Everything is as expected with Windows 10. In the absence of mean, running the statement trajectory U with the J debugger raises the value error and the stack gives the line of the error with the usual spaces before the culprit token:
value error Fold_j_[42] S=:S-e* mean y,V trajectory[0] u F: v y That is much much better and allows to debug complex programs easily. Perfect! === About the Qt IDE, I note Windows: Qt IDE: 1.8.4/5.12.5 Linux: Qt IDE: 1.8.1/5.5.1 Let's hope for a Linux catch up. The solution is perhaps there. ~ Gilles Le 2019-10-17 à 19:25, 'robert therriault' via Beta a écrit : > I commented out mean by putting NB. at the front of the line of its > definition. The script that I ran is the one that is attached to the email > below and I see > >>> NB.mean =: +/ % # NB. definition of mean - VERB > immediately above the definition of trajectory. > > I am running > > iMac Pro (2017) > Processor 3 GHz Intel Xeon W > Memory 64 GB 2666 MHz DDR4 > Graphics Radeon Pro Vega 64 16 GB > > macOS Mojave Version 10.14.6 > > Cheers, bob > >> On Oct 17, 2019, at 4:17 PM, Henry Rich <[email protected]> wrote: >> >> It doesn't fail for me under Windows. The script you sent didn't have >> 'mean' commented out, but when I commented it out it still didn't fail. >> >> Henry Rich >> >> On 10/17/2019 7:03 PM, 'robert therriault' via Beta wrote: >>> Okay Henry, >>> >>> I started a fresh session >>> >>> JVERSION >>> Engine: j901/j64avx2/darwin >>> Beta-m: commercial/2019-10-02T11:28:59 >>> Library: 9.01.10 >>> Qt IDE: 1.8.3/5.12.3 >>> Platform: Darwin 64 >>> Installer: J901 install >>> InstallPath: /users/bobtherriault/j901 >>> Contact: www.jsoftware.com >>> >>> Opened debug (no verbs) >>> >>> loaded the following script from a temp file >>> >>> cocurrent 'base' >>> require 'plot' >>> MAXITER=: 50 NB. safety long-stop >>> S=: 100 NB. height above ground (m) - UPDATED >>> T=: 0 NB. time at end of current epoch (s) - UPDATED >>> e=: 0.1 NB. time interval of simulation epoch (s) - CONST >>> U=: 0 NB. starting velocity (m/s) - CONST >>> V=: U NB. velocity at end of current epoch (s) - UPDATED >>> g=: 9.81 NB. acceleration due to gravity at earth surface (m/s^2) - >>> CONST >>> NB.mean =: +/ % # NB. definition of mean - VERB >>> >>> trajectory=: monad : 'u F: v y' >>> >>> u=: monad define >>> NB. Trajectory of solid object in free-fall >>> _3 Z: MAXITER NB. Stop Fold when there have been too many >>> iterations >>> _2 Z: S<:0 NB. Stop Fold when object hits the ground >>> NB. Recurrence relation is based on free-fall formula: V = U + gt >>> V=: y + g*e NB. free-fall: vertical velocity at end of epoch >>> S=: S - e*mean y,V NB. free-fall: vertical height at end of epoch >>> T=: T + e NB. time of flight at end of epoch >>> if. S<:0 do. smoutput '+++ LANDED: T=',":T end. >>> V return. >>> ) >>> >>> v=: monad : 'S' >>> >>> plot trajectory U >>> >>> >>> Note that 'mean' is commented out, and when I loaded this script the >>> session crashed. >>> >>> I am emailing the crash report directly to you. >>> >>> Let me know if there is anything else you require. >>> >>> Cheers, bob >>> >>>> On Oct 17, 2019, at 3:51 PM, Henry Rich <[email protected]> wrote: >>>> >>>> Certainly, if you have a crash I'd like to know about it. What I need is >>>> a script that will crash when run from a fresh session. >>>> >>>> Henry Rich >>>> >>>> On 10/17/2019 6:47 PM, 'robert therriault' via Beta wrote: >>>>> Sorry about that Henry, I have updated now. >>>>> >>>>> On the last run, I ended up having to go into the Activity Monitor to >>>>> force quit the J process. >>>>> >>>>> When I reopened the session, I updated the addons (including base) and >>>>> restarted and the reloaded the project that the trajectory script was in. >>>>> When I loaded that script the first time: >>>>> >>>>> load '/users/bobtherriault/j901-user/projects/enhanced/jig6.ijs' >>>>> |fold limit: trajectory >>>>> | u F:v y >>>>> |[-346] /users/bobtherriault/j901-user/projects/enhanced/jig6.ijs >>>>> >>>>> Subsequent loading of the script results in everything running smoothly >>>>> even after executing clear '' plots display as expected >>>>> >>>>> Loading with mean commented out and debug off I get >>>>> >>>>> load '/users/bobtherriault/j901-user/projects/enhanced/jig6.ijs' >>>>> |value error: trajectory >>>>> | u F:v y >>>>> |[-346] /users/bobtherriault/j901-user/projects/enhanced/jig6.ijs >>>>> >>>>> Turning debug on after executing clear '' of course revealed no verbs in >>>>> the base locale. Loading the script again caused J to crash ( I can send >>>>> the crash report if that is useful) >>>>> >>>>> Restarting J and running the script I once again get the expected: >>>>> >>>>> load '/users/bobtherriault/j901-user/projects/enhanced/jig6.ijs' >>>>> |value error: trajectory >>>>> | u F:v y >>>>> |[-346] /users/bobtherriault/j901-user/projects/enhanced/jig6.ijs >>>>> >>>>> Then I clear '' and this time load the script first and then start the >>>>> debugger (the default is trajectory). Loading script crashes again ( >>>>> crash report saved for this one as well) >>>>> >>>>> Let me know if you want the crash reports. >>>>> >>>>> Cheers, bob >>>>> >>>>>> On Oct 17, 2019, at 3:17 PM, Henry Rich <[email protected]> wrote: >>>>>> >>>>>> You have not updated the dev/fold addon to 1.0.5 level. >>>>>> >>>>>> Henry Rich >>>>>> >>>>>> On 10/17/2019 6:14 PM, 'robert therriault' via Beta wrote: >>>>>>> Hi Gilles, >>>>>>> >>>>>>> When I comment out the definition of mean and then clear the base >>>>>>> locale and run with debug on I get the following in the top pane of the >>>>>>> debugger I get 59 lines with a '>' indicating line 41 (would be simpler >>>>>>> if I could copy the results that show up in the pane but don't know >>>>>>> how). >>>>>>> >>>>>>> In the second pane I get: >>>>>>> value error >>>>>>> Fold_j_[41] 13!:8 ] 13!:11''[FoldZv_j_=:fzv >>>>>>> trajectory[0] u F: v y >>>>>>> script[0] {: 3 : '0!:0 y [ 4!:55<''y''' jpath_z_&.:> >>>>>>> fn[tacit] script >>>>>>> load[7] fn fl >>>>>>> load[0] 0 load y >>>>>>> >>>>>>> I am not sure how to interpret that, but is the result that I get. I am >>>>>>> running it by loading a script with the definitions and final line: >>>>>>> >>>>>>> plot trajectory U >>>>>>> >>>>>>> I get the same result running that line in the Qt IDE >>>>>>> >>>>>>> And now the debugger has become non operational. Buttons do highlight, >>>>>>> but do not do anything and the debugger will not clear or exit. The >>>>>>> script window can be edited and loaded but the execution window only >>>>>>> responds with the load command and does not update. Toggling between >>>>>>> windows does not include the debug window, although it can be selected >>>>>>> when clicked on. >>>>>>> >>>>>>> Hope this helps! >>>>>>> >>>>>>> Cheers, bob >>>>>>> >>>>>>> >>>>>>> >>>>>>>> On Oct 17, 2019, at 2:26 PM, Gilles Kirouac <[email protected]> wrote: >>>>>>>> >>>>>>>> Bob >>>>>>>> >>>>>>>> The trajectory script uses *mean* in the definition of u. If you run >>>>>>>> trajectory *without* the definition of mean, it should result in a >>>>>>>> value error. Right? >>>>>>>> >>>>>>>> Now run the same under the J debugger to see if it can pinpoint that >>>>>>>> the value error is on *mean*. >>>>>>>> >>>>>>>> I will do the same on Win10 shortly. >>>>>>>> >>>>>>>> ~ Gilles >>>>>>>> >>>>>>>> Le 2019-10-17 à 17:05, 'robert therriault' via Beta a écrit : >>>>>>>>> Trajectory runs fine for me on a Mac. >>>>>>>>> >>>>>>>>> JVERSION >>>>>>>>> Engine: j901/j64avx2/darwin >>>>>>>>> Beta-m: commercial/2019-10-02T11:28:59 >>>>>>>>> Library: 9.01.08 >>>>>>>>> Qt IDE: 1.8.3/5.12.3 >>>>>>>>> Platform: Darwin 64 >>>>>>>>> Installer: J901 install >>>>>>>>> InstallPath: /users/bobtherriault/j901 >>>>>>>>> Contact: www.jsoftware.com >>>>>>>>> >>>>>>>>> I note that the Qt IDE and the libraries differ from Gilles, but this >>>>>>>>> would be natural for a different operating system, right? >>>>>>>>> >>>>>>>>> Cheers, bob >>>>>>>>> >>>>>>>>>> On Oct 17, 2019, at 1:51 PM, Henry Rich <[email protected]> wrote: >>>>>>>>>> >>>>>>>>>> I don't have Linux. It runs correctly on Windows for me. >>>>>>>>>> >>>>>>>>>> Henry Rich >>>>>>>>>> >>>>>>>>>> On 10/17/2019 4:30 PM, Gilles Kirouac wrote: >>>>>>>>>>> I am now getting an interpreter crash. >>>>>>>>>>> >>>>>>>>>>> I updated three scripts through PacMan. dev/fold is now at level >>>>>>>>>>> 1.0.5. Then ran the same Trajectory script. Without the debugger, I >>>>>>>>>>> get >>>>>>>>>>> the normal value error: trajectory. Doing the same with the J >>>>>>>>>>> debugger >>>>>>>>>>> but with the J session run under gdb here is the result: >>>>>>>>>>> >>>>>>>>>>> >>>>>>>>>>> gilles@T530gk ~/j901 $ gdb bin/jqt >>>>>>>>>>> ... >>>>>>>>>>> (gdb) run >>>>>>>>>>> Starting program: /home/gilles/j901/bin/jqt >>>>>>>>>>> [Thread debugging using libthread_db enabled] >>>>>>>>>>> Using host libthread_db library >>>>>>>>>>> "/lib/x86_64-linux-gnu/libthread_db.so.1". >>>>>>>>>>> [New Thread 0x7fffe4b49700 (LWP 8058)] >>>>>>>>>>> >>>>>>>>>>> Thread 1 "jqt" received signal SIGSEGV, Segmentation fault. >>>>>>>>>>> 0x00007fffda5cd1ed in suspset () from /home/gilles/j901/bin/libj.so >>>>>>>>>>> (gdb) >>>>>>>>>>> >>>>>>>>>>> I did it twice to make sure I was not doing something silly. >>>>>>>>>>> >>>>>>>>>>> ~ Gilles >>>>>>>>>>> >>>>>>>>>>> Le 2019-10-17 à 11:51, Henry Rich a écrit : >>>>>>>>>>>> Gilles, >>>>>>>>>>>> >>>>>>>>>>>> I have modified the Fold script so that it allows messages from >>>>>>>>>>>> execution of the user's verb to be displayed and debugged. Use >>>>>>>>>>>> Package Manager to update dev/fold - no need for a new beta. >>>>>>>>>>>> >>>>>>>>>>>> I hope that suits better. >>>>>>>>>>>> >>>>>>>>>>>> Henry Rich >>>>>>>>>>>> >>>>>>>>>>>> On 10/11/2019 11:27 AM, Gilles Kirouac wrote: >>>>>>>>>>>>> I installed beta-k then m on Linuxmint 64b >>>>>>>>>>>>> >>>>>>>>>>>>> My desktop directory is named Bureau, not Desktop. The installer >>>>>>>>>>>>> is thus >>>>>>>>>>>>> unable to install shortcuts on Desktop. Minor annoyance. I >>>>>>>>>>>>> created a >>>>>>>>>>>>> link to Bureau for next time. May be suggest to do so in install >>>>>>>>>>>>> instructions. I added a few qt5 missing packages. >>>>>>>>>>>>> >>>>>>>>>>>>> With the current IDE, to find the syntax of std lib utilities >>>>>>>>>>>>> (e.g. >>>>>>>>>>>>> dates, files, strings, text), I have to go through Help, Help, >>>>>>>>>>>>> Usr, >>>>>>>>>>>>> Standard Library, ... More importantly I have to KNOW that the >>>>>>>>>>>>> std lib >>>>>>>>>>>>> doc is there. >>>>>>>>>>>>> >>>>>>>>>>>>> Do you think a newcomer will easily explore to that level? The >>>>>>>>>>>>> Library >>>>>>>>>>>>> is a selling point. It should be easy to get to. And IMO the Lib >>>>>>>>>>>>> is more >>>>>>>>>>>>> useful to anyone than Foreigns. >>>>>>>>>>>>> >>>>>>>>>>>>> Two years ago, I proposed to add a line to the Help Menu pointing >>>>>>>>>>>>> to the >>>>>>>>>>>>> std lib doc. IIRC the proposal was accepted, but ... so I >>>>>>>>>>>>> reiterate. >>>>>>>>>>>>> >>>>>>>>>>>>> Having launched jqt through a terminal, I received the following: >>>>>>>>>>>>> >>>>>>>>>>>>> QPaintDevice::metrics: Device has no metric information >>>>>>>>>>>>> >>>>>>>>>>>>> Should I worry? >>>>>>>>>>>>> >>>>>>>>>>>>> I briefly experimented with Fold after reading the Release Notes. >>>>>>>>>>>>> Those >>>>>>>>>>>>> could include a link to >>>>>>>>>>>>> https://code.jsoftware.com/wiki/Vocabulary/fcap, >>>>>>>>>>>>> which I found through a web search. >>>>>>>>>>>>> >>>>>>>>>>>>> I tried to run the Trajectory example from that page and got a >>>>>>>>>>>>> value >>>>>>>>>>>>> error. Where? A visual look at the script revealed a susceptible >>>>>>>>>>>>> culprit: mean. Added require 'stats/base' and it ran. >>>>>>>>>>>>> >>>>>>>>>>>>> While trying the Trajectory, I ran the script with the debugger. >>>>>>>>>>>>> [The >>>>>>>>>>>>> debugger did not fail, great!] The value error on mean was not >>>>>>>>>>>>> obvious. >>>>>>>>>>>>> The stack is given as: >>>>>>>>>>>>> value error >>>>>>>>>>>>> Fold_j_[40] 13!:8]13!:11''[FoldZv_j_=:fzv >>>>>>>>>>>>> trajectory[0] u F: v y >>>>>>>>>>>>> >>>>>>>>>>>>> This Trajectory example is a simple case. I suspect that errors >>>>>>>>>>>>> in more >>>>>>>>>>>>> complex cases may not be easy to analyze. >>>>>>>>>>>>> >>>>>>>>>>>>> From a notational point of view, the Fold family is a great >>>>>>>>>>>>> addition, I >>>>>>>>>>>>> think. >>>>>>>>>>>>> >>>>>>>>>>>>> JVERSION >>>>>>>>>>>>> Engine: j901/j64avx/linux >>>>>>>>>>>>> Beta-m: commercial/2019-10-02T09:15:44 >>>>>>>>>>>>> Library: 9.01.10 >>>>>>>>>>>>> Qt IDE: 1.8.1/5.5.1 >>>>>>>>>>>>> Platform: Linux 64 >>>>>>>>>>>>> Installer: J901 install >>>>>>>>>>>>> InstallPath: /home/gilles/j901 >>>>>>>>>>>>> Contact: www.jsoftware.com >>>>>>>>>>>>> ~ Gilles >>>>>>>>>>>>> >>>>>>>>>>>>> ---------------------------------------------------------------------- >>>>>>>>>>>>> For information about J forums see >>>>>>>>>>>>> http://www.jsoftware.com/forums.htm >>>>>>>>>>> ---------------------------------------------------------------------- >>>>>>>>>>> For information about J forums see >>>>>>>>>>> http://www.jsoftware.com/forums.htm >>>>>>>>>> -- >>>>>>>>>> This email has been checked for viruses by AVG. >>>>>>>>>> https://www.avg.com >>>>>>>>>> >>>>>>>>>> ---------------------------------------------------------------------- >>>>>>>>>> For information about J forums see >>>>>>>>>> http://www.jsoftware.com/forums.htm >>>>>>>>> ---------------------------------------------------------------------- >>>>>>>>> For information about J forums see http://www.jsoftware.com/forums.htm >>>>>>>> ---------------------------------------------------------------------- >>>>>>>> For information about J forums see http://www.jsoftware.com/forums.htm >>>>>>> ---------------------------------------------------------------------- >>>>>>> For information about J forums see http://www.jsoftware.com/forums.htm >>>>>> ---------------------------------------------------------------------- >>>>>> For information about J forums see http://www.jsoftware.com/forums.htm >>>>> ---------------------------------------------------------------------- >>>>> For information about J forums see http://www.jsoftware.com/forums.htm >>>> ---------------------------------------------------------------------- >>>> For information about J forums see http://www.jsoftware.com/forums.htm >>> ---------------------------------------------------------------------- >>> For information about J forums see http://www.jsoftware.com/forums.htm >> ---------------------------------------------------------------------- >> For information about J forums see http://www.jsoftware.com/forums.htm > ---------------------------------------------------------------------- > For information about J forums see http://www.jsoftware.com/forums.htm ---------------------------------------------------------------------- For information about J forums see http://www.jsoftware.com/forums.htm
