this a useful calculator for me.
▼
Posts: 1,392
Threads: 142
Joined: Jun 2007
Quote: Does the 30b have degree-minute-second math functions?
No.
Posts: 553
Threads: 57
Joined: Sep 2006
No, but you could write four little programs to do that, and then assign each of those to a key. Then you have what you want!
Thanks,
Bruce
▼
Posts: 74
Threads: 2
Joined: Jul 2005
I know. Just a matter of IP, addition, subtraction, and division and multiplication by 60. But, it seems that the only things of interest to me on the 30b calculator are the speed and trig functions. I have some backup 42's. It is just such a PITA if I have to re-enter those programs on the 42. I guess once every three or four years shouldn't be a concern. I have a a number of calcs at work, but when I really need to do something it is always the 42 that gets picked up. Contrary to popular belief, I think the display is perfect for what it does (annotations and numbers on two lines).
I'll wait until they come out with the 30s or 45s. None of the other recent calcs (since the 42) have been better for my needs.
thanks
▼
Posts: 2,448
Threads: 90
Joined: Jul 2005
They already used the 30S name and it sucked. Have to be the 31S :-P
▼
Posts: 3,229
Threads: 42
Joined: Jul 2006
Or 34S. Much better pedigree :)
- Pauli
Posts: 28
Threads: 6
Joined: Nov 2006
My observations on the 30B are that whilst it adds a lot of great functions over the 12C, the 12C is still preferable as a daily calculator, mainly because its extra shifted functions make it faster to use.
Going through menus for regularly used functions is a huge disadvantage when the speed of a calculation is critical, as it is when used in a financial markets environment.
Disadvantages of 30B over 12C:
- Lack of 12x and 12/ for time value functions, since these are often used.
- Having y^x and 1/x as shifted functions is a problem as these are often used in rate calculations.
- STO is a shifted function - again, this adds time to complex calculations.
▼
Posts: 764
Threads: 118
Joined: Aug 2007
All that is needed is to set 12 as P/Y, then SHIFT N acts as 12x. The interest key is for annual interest, not periodical interest.
Posts: 1,392
Threads: 142
Joined: Jun 2007
To be fair, the 12* and 12/ aren't needed on the 30b because you can set p/yr to whatever you want. And I would think that financial people would not use y^x and 1/x and STO often enough that the shift on the 30b for these would be a problem.
I know most users of the 12c don't write and run programs, but from a programming perspective there is no comparison between the 12c and 30b, the 30b wins hands-down.
I have both, and I like both. But if I'm going to write a program (which is all I generally do with these), I'll do it on the 30b.
▼
Posts: 28
Threads: 6
Joined: Nov 2006
Don,
y^x and 1/x are used a lot in doing interest rate conversions. Similarly, STO is also used a lot for storing several constants (in different registers) and also intermediate results. Also, setting 12 as the number of periods (if I understand correctly) is something that would have to be changed frequently as you alternate between annual compounding and monthly compounding calculations.
Eddie,
Cashflow calculations can be done for different periods, such as monthly when calculating swap rates and valuations. The annual convention is arbitrary, and irrelevant for some calculations, so I think in terms of compounding periods rather than years or months.
An example of the 12C 12/ function use is when an NPV is being calculated and the rate is quoted as an annual equivalent on a monthly compounding basis. Having the 12/ function makes it easy to throw in 12 g i (4 keystrokes) instead of 12 ENTER 12/.
Bruce,
I know the 30B has a faster processor, and I will try it at some point, but I doubt that it will be faster than the 12C at everyday financial market calculations.
Compare the way the 12C and 30B are laid out - the 12C is really laid out in a more organized way. I don't think that serious financial number crunchers were consulted in the design of the 30B. Having the y^x and 1/x keys near the Time Value of Money section is very useful.
As a related aside, I had a conversation with someone at HP a few years ago and mentioned that a limitation of the 12C was that it needed a greater number of digits in the display, and gave the example of calculating swap values, but they seemed unaware of this application (which leads me to believe they didn't consult professional users). Maybe it is more accurate to say that HP DID consult professional users in designing the 12C, but this part of HP's organizational memory has been lost.
In addition to the TVM, the y^x, 1/x, STO and RCL keys are probably the most often used keys in doing everyday rate, cashflow and bond related calculations (i.e. bonds, swaps, futures).
I don't know of any common finance applications that rely on trig functions - I would have thought that that should have been relegated to a menu, rather than on the main keyboard.
FWIW, I have used the 12C for 20+ years (after using a 41C), and have tried various other HP models, but none so far beats the 12C for financial markets usage.
Edited: 21 Mar 2010, 6:24 a.m.
▼
Posts: 1,392
Threads: 142
Joined: Jun 2007
Quote: I have used the 12C for 20+ years (after using a 41C), and have tried various other HP models, but none so far beats the 12C for financial markets usage.
Then I'd say you should stay with the 12c unless another calculator does something you need to do that the 12c does not do.
Posts: 553
Threads: 57
Joined: Sep 2006
Quote:
Compare the way the 12C and 30B are laid out - the 12C is really laid out in a more organized way. I don't think that serious financial number crunchers were consulted in the design of the 30B. Having the y^x and 1/x keys near the Time Value of Money section is very useful.
Tim or Katie can probably comment on the relative speed of calculations between the 30b and 12c, but I would also be interested in hearing that answer.
Remember, if you don't like the keyboard layout of the 30b, you can reassign up to 10 keys. This way you could move y^x, 1/x up where you want them, and you could even flip the behavior of STO and RCL (RCL being shifted, or put on another key, and STO being the default action of that key).
Each key has three possibilities: keypress, shift-keypress and shift-hold-keypress.
All that being said, if you're really that attached to the 12c, I would agree with Don that you probably won't like any other calculator for your daily use.
Thanks,
Bruce
Posts: 193
Threads: 10
Joined: Mar 2008
Hello,
Quote: y^x and 1/x are used a lot in doing interest rate conversions. Similarly, STO is also used a lot for storing several constants (in different registers) and also intermediate results.
But the 30b has an interest conversion function (IConv), so you don't real need ^ and inv.
I agree that STO is important, unfortunately, we ran out of keys. RCL was put as primary because you normally do multiple RCL per STO.
Quote: setting 12 as the number of periods is something that would change frequently as you alternate between annual compounding and monthly compounding calculations.
if you alternate between 1 period per year and 12 periods per year, yes, you will have to change it often. not that it is slow. 1 shift P/YR is only 3 keys...
Quote: Cashflow calculations can be done for different periods,
cash flow calculations are done in the CF menu and it has it's own p/yr setting, so this would not affect the p/yr used for TVM calculations.
Basicaly, on the 30 (and every other modern finance calculator), Interest rates are ALWAYS on a yearly basis with a given compounding.
Since these calculators do not overload keys by having them being used for various semi-related thing for various function, it means that you can have simultaneous TVM calculations with 12 p/yr and TVM calculations with p/yr=1 going on at the same time.
Quote: I doubt that it will be faster than the 12C at everyday financial market calculations.
If you are a power user of the 12c, you are probably right. and the main reason is that the 12c has 1 function per key, threfore resulting in faster user input. However, if you are NOT a power user and do not remember exactly which registers the bond values have to be placed on an what arguments are needed for the BOND price function and what the result means, then the 30b is much better.
Also, realize that the 30b has at least 2 to 3 times more functions than the 12c. So, a 30b with one function per key would require close to 80 additional keys compared with the 12c.
If you do cash flow calculations, you should realy have a look at the 30b as it makes working with cash flow a whole lot easier thanks to the nice CF editor.
Quote: I don't know of any common finance applications that rely on trig functions - I would have thought that that should have been relegated to a menu, rather than on the main keyboard.
This is mostly done because finance student more often than none have to take some math and need the trig functions. This has beed a magor hurtle for the 10b and therefore had to be added proeminently on the 20/30b.
regards, cyrille
Posts: 1,545
Threads: 168
Joined: Jul 2005
Hi bink. I am a longtime user of the 12c as well as a former university instructor who taught business math to freshmen and sophomores for nearly 20 years. Let me comment on some of your observations.
1) y^x and 1/x. The only real way these could have been better positioned for what you want is to have made them primary keys on the 20/30b. However, the primary keys are already used/taken for other functions / menus, such as the IRR / NPV menus and the cash flow environment, etc. Given that they were going to be shifted functions on account of this, the finger travel distance is essentially the same where they are located on the 20b/30b and where they would have been had they been located near the top of the calculator.
2) P/YR vs. periodic compounding. This is essentially a wash, as I have become convinced over a number of years. On the 20b/30b (and it must be said, the 10b2, TI BAII Plus, TI BAII Plus professional, the casio financials, the sharp financials and every other (?) financial calculator presently sold other than the 12c), you must type 12 shift P/YR or 1 shift P/YR or 4 shift P/YR when you want to change the payment frequency. This is **offset** by the requirement on the 12c to divide the nominal interest rate by the under of periods in order to enter a periodic interest rate. Having taught classrooms of students using BOTH types of machines for 20 years, there is no speed difference between the methods that I have ever been able to ascertain. It is a matter of taste and what you are accustomed to.
3) STO as a shifted function. While Tim and Cyrille are correct ... that you tend to store a number/value less often than you recall the value, I too would like to see a primary store key. In practicality, I don't see this as a make/break compromise.
4) Cash flow compounding periods. Given that the 20b/30b are MENU oriented calculators, this is affected by setting a #CF/yr value at the bottom of the cash flow environment. Set it once if you don't need to change it very often and you are good to go. If you need to change it all the time, this might be less convenient.
5) Speed. The new 12c+ is certainly about as fast as the 20b/30b at many financial calculations, certainly. The older 12c seems painfully slow now at some functions, however. Compared to the older 12c, the 30b can be 10-20 times faster. Calculations are simply instantaneous in almost all circumstances.
6) Interest rate conversions. These seem to be the primary reasons the location of the 1/x and y^x key placement seems to be very important to you. However, once you have used the interest rate conversion environment found in IConv, you will be slower at converting interest rates using 1/x and y^x.
7) Professional user involvement. Suffice it to say that there were professional users involved and feedback was very positive.
Bottom line (ha).
For someone used to the 12c, great... buy a brand new 12c+ that runs much faster and continue to enjoy the HP way.
However, the 20b/30b offers much that the 12c does not. You may have no use for these, but they are certainly wonderful to have available if you ever find a need for them.
Advantages of the 20b/30b over the 12c...
1) Customizable keyboard. Bring ANY function out from a menu and assign it to a primary, shifted, or shift-hold key position. Write a program to do any calculation you want, including 12 / and assign it to any key position...such as the shift I/YR position. Now you have the 12 / function from the 12c on the 30b. You can do this for 10 key positions if you wish.
2) Trigonometry, inverse trignometry, hyperbolic trigonometry and inverse hyperbolic trigonometry. Hey, not everyone needs these, but even a financial analyst might want to know how tall a telephone pole is given the length of its shadow and the angle to the sun.
3) Probability distributions. Normal distribution and inverse normal. Students t distribution and inverse. F Distribution and inverse. Chi-square distribution and inverse. And depending on the rom revision, the binomial distribution. No substitute for having these built in when you might need them.
4) Much more advanced statistics. Built-in population standard deviation of X and Y. No more computing the mean and doing Sigma+ to fake this. Covariance. Nice financial computation built-in. Standard error of the X and Y means. Very handy for hypothesis testing or confidence intervals. Built-in permutation and combination calculations. SEVEN regression models including Best fit. Just no comparison here. If you don't use statistics, these don't mean much to you. Sure.
5) Increased financial functionality. 30b has MIRR built-in. !2c requires a program or a move to Excel. 30b has FMRR (financial management rate of return) as an alternative to MIRR. For NPV calculations, there is also Net FV, Net Uniform Series, Payback, discounted payback, and a total cash flow summation. The 12c does not directly offer ANY of that.
6) Bond calculations. The 30b offers two bond duration calculations. 20b/30b directly compute yield to call as well as yield to maturity.
7) Breakeven. 20b/30b have this built in. 12c does not.
8) Black Scholes Option pricing. Built in. 12c requires a wonderful program written by Tony Hutchins.
9) Programmability. The 12c can be programmed. Sure. The 30b has more program space and is MUCH faster. It also allows for up to 10 separate programs. Has subroutine abilities, text messaging abilities, etc. 12c looks old by comparison.
So...
Don't get me wrong. I was raised on the 12c. I still THINK like the 12c works. But, the 30b is well worth a look by anyone who has used a 12c. The VERY few changes required to use the 30b (which works like all other financial calculators in existence today) are well rewarded by a host of improvements and things you just cannot do with a 12c. Part of me is pained to write that sentence, but I believe it to be true.
My 2 cents...and that's a distinct present value.
▼
Posts: 42
Threads: 4
Joined: Sep 2008
You didn't add the solver, which is a huge addition. The main reason I stayed with my 17BII throughout my MBA was the solver, which is much easier to quickly use than key stroke programming (and this is coming from a programmer).
--Sancerre
▼
Posts: 1,545
Threads: 168
Joined: Jul 2005
Agree about the solver.
I was thinking of it with regard to programming, since you write a program to evaluate the function so that it is equal to zero.
But, you are correct. The ability to have a program that evaluates:
A - B / C + D*E = 0
and to easily store values into A, B, C, D or E and solve for the unknown is very nice.
The learning module on using the HP Solve in RPN mode contains an entire solver program that handles loans with odd days first periods. 2 odd days, 34 odd days, etc. You can even SOLVE for the odd days if you know the other pieces.
Rather difficult to do on the 12c, certainly with only one program...and especially since the 30b program is simply written as the formula set to equal zero. And, again, I say things like that about my old friend the 12c with a twinge of pain.
Gene
▼
Posts: 850
Threads: 10
Joined: Mar 2009
Is there a manual available? I can't seem to find any on the HP site. I know there are modules, but in this electronically connected age, manuals are usually available online and I like to look through them before deciding to buy (or not).
The 30b seems like a nice almost-do-everything calculator (financial with a good bit of maths & stats), I'm just dissapointed by the lack of programming space. Probably because initial information said it would have 32k. But I guesss they wanted more reasons to place it below the 17bii+. However, a bit more e.g. 4k would have been nice.
▼
Posts: 1,392
Threads: 142
Joined: Jun 2007
Agree wholeheartedly about the need for more program space. I think the 290 byte limit has more to do with the characteristics of the ARM processor than a marketing consideration.
To those who have purchased the 30b, did it come with a manual on CD?
▼
Posts: 3,229
Threads: 42
Joined: Jul 2006
Quote: I think the 290 byte limit has more to do with the characteristics of the ARM processor than a marketing consideration.
The CPU has 2048 bytes of non-volatile RAM. The low level drivers don't use any of this (at least they don't on the 20b and I suspect they are identical in this respect).
So 1758 bytes are being used for other purposes (and probably a bit spare for bug fixing etc). I smell a lot of waste here :)
Pauli
▼
Posts: 850
Threads: 10
Joined: Mar 2009
Surely it can handle a bit of external RAM?
▼
Posts: 3,229
Threads: 42
Joined: Jul 2006
There is another 4096 bytes of volatile RAM which is powered down most of the time.
Yes, external storage could be added for additional cost.
I'd personally prefer some kind of SD or USB host capability. There is no reason user code couldn't be run from such.
- Pauli
▼
Posts: 850
Threads: 10
Joined: Mar 2009
Just to clarify, with "external" I meant "not part of the processor" but still on the PCB, not external to or removable from the calculator (although that would be nice). Maybe HP were afraid of the complaints again "what's the use use of having all that memory without ways to backup and restore it". They could see it coming: add memory, add I/O. Price, positioning (again I make my point - relative to the 17Bii+) all had their influence I suppose. (Like the 42s was a crippled upgrade to the 41, but I digress).
From what information I've gathered, it seems a very nice calculator for the price. (but I would have been happy with just that bit more memory*, even without I/O capability).
* But as any programmer will tell you "how much memory is enough? just that little bit more ;)
▼
Posts: 1,545
Threads: 168
Joined: Jul 2005
I don't know what is stored where, but there are a lot of financial variables that are kept by the continuous memory (think things like PV, N, etc and cash flows and statistics data and ...) all of which has to be stored too.
Posts: 1,278
Threads: 44
Joined: Jul 2007
A lot of waste? There isn't a spare byte of space left in there. 290 bytes is all that is left for user program space.
You have to store all of the various menu items, saved data, settings and so on. Remember the calculator is essentially turned off all the time. It has to be able to retain all of those items in saved memory or they disappear when you release the key.
There are approximately 200, 12 digit+3exponent numbers that need to be maintained, plus space for the lower and upper lines of text. . .
TW
Edited: 25 Mar 2010, 8:04 p.m.
▼
Posts: 3,229
Threads: 42
Joined: Jul 2006
Quote: A lot of waste? There isn't a spare byte of space left in there. 290 bytes is all that is left for user program space.
Just because there isn't a spare byte doesn't mean there isn't any waste :)
Quote: There are approximately 200, 12 digit+3exponent numbers that need to be maintained, plus space for the lower and upper lines of text. . .
Financial registers -- a complete waste ;-) For the sarcasm impaired both this and my original comment have a degree of jest present. Opps, that was sarcasm too: damn an infinite loop.
Given that the bulk of the non-volatile RAM is used for these numbers, how much space is left? 12 digits + 3 exponent digits + 2 sign bits can be packed into seven bytes without any difficulty: three digits per ten bits plus two sign bits is 52 bits leaving 4 unused bits. 200*7 = 1400, leaving 600 bytes for other stuff. System state and the lines of text (the lower line isn't really text but we will ignore that) will consume some of this but over half seems very high.
However, using 8 bytes per register and doing a dumb BCD encoding leaves about 400 bytes minus the 290 for programming and we're down to a size I can believe for the other overheads (although it still seems a bit on the high side). However this is wasting precious memory.
Memory is very short on this CPU and the CPU is very fast. Trading CPU cycles for memory is beneficial here: store the numbers space efficiently and convert them to/from an unpacked format for calculations.
Yeah, I know existing code, time to market, etc will stop this form being a possibility...
- Pauli
▼
Posts: 3,283
Threads: 104
Joined: Jul 2005
Another idea would be t have a curtain between the numbers and program space which moves automatically. So you have probably less room for cashflow data if you have stored a bigger program but who cares?
Having lot of UP/DOWN steps in a program isn't very memory efficient either. Why not code each function as such? That would make reading programs easier, too.
Spending a lot of CPU cycles for routine tasks my drain the batteries faster.
Edited: 26 Mar 2010, 4:19 a.m.
▼
Posts: 1,477
Threads: 71
Joined: Jan 2005
I agree, the 30b should have dynamic memory allocation. The 290 steps are not "fully merged", so sequences like "STO + 2" take up 3 bytes of RAM. After writing a good number of programs for the 30b I've concluded that the program memory is effectively not much more than the 99 steps of the 12C. OTOH, having an array of 100 bytes that's always available is awfully nice even if the programs that use that array tend to be quite long.
-Katie
Posts: 528
Threads: 40
Joined: Dec 2008
Quote:
A lot of waste? There isn't a spare byte of space left in there. 290 bytes is all that is left for user program space.
Although I completely understand the hardware and cost limitations here, it's still very unfortunate that in 2010, when memory costs something like 10 one-millions of a penny per byte, that a calculator has only 290 bytes of memory available for programming. I'd actually view this as a marketing snafu. A consumer who reads the specs carefully is likely to think "290 BYTES of memory? Didn't they have more than that on the HP 41C 30 years ago?" A consumer would wonder why HP didn't raise the cost by a penny and put in a hundred K of memory. Maybe another processor with more embedded NVRAM will come available.
Still, I'm very encouraged by the 30b, mostly because it seems to imply a push towards native ARM code!!
Posts: 1,545
Threads: 168
Joined: Jul 2005
Think of the 30b this way...
1) A nice, inexpensive "do everything" calculator. Well, almost everything.
2) Programmable enough for one or two big programs (prime factor, base conversion and lunar lander all fit at once) or use the program space to reassign up to 10 keyboard locations with frequently used functions.
3) Programming is more like a macro recording language than we've seen before. Keep that in mind and you won't get all worked up over UP and DOWN showing as program steps to access menu items.
4) An advanced financial model containing functions never before found on any financial calculator.
5) An advanced statistics calculator too.
I will be curious what the community at large think of it. I know Don and I like it. :-)
▼
Posts: 850
Threads: 10
Joined: Mar 2009
I have received a copy of the manual. It certainly has some handy functions, including ones that the 35s hasn't got (apart from financial that is). I will try and plan some programs and see what I can can fit in once I get one (if and when it finally lands in the UK).
▼
Posts: 1,545
Threads: 168
Joined: Jul 2005
The 30b learning modules have quite a few programming examples in them too.
Examples
Example 1: Calculating digits of PI
Example 2: Finding prime factors of an integer
Example 3: Base conversions for bases 2-10
Example 4: Lunar lander game from the 1975 HP 25 Applications Program book
Example 5: Guess the secret number game
and here:
HP Solve examples
Example 5: Generalized odd-days loan solver program
All the 30b learning modules can be found here:
30b learning modules
Posts: 553
Threads: 57
Joined: Sep 2006
I too would prefer more programming space, but -- honestly -- once you start playing around with it, you'd be surprised at just how much you can fit in there. And how much more effort you put into thinking them through and making them smaller. It's actually rewarding!
Give it some time. My bet is that as you write your first dozen programs, you'll find that there's more space than you imagined.
thanks,
Bruce
Posts: 23
Threads: 5
Joined: Mar 2010
Enjoyed reading your comments on the calculators, I liked all the information. I am a Financial Planner and very much prefer the HP 12C but that was not always the case.
When I first started as a planner I would use a different financial calculator every day. I used the; Sharp EL-738, TI BA II, HP 10B, etc. Both older models purchased on e-bay and new models. I did not use the HP 12C as I purchased one but did not use it enough to see the advantages of RPN. Several years ago I tried it again. I learned the advantages of RPN and also loved the fact I could do most financial calculations with a lot fewer keystrokes. Now I love it and use it most of the time.
I do have a HP17bll+ which I have also put a few programs in but still prefer the HP 12C.
I also purchased two HP 30b calculators and have input one program just for fun.
For a person who truly is using the calculator as a financial calculator, the key set up on the HP 12C is unbeatable. I have several of the newest model that will calculate the FV, for example, right away.
▼
Posts: 1,392
Threads: 142
Joined: Jun 2007
Hi Joe. If you are going to write programs for the 30b, this file may be of assistance to you. Capitalization in the file name is important, BTW.
Posts: 553
Threads: 57
Joined: Sep 2006
As both Eddie and Don have pointed out, I think you'll find that the 30b is much easier and faster to use than it first seems. You really should look at the keyboard and give it a shot.
Thanks,
Bruce
▼
Posts: 349
Threads: 66
Joined: Apr 2007
Now that the machine is announced and shipping, I would *love* to see where the HP marketing folks have advertised this unit. How can the world learn about it? I receive a couple of emails from HP each week, advertising PC-related products and I have found none so far with any calculator items. Does HP advertise in print any more, even in financial pubs? If anybody knows of an ad for HP calcs *anywhere* in print or on line (but not HP's website), please let me know. And if they're not doing it, why not? It can't cost THAT much to sneak something into an HP-originated email about a calculator, could it? I know, corporate rules, blah blah blah, but you can't buy something you don't know exists.
▼
Posts: 1,248
Threads: 33
Joined: Aug 2007
Jake, you raise a VERY interesting point, which applies to the 35s, new faster 12c, etc. We have had many threads on here debating cost to add this or that feature, or cost to improve keyboard quality, or how $5, $10, or $20 added to the final price would affect sales numbers, what market is HP going after, etc. etc. ad infinitum.
But I do not remember any discussions on HP's marketing efforts! Come to think of it, neither do I recall seeing any HP calculator ads, even in their own emails. You have to dig into the website to find the calculators. What does this say about HP's investment in the calculator market?
This is an important topic, so since your post was No. 51, few are going to read it; I suggest you start a new thread with this topic.
Edited: 27 Mar 2010, 7:41 p.m.
|