HP Forum Archive 20

[ Return to Index | Top of Index ]

OT: Limited Ed. Message #1 Posted by Egan Ford on 9 Sept 2011, 12:02 p.m. http://nikemag.ebay.com/shoes Only 110 pairs. A previous article stated 1500. I guess there will be more auctions. To bring this a bit into topic, use your fav calc and see if the shoe sizes are a normal distribution.

Re: OT: Limited Ed. Message #2 Posted by Mike Morrow on 9 Sept 2011, 3:32 p.m., in response to message #1 by Egan Ford I just got my two from buy.com. Ordered them yesterday. HP 15C LE, Batch CNA132087X, Limited Edition Numbers 1577 and 1614. It's for real! Edited: 9 Sept 2011, 3:38 p.m.

Re: OT: Limited Ed. Message #3 Posted by Lode on 9 Sept 2011, 4:00 p.m., in response to message #2 by Mike Morrow Not sure if you saw it, but the original post was about Nike shoes, not HP 15C LE calculataors!

Re: OT: Limited Ed. Message #4 Posted by Mike Morrow on 9 Sept 2011, 4:30 p.m., in response to message #3 by Lode Yea...but when talking about limited edition pairs...I had a pair to talk about. :-) Self test don't work though...not too unexpected. Overall quality below original. Edited to change "far below" to "below". It is unrealistic to expect the quality of the original. Edited: 9 Sept 2011, 6:40 p.m. after one or more responses were posted

Re: OT: Limited Ed. Message #5 Posted by Michael de Estrada on 9 Sept 2011, 5:12 p.m., in response to message #4 by Mike Morrow Do you mean that they don't include self-tests, or that they fail ?

Re: OT: Limited Ed. Message #6 Posted by Steve Fennell on 9 Sept 2011, 5:31 p.m., in response to message #4 by Mike Morrow Could you post a little more detail about the quality and differences between the two generations of the 15C?

Re: OT: Limited Ed. Message #7 Posted by Tim Wessman on 9 Sept 2011, 5:48 p.m., in response to message #6 by Steve Fennell Looks like cyrille forgot to patch the old self tests away, or at least slow them down. The old self tests are there. The keyboard one works but times out quickly. If you start pressing keys immediately and don't wait more then about 2 seconds between each, you pass it fine. The second one runs too quickly and thus fails. The real menus used now are the 12c+ ones that have the checksum and LCD test. If you think the quality on a 12c+ is crap, you will think the 15c+ is crap too. They are the same except for firmware and cosmetics. TW Edited: 9 Sept 2011, 5:48 p.m.

Report: 15C LE Self Tests and Benchmark Run Times Message #8 Posted by Mike Morrow on 9 Sept 2011, 6:25 p.m., in response to message #7 by Tim Wessman I was unfair stating that the 15C LE quality was far below that of the original 15C. I would say it is perceptibly less than, but not far below, the original. It is certainly serviceable. I was initially influenced negatively by the self test performance failures. In that area: "ON with /" tries to work for a fraction of second, then ends with "Error 9". "ON with *" immediately results in "Error 9". "ON with -" immediately results in "Pr Error" and clears the memory and programs, ALL as expected. "ON with +" immediately results in "Error 9". The manual, page 261, describes the self tests as functioning in the manner of the original 15C, but they do not. In other areas: The 15C LE produces identical numerical results as the 15C, including the sequence of RAN# outputs. Performance of a 2500-iteration Savage benchmark takes 48 seconds on the 15C LE, and 5840 seconds on the 15C. That shows a runtime improvement by a factor of 122. It's worth the money, IMHO. Still, I'd prefer a 42S LE...greatly. :-) Edited: 9 Sept 2011, 8:02 p.m.

Re: OT: Limited Ed. Message #9 Posted by Mike Morrow on 9 Sept 2011, 6:43 p.m., in response to message #7 by Tim Wessman Quote: The real menus used now are the 12c+ ones that have the checksum and LCD test. So how are the real tests accessed? I don't find mention of them in the manual.

Re: OT: Limited Ed. Message #10 Posted by Steve Fennell on 9 Sept 2011, 6:51 p.m., in response to message #7 by Tim Wessman Haven't seen a 12c+ yet. I have a 12c 30th and 15c LE on order though.

Re: OT: Limited Ed. Message #11 Posted by uhmgawa on 9 Sept 2011, 7:08 p.m., in response to message #7 by Tim Wessman Quote: Looks like cyrille forgot to patch the old self tests away, or at least slow them down. The old self tests are there. The keyboard one works but times out quickly. If you start pressing keys immediately and don't wait more then about 2 seconds between each, you pass it fine. The second one runs too quickly and thus fails. Unless the tests are stumbling on an incorrect checksum due to rom patching, it likely isn't a question of speed, but rather quirks in the r2d2 which cause the "on" + "x" and "on" + "+" tests to fail in your emulation. I puzzled on that early when writing KEMU on the unfulfilled belief those tests (well they're essentially the same) would ferret out implementation bugs in KEMU. Unfortunately for as much as they try to be, they aren't able to exhaustively functional test all instruction semantics. Rather what I'd done as a more definitive benchmark was to take one of the NUT cpus left over from a KINOMI conversion and place it in a test jig PCB such that I could instrument it via feeding it instructions and examining the results. Some of the internal NUT state needs to be inferred, but IIRC it should provide complete coverage. Regrettably that project has fallen derelict due to available time (what's that?), but I should have extra break-out NUT PCBs if anyone would be interested to complete the effort. Concerning rom patching, I'd sidestepped the issue as while KEMU running on a atmega1284p realizes a 5-6x speed increase over the stock NUT, it wasn't as drastic as you'd find on a 32bit SoC running at a substantially higher clock rate. And while keyscan debounce timing was made independent of emulation speed, other scenarios exist such as <f> + "prefix" which you can see here, timeout the delay at the actual emulation speed. Quote: If you think the quality on a 12c+ is crap, you will think the 15c+ is crap too. They are the same except for firmware and cosmetics. Please don't say that in earshot of the Evilbay sellers now stuck with $400 NUT Voyagers lest they start marketing them as "Classic Voyagers" and we'll never be rid of them. Edited: 9 Sept 2011, 8:20 p.m.

Re: OT: Limited Ed. Message #12 Posted by Katie Wasserman on 9 Sept 2011, 10:18 p.m., in response to message #11 by uhmgawa On My 12C+ (not LE) the "+" and "x" self tests work just fine. This is version 2009-11-19. The "/" self test does time out if your wait more than 2 seconds between key presses but this is because it's running 150 times faster then the original 12c. The "/" self test on the original 12c (10c, 11c, 16c and 15c) also times out but you need to wait for around 5 minutes. I don't understand why these self tests didn't make it into the 15C LE firmware. -Katie

Re: OT: Limited Ed. Message #13 Posted by uhmgawa on 9 Sept 2011, 10:50 p.m., in response to message #12 by Katie Wasserman Quote: On My 12C+ (not LE) the "+" and "x" self tests work just fine. This is version 2009-11-19. The "/" self test does time out if your wait more than 2 seconds between key presses but this is because it's running 150 times faster then the original 12c. The "/" self test on the original 12c (10c, 11c, 16c and 15c) also times out but you need to wait for around 5 minutes. I don't understand why these self tests didn't make it into the 15C LE firmware. I'm sure it's in there, but apparently needs a little further emulator support. Actually I wasn't clear in my earlier comment. The 15c requires additional emulator support for the "on" + "x" system self test beyond the equivalent tests of the 11c/12c/16c. There are other model specific quirks such as the 12c having a slightly different power-on detect of the "x" key used to enter the self test relative to all other voyager models. Unsure whether that could be coming into play for the 15c here. With the exception of the <f> + "prefix" sequence which is just a countdown loop, thus far I've been able to push the time sensitivity of key input into the emulator. It may be possible to deal with the <f> + "prefix" as well in the emulator by using the PC at the time of associated "CHK KB" instruction (and perhaps partial stack content) to isolate the associated key release polling scenario, and rate limit the emulation instruction execution speed until it's complete. Depending on religion it may be preferable to patching the firmware to achieve the same result.

Re: OT: Limited Ed. Message #14 Posted by M. Joury on 9 Sept 2011, 11:20 p.m., in response to message #12 by Katie Wasserman The ON + "+" on my 12C+ never seems to come back from the "running" screen. ON + "\" and ON + "X" seem to work fine except for the 2 second time out.

Re: OT: Limited Ed. Message #15 Posted by Katie Wasserman on 10 Sept 2011, 12:14 a.m., in response to message #14 by M. Joury Quote: The ON + "+" on my 12C+ never seems to come back from the "running" screen. Mine stops almost instantly upon pressing any key. What version of the firmware do you have in your 12c+?

Re: OT: Limited Ed. Message #16 Posted by M. Joury on 10 Sept 2011, 12:29 a.m., in response to message #15 by Katie Wasserman Hi Katie, Yes, indeed it does (with all LCD segments other than the battery low indicator "lit"). I got my tests confused and forgot that you had to press a key to terminate the On + "+" test. My bad. Thanks for reminding me! Cheers, -Marwan

Re: OT: Limited Ed. Message #17 Posted by Howard Owen on 9 Sept 2011, 4:28 p.m., in response to message #2 by Mike Morrow But that's good news anyway!

[ Return to Index | Top of Index ]

Go back to the main exhibit hall