I did all my javascript development the hard way (code staring and alert sprinkling). I only very recently started playing with javascript development/debug tools in FF, Chrome, and IE. I sure wish I'd invested in them in the beginning as it would have saved me lots of time. They are quite good and I am increasingly comfortable with html/css/javascript/jhs for good gui development/delivery.
On Tue, Jan 11, 2011 at 4:35 PM, David Mitchell <[email protected]>wrote: > You are welcome. It was an interesting introduction to the IE8 Developer > Tool > kit. I'll be using this more in the future, I suspect. > -- > > On 1/11/2011 16:25, Eric Iverson wrote: > > Thank you for finding this and forging through to a solution! > > > > On Tue, Jan 11, 2011 at 3:45 PM, David Mitchell<[email protected] > >wrote: > > > >> Yes, your suggested change fixed the problem. I no longer get the alert > >> box. > >> > >> Thanks. > >> -- > >> On 1/11/2011 15:08, Eric Iverson wrote: > >>> Thanks for the additional info. I am getting a glimmer of > understanding. > >> I > >>> see that the jsetcaret code for FF (and similar browsers) handles the > >> case > >>> where "caret" doesn't exist. I think I need to do the same for IE. I > >> think > >>> there is a race between the document completely settling down and the > >>> javascript manipulating it. > >>> > >>> If you are interested you could try editing utiljs to be as follows: > >>> > >>> *** utiljhs.ijs > >>> function jsetcaret(id,collapse) > >>> { > >>> var p= jbyid(id); > >>> if(null==p)return; > >>> if (window.getSelection) > >>> ... > >>> > >>> On Tue, Jan 11, 2011 at 2:11 PM, David Mitchell<[email protected] > >>> wrote: > >>> > >>>> I did not see the alert after I tried your test. All I see in IE is > the > >>>> JHS > >>>> menu bar and this text: > >>>> > >>>> WARNING: file read failed! > >>>> fubar > >>>> > >>>> When I look at the MS debugging info for the error I am getting with a > >> open > >>>> file, I see this call sequence: > >>>> > >>>> ev_body_load > >>>> color > >>>> jsetcaret > >>>> tst.moveToElementText(p); > >>>> > >>>> Where p is null. I suspect that p should have been set to a non-null > >> value > >>>> by > >>>> > >>>> var p= jbyid(id); > >>>> > >>>> I'm not sure why the call to jbyid(id), where id="caret" returned > null. > >>>> I'll > >>>> see if I can find out why "caret" is not available at this point. > >>>> -- > >>>> > >>>> On 1/11/2011 13:28, Eric Iverson wrote: > >>>>> David Mitchell: > >>>>> > >>>>> I don't understand and your extensive documentation just confuses me > >>>> more. > >>>>> > >>>>> Your final state seems to have the right text displayed and it is > >>>> colored. > >>>>> This would seem to indicate that the ev_body_load is 99% finished and > >> has > >>>>> done all the hard bits. Strange. > >>>>> > >>>>> My guess is that the failure has nothing to do with the fact that the > >>>> file > >>>>> is opened twice or in the switch from jfile to jijs. I'm guessing it > >> has > >>>>> something to do with the warning message. > >>>>> > >>>>> Have you ever seen the alert box except in the 'file already open' > >> case? > >>>> Do > >>>>> you ever see it of the ran or saved messages? > >>>>> > >>>>> Please try the following after a clean jhs start: > >>>>> browse to url: localhost:65001/jijs?mid=open&path=fubar > >>>>> > >>>>> Does this get the alert box? Try this in various ways to see if it > ever > >>>>> triggers the alert. > >> ---------------------------------------------------------------------- > >> 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
