Thanks, Bill, for keeping a record of the changes. (I neglected to do so.) I applaud the major effort you've put in to getting jgtk-wd working, and TABULA with it. I also feel it's important to aim at one single script which works under both j6 and j7, and more importantly on both Mac and Windows, and I am glad you seem to think the same way.
I'm not in a position to say whether any one of the changes listed below are still needed or not, on one of the many platforms. But I can use your list of changes to guess at what features need re-testing. I fear, however, that lacking an up-to-date Windows machine, I can't test on anything later than Win2000. I shall test on the Mac under Snow Leopard, and sit back and leave it to others to report and fix bugs under Windows, particularly 64-bit, which is one of my permanent blind-spots. On Fri, Apr 27, 2012 at 4:52 AM, bill lam <bbill....@gmail.com> wrote: > Hi Ian, > > the effect of closeok is what you has observed. It should be > artefacts for workaround some bugs in gtkwd at that time. Please > revert it. Other changes from me are list below, please review > if they are also reduncdant. > > 4) require 'gtkwd' NB. for: J7 wd > 5) require 'strings' NB. for: rplc > 6) require 'gl2' NB. load gl2 definitions in jgl2 locale > 7) > 8) coclass 'tab' > 89) pc tab closeok; > 489) consts_cappend_button=: newc > 490) consts_casec_button=: empty > 492) consts_cons_button=: newc > 493) consts_cons_select=: empty > 508) consts_searchc_button=: fillconsts > 520) if. -.IFJ6 do. > 521) NB. flip if 255 is red > 522) z=. fliprgb^:(-.RGBSEQ_j_) z > 523) end. > 650) functs_casef_button=: empty > 653) functs_fappend_button=: newf > 654) functs_func_button=: newf > 655) functs_func_select=: empty > 669) functs_searchf_button=: fillfuncts > 1189) if. IFJ6 do. > 1190) z=. z,LF,LF,LF > 1191) z=. z,tabengine 'CTBU' > 1192) ijsstr z > 1193) else. > 1194) require 'print' > 1195) if.''-.@-: wd 'mbprinter' do. print_jprint_ z end. > 1196) end. > 1481) ttable_run'' > 1484) inf_run'' > 1487) wd^:(-.IFJ6) 'pshow;pshow sw_hide' > 1495) wd^:(-.IFJ6) 'pshow;pshow sw_hide' > 1502) wd 'pshow' > 1503) evtloop^:(-.IFJ6)'' > 1552) 3 : 0'' > 1553) if. -.IFJ6 do. > 1557) end. > 1558) '' > 1559) ) > 1642) ttable_calco_button=: calcmd > 1645) ttable_panel_button=: clickpanel > 1646) ttable_panel_select=: clickpanel > 1647) ttable_preci_select=: setpreci > 1662) ttable_xunit_button=: empty > 1663) ttable_xunit_select=: pickunits > 1719) > 1720) 3 : 0'' > 1721) if. IFJ6 do. > 1722) wd=: wd_probed > 1723) end. > 1724) '' > 1725) ) > 1729) 11!:0 y > 1738) if. IFWIN +. IFWINNT do. y rplc SL ; BS end. > 1745) wd`(11!:0)@.IFJ6 'psel tab; pclose;' > > Птн, 27 Апр 2012, Ian Clark писал(а): >> @Bill >> >> I can't understand the purpose of 'closeok' in the definition of a jwd >> form. It has appeared recently in the definition of form: 'tab' in >> TABULA (contained in a noun: TABU). It looks like this: >> >> pc tab closeok; >> >> When I remove closeok, giving: >> >> pc tab; >> >> clicking the go-away box (on the Mac) runs the handler: tab_close, as >> it's meant to. But with closeok, tab_close is not run, and TABULA does >> not remember its screen position and other settings properly. >> >> Why is closeok needed? (I assume it was you that put it there). >> >> JVERSION >> Installer: j602a_mac_intel.dmg >> Engine: j602/2008-03-03/16:45 >> Library: 6.02.062 >> ---------------------------------------------------------------------- >> For information about J forums see http://www.jsoftware.com/forums.htm > > -- > regards, > ==================================================== > GPG key 1024D/4434BAB3 2008-08-24 > gpg --keyserver subkeys.pgp.net --recv-keys 4434BAB3 > ---------------------------------------------------------------------- > For information about J forums see http://www.jsoftware.com/forums.htm ---------------------------------------------------------------------- For information about J forums see http://www.jsoftware.com/forums.htm
