Looking forward to the DM42 update utilizing the Free42 2.023 update...
Looking forward to the DM42 update utilizing the Free42 2.023 update...
An up to dat DM42 is a better DM42
Re: Looking forward to the DM42 update utilizing the Free42 2.023 update...
I disagree. The changes in 2.023 are small, highly specialized and very unlikely to matter to almost anyone, and simply do not warrant a new release of DM42 on their own. It makes more sense to spend the limited time David has to continue work on other forthcoming DM42 enhancements, and then roll-in the 2.023 changes when the other enhancements are ready. YMMV...
--bob p
DM42: β00071 & 00282, DM41X: β00071 & 00656, DM10L: 071/100
DM42: β00071 & 00282, DM41X: β00071 & 00656, DM10L: 071/100
Re: Looking forward to the DM42 update utilizing the Free42 2.023 update...
1 +rprosperi wrote: ↑Tue Feb 12, 2019 3:14 amI disagree. The changes in 2.023 are small, highly specialized and very unlikely to matter to almost anyone, and simply do not warrant a new release of DM42 on their own. It makes more sense to spend the limited time David has to continue work on other forthcoming DM42 enhancements, and then roll-in the 2.023 changes when the other enhancements are ready. YMMV...
Personal note: I hate the never ending update cycle brought by apps on smartphones.
Greetings,
Massimo
ajcaton
-+×÷ left is right and right is wrong Casted in gold
Massimo
ajcaton
-+×÷ left is right and right is wrong Casted in gold
Re: Looking forward to the DM42 update utilizing the Free42 2.023 update...
I'm with Bob on this. However, if there is a serious enough bug discovered in Free42 that will have a definite impact on the DM42, then the updated version of Free42 will warrant a new DM42 release. This hasn't yet happened. On the few occasions that bugfixes have been needed in Free42, there was a new bunch of features for the DM42 going out anyway so it all got folded in together.
Now, take the example of Update 2 of the Intel Decimal Floating Point library. We rushed to add it in and guess what, it created problems. So we rolled back to the tried and tested Update 1 and can't see any reason yet to update again.
Now, take the example of Update 2 of the Intel Decimal Floating Point library. We rushed to add it in and guess what, it created problems. So we rolled back to the tried and tested Update 1 and can't see any reason yet to update again.
There are only 10 kinds of people in the world: those who understand binary and those who do not.
-
- Posts: 1100
- Joined: Tue May 02, 2017 5:48 pm
- Location: Netherlands
- Contact:
Re: Looking forward to the DM42 update utilizing the Free42 2.023 update...
It does look like I am running out of bugs to fix. There's still plenty on my to-do list, but looking at which items affect the core, that is, are not platform-specific user interface issues, this is all I have at the moment:
Regarding 2.0.23, the SOLVE fix addresses a case that is also rare, but which had been bugging me for quite some time:
With starting values 0 and 100, this returns 50 instantly, which is obviously wrong. The reason is that Γ(x) is undefined at 0, and SOLVE assumes infinity, causing the secant extrapolation to get stuck right away.
This scenario is mentioned in William Kahan's HP Journal article on HP-34C SOLVE. The real HP-42S does handle it, though it takes its time, evaluating the function 217 times. Free42 2.0.23 takes 223 evaluations to get there.
N.B. I sometimes get annoyed at the never-ending flood of updates on my smartphones, too, and after receiving complaints about the spate of updates between 2.0.7 and 2.0.20, I have started saving up more fixes before releasing.
2.0.23 could have waited, but I wanted to roll it out because the iPhone version had fallen behind due to my problems with iPhone X support, which I finally managed to solve last Saturday. That iPhone release actually went out with a lame iPhone X skin, so another update to address that is around the corner (waiting for approval right now), but after that, expect things to be quiet for a while.
Unless someone reports a serious bug, anyway.
- More accurate 2x2 matrix operations
- More accurate dot product using FMA
- Better handling of edge cases in →HMS, →HR, HMS+, and HMS-, specifically in the Binary version
- Better termination condition for integrals that converge to zero
- More lenient program pasting, for example, accept # for ≠
Regarding 2.0.23, the SOLVE fix addresses a case that is also rare, but which had been bugging me for quite some time:
Code: Select all
00 { 18-Byte Prgm }
01▸LBL "GM"
02 MVAR "X"
03 RCL "X"
04 GAMMA
05 10
06 -
07 END
This scenario is mentioned in William Kahan's HP Journal article on HP-34C SOLVE. The real HP-42S does handle it, though it takes its time, evaluating the function 217 times. Free42 2.0.23 takes 223 evaluations to get there.
N.B. I sometimes get annoyed at the never-ending flood of updates on my smartphones, too, and after receiving complaints about the spate of updates between 2.0.7 and 2.0.20, I have started saving up more fixes before releasing.
2.0.23 could have waited, but I wanted to roll it out because the iPhone version had fallen behind due to my problems with iPhone X support, which I finally managed to solve last Saturday. That iPhone release actually went out with a lame iPhone X skin, so another update to address that is around the corner (waiting for approval right now), but after that, expect things to be quiet for a while.
Unless someone reports a serious bug, anyway.
Re: Looking forward to the DM42 update utilizing the Free42 2.023 update...
Thanks for explaining the status of Free42 and your methodology for releasing updates Thomas. To clarify my post above, I was not suggesting the Free42 update was not necessary, only the DM42 update based solely on the changes in your 2.023. The multi-platform nature of Free42 dictates that there is probably always something that is needed to chase the various platform changes and tweaks, so it's not surprising if there are updates on one platform which seem unwarranted on another. That said, I'm not sure why iOS or Android Free42 users would complain about frequent updates, as they are (nearly always) instant and painless (e.g. no loss of calculator state, etc.) and Mac and Win users can simply choose to not update. Still, if they aren't reading the change log to understand what is being fixed, it may seem the updates are more frequent than "necessary".
--bob p
DM42: β00071 & 00282, DM41X: β00071 & 00656, DM10L: 071/100
DM42: β00071 & 00282, DM41X: β00071 & 00656, DM10L: 071/100
-
- Posts: 1100
- Joined: Tue May 02, 2017 5:48 pm
- Location: Netherlands
- Contact:
Re: Looking forward to the DM42 update utilizing the Free42 2.023 update...
I sympathize with those complaints to some extent... When my phone updates a dozen apps, less than a week after the last time I checked for updates, sometimes I do get a bit annoyed myself, and so I try not to add too much to that traffic. But people can complain about the weirdest things! I just had someone write a review in the Play Store where he criticized Free42 for not being a true RPN calculator, and in general not behaving like a real calculator. (Registers being initialized to zero, "calculators don't do that.") Claims to have 20 years of experience with HP calculators. None of them RPN models, I'm sure.
Re: Looking forward to the DM42 update utilizing the Free42 2.023 update...
One mans edge case is another man's national emergency. Thanks Thomas for maintaining such a great project for so long. I may have mentioned that I started using free42 back in the days before iPhone and Android. It's near unheard of to see that kind of longevity in pocket sized tech.
DM42 SN: 00076
DM15, HP 28c, HP 12c, HP 10b, HP 20b, HP 30b, HP 17bII, HP 42s, HP 35s, WP 34s, NP-25, TI-83+, teenix multicalc
DM15, HP 28c, HP 12c, HP 10b, HP 20b, HP 30b, HP 17bII, HP 42s, HP 35s, WP 34s, NP-25, TI-83+, teenix multicalc
Re: Looking forward to the DM42 update utilizing the Free42 2.023 update...
complaints? some people are just bloody ungratefulThomas Okken wrote: ↑Tue Feb 12, 2019 11:58 amN.B. I sometimes get annoyed at the never-ending flood of updates on my smartphones, too, and after receiving complaints about the spate of updates …
thank you for all your hard work.
Cambridge, UK
41CL/DM41X 12/15C/16C DM15/16 17B/II/II+ 28S 42S/DM42 32SII 48GX 50g 35s WP34S PrimeG2 WP43S/pilot
Casio, Rockwell 18R
41CL/DM41X 12/15C/16C DM15/16 17B/II/II+ 28S 42S/DM42 32SII 48GX 50g 35s WP34S PrimeG2 WP43S/pilot
Casio, Rockwell 18R
Re: Looking forward to the DM42 update utilizing the Free42 2.023 update...
I totally agree : thanks Thomas for your impressive job !cdmackay wrote: ↑Tue Feb 12, 2019 5:58 pmcomplaints? some people are just bloody ungratefulThomas Okken wrote: ↑Tue Feb 12, 2019 11:58 amN.B. I sometimes get annoyed at the never-ending flood of updates on my smartphones, too, and after receiving complaints about the spate of updates …
thank you for all your hard work.
HP33E, HP34C, HP41C, HP16C, HP28S, HP42S, HP48S, HP12C, HP35S, DM42, DM41L, DM15L, DM41X (SN#15) and HP3000 @ work during 18 yrs as IT Mgr