LB Booster
General >> General Board >> debug-yes;regular-no
http://lbb.conforums.com/index.cgi?board=general&action=display&num=1512325567
debug-yes;regular-no
Post by Alincon on Dec 3rd, 2017, 5:26pm

Is it possible in LBB to have a program run differently in debug mode than in regular mode - like failing to add a record to a random file?

I have posted a message similar to this one, and some of the code on the 'other' forum.

r.m.
Re: debug-yes;regular-no
Post by Alincon on Dec 3rd, 2017, 9:36pm

I should have noted that my program works as intended in LBB, regardless of mode. (yea!)

r.m.
Re: debug-yes;regular-no
Post by Richard Russell on Dec 3rd, 2017, 10:31pm

on Dec 3rd, 2017, 5:26pm, Alincon wrote:
Is it possible in LBB to have a program run differently in debug mode than in regular mode - like failing to add a record to a random file?

The only (intentional!) differences between LBB's 'normal' and 'debug' modes are as follows:
  1. 'Temporary' breakpoints (inserted by double-clicking in the left margin) cause the program to pause at that statement in 'debug' mode, but not in 'normal' mode.

  2. A 'permanent' breakpoint (i.e. a TRACE 2 statement) causes the program to pause when in 'debug' mode but has no effect - other than a slight impact on execution time - when in 'normal' mode.

  3. The NOMAINWIN statement is ignored in 'debug' mode, causing the mainwin always to be displayed.
Does that answer your question?

Richard.
Re: debug-yes;regular-no
Post by Alincon on Dec 5th, 2017, 7:19pm

(As frequently happens) the problem was really mine: I did not close the random file before closing the program. In my defense, I got no error message in LB. In LBB there was no message either, but it worked.
Evidently, LBB handles unclosed files differently.

r.m.
Re: debug-yes;regular-no
Post by Richard Russell on Dec 5th, 2017, 7:50pm

on Dec 5th, 2017, 7:19pm, Alincon wrote:
Evidently, LBB handles unclosed files differently.

LBB should automatically close files on exit; I would have thought that was normal behavior. If LB 4 doesn't work that way I am surprised, but I suppose you should not have assumed it would.

Richard.
Re: debug-yes;regular-no
Post by Rod on Dec 6th, 2017, 07:09am

The problem was a little different to that described. The program did not END, it simply closed its window. I have not tested but I sssume LBB might behave the same way as LB. it hangs quietly in the background doing what it was instructed to do.
Re: debug-yes;regular-no
Post by Richard Russell on Dec 6th, 2017, 08:30am

on Dec 6th, 2017, 07:09am, Rod wrote:
The problem was a little different to that described. The program did not END, it simply closed its window.

Ah, that is indeed very different. In that case the file will remain open indefinitely and there is no guarantee that data written with a PRINT# or PUT# will have been 'flushed' to the actual storage device (and thus be visible to, for example, another program looking at the same file).

In fact I might expect that LBB would, if anything, be more vulnerable to that possibility, because for performance reasons it stores output data in a temporary buffer which gets written only when full (or when the file is closed).

If you need to commit data to disk without closing the file there is a 'trick' you can use in LBB:

Code:
    SEEK #file, LOC(#file)

Whilst you might expect that setting the file pointer to its current position would do nothing, it has the side-effect of flushing LBB's buffers.

Richard.