I've found that when learning a new system, it is easy to find bugs in it - which turn out to be a misunderstanding on my part. Take it from someone with years experience doing nothing but APL, GNU APL is pretty solid. If you encounter a bug, definitely look closely at your code.
Glad to help. --blake On Tue, Mar 10, 2015 at 8:49 PM, Christian Robert < christian.rob...@polymtl.ca> wrote: > You are right. QuadNC do the right thing. > > btw: this is why I asked for some kind of a "newbie" list for gnu apl. I > don't want to bother the developpers at each time > *I think* of a problem in the software. 9/10+ are my own ignorance. > > thanks, and still learning from near zero. > > Xtian. > > > On 2015-03-10 21:09, Blake McBride wrote: > >> I think you are confusing ⎕NC and ⎕NL >> >> On Tue, Mar 10, 2015 at 7:16 PM, Christian Robert < >> christian.rob...@polymtl.ca <mailto:christian.rob...@polymtl.ca>> wrote: >> >> Hi, >> >> I'm pretty sure this was working a few days ago to test if a variable >> has a value or not at entry to a function. >> >> Is it broken or I miss something ? >> >> thanks, >> >> Xtian. >> >> >> ∇Time[⎕]∇ >> ∇ >> [0] z←ncpu Time what;⎕io;start;stop;day;old___ncpu >> [1] →(1=⎕NL 'ncpu')/already_set >> [2] ncpu←1 ⍝ the default >> [3] already_set: >> [4] ⎕io←0 >> [5] start←¯5↑⎕ts >> [6] day←start[0] >> [7] start ← 32 24 60 60 1000 ⊥ start >> [8] old_ncpu←⎕SYL [25;1] >> [9] ⍎what >> [10] stop←¯5↑⎕ts >> [11] ⎕SYL [25;1]← old_ncpu >> [12] →(stop[0]=day)/ok >> [13] stop[0]←stop[0]+1 >> [14] ok: >> [15] stop ← 32 24 60 60 1000 ⊥ stop >> [16] z←(⍕(stop-start)÷1000),' seconds at num core = >> ',(⍕⎕syl[25;1]),' of ', ⍕⎕syl[24;1] >> ∇ >> Time 'z←⍳1000' >> DOMAIN ERROR >> Time[1] →(1=⎕NL 'ncpu')/already_set >> ^ ^ >> >> >> >> >>
