Re: Crash-a-lot just became critical

On Sun, Jan 11, 2015 at 4:53 PM, Peter Haworth wrote: > Since runtime errors in modal dialogs never show up, you might try opening > it in any other mode and put a breakpoint right after the open command to > stop your code continuing. > Tried that, just to get at the debugger. Never crashes, a

Re: Crash-a-lot just became critical

Since runtime errors in modal dialogs never show up, you might try opening it in any other mode and put a breakpoint right after the open command to stop your code continuing. That way, if there is an runtime error somewhere in the modal stack scripts, it will show up as a runtime error. It doesn

Re: Crash-a-lot just became critical

On Fri, Jan 9, 2015 at 10:34 AM, Dr. Hawkins wrote: > > > I remembered an address they have; livecode has a crashing stack. > I have it tracked down. It seems to occur between my code calling for a stack to open and the preOpenStack handler for that stack: I can't do any logging *within* live

Re: Crash-a-lot just became critical

On Fri, Jan 9, 2015 at 11:57 AM, J. Landman Gay wrote: > Thanks so much. You've probably saved the rest of us a bunch of headache > Now if only they'll fix it fast enough to end mine . . . :) -- Dr. Richard E. Hawkins, Esq. (702) 508-8462 ___ use

Re: Crash-a-lot just became critical

On Fri, Jan 9, 2015 at 11:55 AM, J. Landman Gay wrote: > But you're right that problems would occur with a card or stack level > handler that assumes a mouseEnter was always triggered by a control. > Right. Pretty sure I have some card level mouseEnter handlers that check the mouseControl and a

Re: Crash-a-lot just became critical

On 1/9/2015 12:34 PM, Dr. Hawkins wrote: I remembered an address they have; livecode has a crashing stack. Thanks so much. You've probably saved the rest of us a bunch of headaches. -- Jacqueline Landman Gay | jac...@hyperactivesw.com HyperActive Software | http://www

Re: Crash-a-lot just became critical

On 1/9/2015 12:41 PM, Devin Asay wrote: >It looks like that change actually happened in LC 6.7. In 6.5 no mouseEnter is sent if there is no control object under the mouse. Starting in 6.7 a mouseEnter is sent to the card even if there is no control object under the mouse. Is this possibly an e

Re: Crash-a-lot just became critical

On Jan 9, 2015, at 11:30 AM, Devin Asay wrote: > > On Jan 9, 2015, at 11:03 AM, J. Landman Gay wrote: > >> On 1/9/2015 8:02 AM, Peter M. Brigham wrote: >>> Are people saying that in some newer versions of LC the mouseEnter >>> message is not sent to the control being entered but rather direct

Re: Crash-a-lot just became critical

On Fri, Jan 9, 2015 at 10:30 AM, Dr. Hawkins wrote: > It must still contain "garbage", so I need to send it somehow, but not > publicly post (it could have my master key in it, which would be > catastrophic). I remembered an address they have; livecode has a crashing stack. -- Dr. Richard E.

Re: Crash-a-lot just became critical

On Jan 9, 2015, at 11:03 AM, J. Landman Gay wrote: > On 1/9/2015 8:02 AM, Peter M. Brigham wrote: >> Are people saying that in some newer versions of LC the mouseEnter >> message is not sent to the control being entered but rather directly >> to the stack? > > No, that part behaves as it always

Re: Crash-a-lot just became critical

On Thu, Jan 8, 2015 at 10:21 AM, J. Landman Gay wrote: > I understand. But they can't fix it unless they can reproduce it. So you > can either trust them or live with it until/unless someone else has the > same issue and submits an example. The team is very keen to fix these > issues, and crashe

Re: Crash-a-lot just became critical

On Fri, Jan 9, 2015 at 6:02 AM, Peter M. Brigham wrote: > 'm not sure I understand. Are people saying that in some newer versions of > LC the mouseEnter message is not sent to the control being entered but > rather directly to the stack? If this is so, what happens to the target > function? Does

Re: Crash-a-lot just became critical

That is definitely different. I won't need to put a background grc or image (or whatever) to catch the mouseenter anymore, but I can see it causing weird stuff. What about empty space in groups? Just checked, they still behave the old way. Enter a group, but not over a control, and no mouseente

Re: Crash-a-lot just became critical

On 1/9/2015 8:02 AM, Peter M. Brigham wrote: Are people saying that in some newer versions of LC the mouseEnter message is not sent to the control being entered but rather directly to the stack? No, that part behaves as it always did. Controls receive mouseEnter normally. Previous versions of

Re: Crash-a-lot just became critical

Just did a test with 7.0.1, and things seem to work as I’d expect. If I put this into a button, card, and stack: on mouseenter answer “button” — or “card" or “stack” end mouseenter the answer shows up that I would expect. If I do this: on mouseenter answer “button” — or “card or “stack” p

Re: Crash-a-lot just became critical

I'm not sure I understand. Are people saying that in some newer versions of LC the mouseEnter message is not sent to the control being entered but rather directly to the stack? If this is so, what happens to the target function? Does it still return the control the mouse is entering? I'm still u

Re: Crash-a-lot just became critical

It fails on Mavericks too with versions of LiveCode before 7. I didn't know it ever worked on other platforms, actually. On January 9, 2015 12:44:20 AM CST, Mark Wieder wrote: >Jacque- > >Thursday, January 8, 2015, 7:06:00 PM, you wrote: > >> I think he means, he relies on the message being sen

Re: Crash-a-lot just became critical

Jacque- Thursday, January 8, 2015, 7:06:00 PM, you wrote: > I think he means, he relies on the message being sent to the stack where > there is no other control present. LC 7 sends it, but in all previous > versions, mouseEnter isn't sent directly to cards or stacks, only to > controls on the car

Re: Crash-a-lot just became critical

On 1/8/2015 8:16 PM, Mark Wieder wrote: Richard- Thursday, January 8, 2015, 1:50:34 PM, you wrote: There are an annoying number of apparently undocumented changes in the new engine with regard to message handling--I had apparently been relying on mouseEnter being sent to a stack itself, which

Re: Crash-a-lot just became critical

Richard- Thursday, January 8, 2015, 1:50:34 PM, you wrote: > There are an annoying number of apparently undocumented changes in the new > engine with regard to message handling--I had apparently been relying on > mouseEnter being sent to a stack itself, which 5.x does not do. I've seen > and for

Re: Crash-a-lot just became critical

On Thu, Jan 8, 2015 at 2:04 PM, Richard Gaskin wrote: > Dr. Hawkins wrote: > > I've been living through the frequent random crashes... >> > > Computers are deterministic systems: consistent inputs will yield > consistent results. > So they would like us to believe :) > > If something may appe

Re: Crash-a-lot just became critical

On Thu, Jan 8, 2015 at 1:50 PM, Dr. Hawkins wrote: > It appears that it will, only sometimes and randomly, happen when > referencing an object that doesn't exist in a stack that was opened as a > palette. When I changed how the stack was opened (stack command instead of > palette), normal errors

Re: Crash-a-lot just became critical

Dr. Hawkins wrote: I've been living through the frequent random crashes... Computers are deterministic systems: consistent inputs will yield consistent results. If something may appear to be random, it merely suggests we haven't found the recipe for reproducing it yet. All problems with

Re: Crash-a-lot just became critical

On Thu, Jan 8, 2015 at 1:11 PM, Colin Holgate wrote: > Is it really the case that LiveCode does no logging to Console at all? > I'm not sure where those would go when launched from a GUI. But all I get is "application quit unexpectedly" > > That aside, would a screensharing session with Edinb

Re: Crash-a-lot just became critical

Is it really the case that LiveCode does no logging to Console at all? That aside, would a screensharing session with Edinburgh have any value? Them watching it crash might lead to some ideas, without needing the source code. > On Jan 8, 2015, at 1:21 PM, J. Landman Gay wrote: > > On 1/8/2015

Re: Crash-a-lot just became critical

You should tell all this to the Engineers, not to the list. No one here caused this or can do anything about it. Bob S On Jan 8, 2015, at 08:41 , Dr. Hawkins mailto:doch...@gmail.com>> wrote: I've been living through the frequent random crashes, believing they are an IDE issue. Apparently, t

Re: Crash-a-lot just became critical

On 1/8/2015 10:41 AM, Dr. Hawkins wrote: It is also unacceptable to expect me to turn over source code with over $100k of development time to be able to try to track down these crashes. I understand. But they can't fix it unless they can reproduce it. So you can either trust them or live with

Re: Crash-a-lot just became critical

FIled as bug 14350 -- Dr. Richard E. Hawkins, Esq. (702) 508-8462 ___ use-livecode mailing list use-livecode@lists.runrev.com Please visit this url to subscribe, unsubscribe and manage your subscription