hp 35s hexadecimal math....what were they thinking?



#22

Here's a line from the manual that we should take note of....

"In RPN mode, the keys sin, cos, tan, sqrt, y^x and 1/x act as a short cut to enter the digits A to F. In ALG mode, press the RCL A,B,C,D,E,F to enter the digits A to F."

To finish an entry, you have to hit the base key again, use the down arrow key to append a "h" to the number and you are all done! You have to do that with each entry... I guess they want to punish the user for trying to do hexidecimal math...


#23

Yes, this is a major PITA. But it gets even worse. If you forget
to append the "h" one of two things will happen... If the hex
number contains one of "ABCDEF" it will complain about a syntax
error and leave the cursor under the first digit. But if the number
does not contain one of "ABCDEF"... it will happily enter the
number... and assume that the number is decimal and convert it
to hexadecimal.

I end up cursing at least once for every time I try to use the
hexadecimal, which is a large fraction of what I use a calculator
for.

But the 35s is so much nicer to look at, with the ENTER key in
the right place, that I haven't gone back to the 33s or 16c yet.

Monte

#24

Agreed on almost all comments. It's a mess.

The good news is that this begs for a decent set of quick and easy programs for doing this the right way. I.e., taking what they've given us and creating a proper implementation.

Anyone?

thanks,
bruce


#25

Now that's the spirit!

#26

I'll probably get completely raked over the coals for this comment; I'm ready for it.

How many users of "scientific calculators" actually use hexidecimal? I realize some programmers and computer techs may. But I'm talking engineers, chemists, physicists, mathematicians, statisticians, etc. I know a lot of people rely on binary, octal, and hexidecimal in their professions, but I'm curious to know a percentage of actual users who do. I'm guessing a lot fewer than 1%. Is the 16C not the preferred device for this group?

I'm more interested in a calculator that can operate in any base, includign e, pi etc. Some extremely interesting mathematics occurs in these bases. I wrote a program for the 50g to do this. Maybe one for the 35s is the way to go. But I still have two days to wait for mine. :(

Let me have it! :)

CHUCK


#27

Coals? No -- you've asked a reasonable question.

As a programmer, I probably use hex or binary once a week or so. I probably do at least some sort of basic functionality in these bases that often, sometimes more. It's generally limited to switching between bases, but frequently includes basic math, and basic logic operations too.

Personally, I think the issue is less about how MUCH we need it, but more about how EASY it is to use.

I would really prefer to use the 16c for this work -- as you postulated -- but it is such a rare and valuable calc that I won't subject it to use at work; dust, spills, theft, etc., are all concerns. If the 16c's were still in production, or available in copious quantities, then yes, I would totally be using that calc instead. There has been nothing like it since it came out.

Since they aren't prolific, I rely on my "desk calc", whatever model that is, to do my occasional base work. Here lies the second half of the problem...

If it's difficult to use, or not obvious in how to do something, then I have a big problem with that calc. Partly because I want to be able to just pound out my answers, but also because I have to trust the numbers. If I have to really carefully watch as I type in each number, make sure I've pressed the right operation, check to see the number makes sense, no errors, etc., then I've worked too hard at it. Some would argue that I should *always* do that anyhow. True. But with the 16c, for example, I never have to doubt my work -- it was so intuitive that the results were rock solid. With the new 35s (and the 33s before it; the 50g which is still different yet; etc), I always had to make sure I clearly was doing the right thing and the answers were plausible in context. That sucks. It needs to be easy to use, and rock solid in what it does without making me doubt my work.

A good example is what someone already posted. Go into hex mode and enter a number. Notice that if you enter "25c", it complains of syntax error. Ugh. If I just enter "25", it takes it and converts it. Dangerous. I have to go put a "h" at the end to make sense in that context (a multi-keystroke step). At that point, I'm working too hard. That's not even talking about math or logic -- that's just entering a number.

I still think a program to emulate the 16c behavior is needed -- or something of that sort. Anyone else have ideas or comments?

My $0.02 anyhow.

thanks,
bruce

Edited: 23 July 2007, 8:44 p.m.


#28

Quote:
I still think a program to emulate the 16c behavior is needed -- or something of that sort.


Jake????

#29

Quote:
How many users of "scientific calculators" actually use hexidecimal?

I, for one, do on a daily basis.


Quote:
Is the 16C not the preferred device for this group?

I keep a 16c with me at work for just this reason. However I frequently have to reach for a different model to do other calculations. The 16c is retarded for most things apart from binary/hex stuff :-(

A simple programmable calculator which had base arithmetic and the normal scientific functions all equally accessable would be a real win for me. I don't care for complex support, matricies or hyperbolics...


- Pauli

#30

Quote:
How many users of "scientific calculators" actually use hexidecimal?

I don't know how many, but I'm one of them. I'm actually a programmer by trade, who happens to want the "scientific" functionality because I'm also interested in math, physics, and whatnot.


If only they'd implemented bin/oct/hex like they did in the HP-42S! I use that functionality a lot, and in the 42S it never trips me up, and it doesn't annoy me by requiring lots of keystrokes. (Which is more than I can say for the RPL series, and, from what I hear, the 35S!)

- Thomas

Edited: 23 July 2007, 8:55 p.m.


#31

Quote:
If only they'd implemented bin/oct/hex like they did in the HP-42S! I use that functionality a lot, and in the 42S it never trips me up, and it doesn't annoy me by requiring lots of keystrokes. (Which is more than I can say for the RPL series, and, from what I hear, the 35S!)

Certainly agreed here: For bin/oct/hex and logic, I'd say that the HP-16C is most robust, the Pioneer-series models are well-conceived, and the RPL-based models are cumbersome. Here's a lengthy post of mine from the Archives, which morphed into a discussion of simulators vs. emulators:


http://www.hpmuseum.org/cgi-sys/cgiwrap/hpmuseum/archv015.cgi?read=77331#77331

#32

Quote:
How many users of "scientific calculators" actually use hexidecimal? I realize some programmers and computer techs may. But I'm talking engineers, chemists, physicists, mathematicians, statisticians, etc. I know a lot of people rely on binary, octal, and hexidecimal in their professions, but I'm curious to know a percentage of actual users who do. I'm guessing a lot fewer than 1%. Is the 16C not the preferred device for this group?

I'm an electronics design engineer and I need both binary and hex capability (rarely care about octal). Not on a daily basis, but when I need it, I need it, and as a not often used function it must be easy and foolproof to use. I don't want to have to read the manual after 6 months to figure it out again.

For that purpose I find the way the base functions work on my 20S to be ideal. It's the same as the Casio models where the HEX, BIN, OCT, DEC keys put the calc into the required mode and automatically convert the current displayed number from whatever mode it was in.
The SHOW button switches between groups of 8 bits in binary mode.

Dave.

#33

No coal raking here...sounds like a legitimate question.
Most people don't need this functionality everyday, but some of us do.

As an embedded device developer, I use Hex-Bin every day.
This is the only way to make any sense of complex port register settings and controlling things like LED matrices.

Another big use is in explaining in my comments how the hex numbers being used in a program reflect the settings of bitwise flags, pointers and port control settings.

Also, I am an Engineer, both Computer Science and Electronics

Edited: 23 July 2007, 10:14 p.m.

#34

Well in electronics and some programming I use hex and binary conversions to and from decimal now and then. I can do binary <-> hex in my head. I'm surprised it looks so clumsy on the new HP35S with those extra button presses needed and it can trip you up with some numbers if you just press enter (sounds like a recipe for disaster), my HP32Sii seems to be much more elegant.

For those who don't use hex: I hardly ever use statistic functions (or SIN COS and TAN) but I'd be disappointed if my shiny new HP calc didn't do them very well.

#35

I'm writing a subnet calculator suite. Hex mode has me so annoyed, I do all the binary math in decimal mode. I do switch to hex occasionally to do a quick conversion or to verify I have my masks working right. It's really awkward, but not nearly so as the native base modes are.

The one advantage I can see to tagging the integer on the stack is so you can do mixed-base calculations. But who needs to do integer math with numbers of different bases? The only other thing I can figure is that, for some reason, they couldn't use the current base as a clue for what base to put an entry into. I'd hate to think that this major annoyance was intentional and optional! I can see requiring the tag when you are in, for instance, binary mode, and want to do hexadecimal input. But the assumption of decimal input in base modes other than decimal is evil, evil, evil!

Regards,

Howard

#36

I'd be curious to see the rationale for having to add the trailing 'h'. The only justification I can see would be that from a decimal mode, you can enter some 1s and 0s and decide it's a binary and hit BASE 8.
But to enter an hex number, the HEX mode has to be active to access ABCDEF, so ENTER should be enough!

#37

I've never ever used alternative bases for any real work, ever. But I'm not a "programmer" nor do I write machine code (other than RPN programs:-)

#38

Easy to use (fast) logic operations are one of the main reasons I use a calculator.

I was trying to understand what was going on in the Base Logic function learning module, and now I understand... this is crazy. While I'm happy that the logic functions are there (I really miss them on the 33s), when I'm working in hex, I'm _only_ working in hex, and my mind just switches as if decimal doesn't exists. I'm going to forget the little h, and it's going to be a mess.... Darn it!

-Jonathan


#39

Well, it doesn't help matters that I have a typo in the answer for problem 1, does it?

I've sent the correction off to HP to be replaced on the site.

At least the answer shown in the screen captures is correct.

#40

Easily, the BASE handling is the most unfortunate aspect of an otherwise delightful 33s upgrade.

And why did they make the SIN, COS, etc. row serve as the hexadecimal A-F input keys?

I can see that many of the keys actually labeled A-F are assigned primary functions that should terminate number entry.

So why, then, not put the SIN, COS, etc. row across the top of the keyboard and label its keys A-F?

Well, yes, then the arrow keys would be one row lower, but would that have made any real difference?

So, let me get this straight: The roll down and swap keys are removed from logical proximity with ENTER, etc. -- to make room for a six-key row of real-number functions (SIN, COS, etc.) -- which can then run completely across the keyboard and hint at an otherwise unapparent A-F sequence -- so that the (some might say ill-advised) arrows may be placed immediately below the display?

My, what tangled webs we weave . . .

And the requirement for trailing bits of syntactic sugar ("h", "o", etc.)? Bummer!

I'm trying to remember what was the device or system for which the simple use of a leading zero during integer entry was the sole requirement for forcing conversion in the chosen alternate base rather than the default base 10 . . . There are simpler ways!

Even the lowly 33s had a much more straightforward implementation of the BASE feature.

--

This is, of course, a design critique benefiting from my total ignorance of any other pertinent issues there may have been, as well as something like 20/20 hindsight. Still, it does all seem more than a bit strained.

But, hey! The 35s is certainly a refreshing piece of work, otherwise & overall.


#41

Quote:
I'm trying to remember what was the device or system for which the simple use of a leading zero during integer entry was the sole requirement for forcing conversion in the chosen alternate base rather than the default base 10

I don't know where that convention first appeared, but the C programming language has it, and it was inherited by several other languages (C++ of course, and also tcl, perl, Java, JavaScript...)


So, should they also add the 0x prefix for hexadecimal? Do you think prefixes are better than suffixes? ;-)

- Thomas


#42

No, what I remember is this:

* Entering something like "123" was assumed decimal.

* Entering something like "123A" was assumed hexadecimal.

* But if you wanted "123h", you simply entered "0123".

. (No special character, no "x".)

Maybe it was WYLBUR, an obscure text editor / RJE facility on IBM mainframes. (I was the resident expert . . . )


Possibly Related Threads...
Thread Author Replies Views Last Post
  Need help understanding math.... cyrille de Brébisson 9 757 12-13-2013, 02:23 AM
Last Post: Didier Lachieze
  OT: a math competition site Pier Aiello 0 213 09-16-2013, 06:03 AM
Last Post: Pier Aiello
  HP's thinking behind the 20b/30b? John Ioannidis 3 374 09-07-2013, 10:21 AM
Last Post: Tim Wessman
  Simple Math Question Namir 2 294 08-09-2013, 06:13 PM
Last Post: Eddie W. Shore
  Cool math clock Bruce Bergman 28 1,688 04-10-2013, 03:13 AM
Last Post: Siegfried (Austria)
  HP-71B - thanks to Marcus von Cube for MATH ROM article Michael Lopez 2 288 03-03-2013, 07:19 AM
Last Post: Paul Berger (Canada)
  Math Challenge I could not solve Meindert Kuipers 22 1,348 01-05-2013, 04:43 PM
Last Post: Thomas Klemm
  Math Question Namir 0 185 11-06-2012, 07:43 AM
Last Post: Namir
  OT-Novus 3-level RPN. What were they thinking? Matt Agajanian 15 847 07-01-2012, 11:53 AM
Last Post: Marcus von Cube, Germany
  Survey for Special Math Problem Namir 7 532 06-03-2012, 09:46 PM
Last Post: Namir

Forum Jump: