34s conversions



#22

When I started typing this, it was to say that "Looks like the km -> miles conversion is wrong" because I had input 1 and the km->mile conversion was showing 0.9 something, which of course is not right. However, strangely enough, in the time it took me to get ready to type this, I stepped through the entire CONV catalog with a 1 as an input and checked against the values shown in the catalog (or the formulae shown) to see whether it matched. Very odd, because now 1 km->miles shows the 0.62... value it should. Could anything be off in the code of this section that might cause what I observed? I truly :-) don't think it was user error.

However, stepping through the catalog and the manual...

bar->Pa and Pa->bar are not in the manual on the table starting on page 66 but are in the machine.

mbar->Pa and Pa->mbar are in the table but not in the CONV catalog in the machine.


I would suggest checking the table on pages 68ff against the contents of the CONV catalog on the 2.0 machine.


#23

Quote:
I would suggest checking the table on pages 68ff against the contents of the CONV catalog on the 2.0 machine.

I would suggest checking said table in the current manual against the contents of the CONV catalog on the 2.0 machine ;-) We exchanged two pressure conversions recently (exactly those you found).

When testing, please make sure you start with the wp34s in a defined state. I've experienced strange results on the emulator sometimes, too, but only in the first operation. Especially if starting the emulator after loading a new build a RESET helps.

Hope that explains what you've seen.

Walter


#24

Ah, but I was looking at the 2.0 documentation and the flash version is the 2.0 one, revision 1085 from yesterday.

Check commands for 2.0: RCFRG and RCFST are present in X.FCN catalog on my flashed physical 34s. That's 2.0 only, isn't it? These commands are present in my machine and my CONV catalog does not have the mbar conversion that the 2.0 documentation suggests is present.

So, the issue existed at least with the 1085 flash version.

Were you checking with an emulator or a physical flashed unit?


#25

Hmmh, I don't know where you were looking at. Take the manual with #1085 on the title (committed under #1086), and you'll find on page 68 the conversion bar->Pa and on page 70 Pa->bar, perfectly matching the SW. So, what's your issue?

This holds for the flash as well as for the emulator.

Warning (though repeated some times already): The build numbers of SW and manual shall match, though not every build will result in a new manual revision. Moving targets, you know.

Walter


#26

My "issue" as you so helpfully framed the discussion was that the 2.0 manual which has no identifier in it has a table of conversions different from the build from a few days ago. The Emulator presently downloadable does not seem to have a "build" number associated with it, that I can find.

You have now successfully solved my "issue" by pointing out that the 2.0 manual in the emulator does not have a build number associated with it, hence the discrepancy.

Edited: 13 June 2011, 7:46 a.m. after one or more responses were posted


#27

Gene,

Calm down. The only point is you publicly claim a discrepancy which - based on the information you kindly gave us - cannot be reproduced. If stating that is what you call an insult then I'm sorry - I wasn't aware insulting is so easy. But, master - for sure you know already, then I apologize three times in advance, but nevertheless - VERS will tell you the build number. And:

Quote:
You have now successfully solved my "issue" by pointing out that the 2.0 manual in the emulator does not have a build number associated with it, hence the discrepancy.

I wonder where you saw me "pointing out" that? Since it's plain wrong: When you open said "manual in the emulator" you will immediately see the title picture telling you "34s 2.0 1089" right now. The last four digits are the "build number associated with it". So each and every manual we publish for months has an identifier in it. Your luck I'm not as insult-prone ... [;-)

Sorry, if anybody claims our work being faulty, then this person shall prove it. If (s)he can - fine - we've shown many times we listen and improve. If (s)he can't - well - "wer austeilt, muss auch einstecken können" as people say here. FWIW.

Walter


#28

Thank you, Walter. VERS returns no # when executed on this physical machine for reasons of which I am aware.

Perhaps you assume too much from those with whom you communicate. That was my original reason why I suggested creating a much more user-friendly manual and why I had requested the Word version of your manual as a starting point.

I would still like to have a Word version of your document to save me (and others) a great deal of time to create a manual with much less "assumed".

Thanks...

Edited: 13 June 2011, 7:46 a.m. after one or more responses were posted


#29

Thanks, Gene, for the explanations. I googled the phrase that you are "not particularly upset" about - BTW then I don't want to hear you upset [[;-) - and found nothing justifying any upset IMHO. But "issue" turns out having a lot of meanings, so maybe I missed one using English as a foreign language. I used that expression thinking of "Was ist jetzt eigentlich dein Problem?" being a very common phrase in such context here. Since "translates poorly" isn't a real explanation IMHO, I'd appreciate an enlightenment what was wrong with this sentence - just for sake of curiosity and continuous improvement.

OTOH I read texts here claiming plain wrong stuff more than once (well, they look plain wrong to me at least, like e.g. "there's no identification number in the manual"). So I try rectifying this but it seems the author didn't even notice the error :-? Very confusing!

Maybe I got it fundamentally wrong: I thought this being an international forum here. But it seems knowledge of all sorts of slang, local TV shows and movies is required from participants coming from beyond the English speaking world (yes, there is one, and it will survive), while the requirements on the other side are a bit lower ... :-(

(I'll never forget my family and me ROFL reading a "professional" multilingual ad in a San Diego hotel some years ago: it was "translated" into German - well, into what the translator meant (thought??) being German ... And we had good reasons to assume it happened in more languages there.)

So I sign off modestly wishing the educated people of this planet would learn one (!) second language at least, just to get a broader view (and maybe for symmetry reasons as well). But I know Xmas is still some months to come. My apologies for any wrong words, terms, contextual meanings, spellings etc. I may have picked.

Walter

Addendum: Though your sentence sounds different, you got a Word version of our manual many weeks ago already. No - I won't ask "what's your issue" again [[[;-)

Addendum 2: Just read this thread once again from the beginning and found the term "issue" was introduced by *you*. Confusion rising :-?


Edited: 13 June 2011, 4:20 a.m.


#30

Walter, why not check in the Word document in SVN where it belongs?

#31

What word version? How did I get it?


#32

It was on the 12th of April. Now, I just did forward the file to you once more at genewright143AThotmailDOTcom. HTH

#33

Quote:
You have now successfully solved my "issue" by pointing out that the 2.0 manual in the emulator does not have a build number associated with it, hence the discrepancy.

All versions have a build number embedded in the code that is automatically updated by SVN. The problem you may have encountered is the web download from the subversion tree which returns this field blanked out. A regular SVN client returns the files with the build number intact.

BTW, I'm not very pleased with the discussion as it surfaces from time to time that deals with style and personal matters. Would you be so kind and move the discussion out of the public forum, please?

#34

Quote:
When I started typing this, it was to say that "Looks like the km -> miles conversion is wrong" because I had input 1 and the km->mile conversion was showing 0.9 something, which of course is not right.

Weird. I've not seen any problems with this catalogue for a while now and it is one I always go through and test. We've had a few hiccups but none recently that I'm aware of. If you do manage to replicate this please let us know.

There was a (recently fixed) bug with stack lift and turning the device off (stack lift got itself disabled always over a power off and on) which was very confusing but is fixed now. Still, I don't see how this would have caused what you saw.

The VERS command in the emulator should include the subversion build number.


- Pauli

#35

While we are on conversions (hopefully this is not too much of a thread hijack) I have had a thought about how I work with conversions and noticed that I generally need to go back and forth a few times with complimentary conversions. I.e. hp->W and then W->hp.

For me, I would prefer the table of conversions to be organized with complementary conversions right next to each other. I know this would probably kill the ability to search for a conversion alphabetically but with the position within the table saved between uses, it means the complementary conversion is just an arrow away.

Any other thoughts on this?


#36

Sure it would be possible, not even terribly difficult.

However, I'm not keen because, as you point out, it would destroy the type alpha characters to get to the command functionality which is very useful. That code relies on the catalogue being sorted correctly.

As things stand, the catalogues are automatically sorted as part of the build process so we'd have to modify that as well which would mean going back to manual ordering which was a right pain in the a**.


If there is a conversion you require a lot, write small wrapper programs and use the hotkeys. Not as good as an assignable keyboard but workable nonetheless.


- Pauli


#37

Pauli, we might use a key in the conversion catalogue which just jumps to the reverse operation. It would need some code and rely on the spelling of the conversion commands (slow) instead of a special table (too demanding on flash space) but that seems possible.


#38

Marcus, good idea. Browsing a catalogue, however, alpha mode is set so finding an unused *and* comprehensive key will be challenging. Right now, f-shifted B is the only one I find.

Walter

Edited: 13 June 2011, 5:07 a.m.

#39

The opcodes are closely related for most of the conversions but there are enough exceptions to be annoying.

Rather than a key to jump to the opposite command, why not a key to do the conversion in reverse & exit the catalogue?

f-shift B -> Backwards perhaps?


- Pauli


#40

Coded, committed and built. The subversion binaries include f-shift B which performs the inverse of the currently displayed conversion. Note it doesn't go to the inverse of the currently display conversion, it does it.

I've also done some work on the hyperbolic functions and inverses to improve numeric stability and performance.


- Pauli


#41

Thanks. Whether you remember f-shifted B for "backwards" or for "1/x" doesn't really matter. Another issue solved ;-)

Walter

#42

Thank You!


Possibly Related Threads...
Thread Author Replies Views Last Post
  Interesting Base Conversions - Porting a 1975 HP 25 Program to the HP 35S Eddie W. Shore 1 200 10-13-2013, 07:49 PM
Last Post: BruceH
  Decimal/Binary Conversions for the HP 39gii - 3/16/2013 Eddie W. Shore 0 154 03-16-2013, 04:30 PM
Last Post: Eddie W. Shore
  Unit conversions inside programs on 50g Chris Smith 2 227 02-23-2013, 05:35 AM
Last Post: Chris Smith
  17bii unit conversions Dave Britten 12 510 03-27-2012, 11:59 AM
Last Post: Dave Britten
  WP 34S adding own conversions Lourensk 4 218 03-24-2012, 02:45 PM
Last Post: Walter B
  Dave Ramsey et al with early HP 41C conversions to CL Geoff Quickfall 4 258 01-01-2012, 12:51 AM
Last Post: David Ramsey
  [wp 34s] wp 34s picture and scan Jeroen Van Nieuwenhove 2 173 10-27-2011, 09:02 PM
Last Post: Les Wright
  Google onlinecalculator conversions etc. designnut 2 189 08-18-2011, 09:23 AM
Last Post: snaggs
  Wp34s unit conversions Cristian Arezzini 4 266 07-25-2011, 05:02 AM
Last Post: Paul Dale
  wp34S IEEE Floating-Point Conversions Jake Schwartz 12 494 06-08-2011, 10:46 AM
Last Post: Jake Schwartz

Forum Jump: