This is follow up to two earlier postings by me.  I would have replied to
the thread(s) but my digest does not seem to be arriving.  The earlier links
are
http://marc.theaimsgroup.com/?l=php-general&m=98399367205538&w=2
http://marc.theaimsgroup.com/?l=php-general&m=98391734022011&w=2

I believe I have tracked down the nasty heisenbug, but do not know how to
squash it.

I initially had a table with ~60 rows and five columns.  Two of the columns
are simple, the other three are forms with 4 hidden variables and a submit
button.  Initially the table was one long string and about 2/3 the way
through it, the string would terminate prematurely.  I split the string to
an array placing each row of the table in a separate cell.  This changed
where the errors occurred, but did not fix the problem.  I then split the
table into multiple smaller tables with each table containing 5 rows.  Once
again it changes the location of the errors, but does not fix anything.
Throughout all this the page is about 45Kb.  If the  row where the error
occurred is echo'd with htmlspecialshars the row looks correct(and the error
moved to the line before).  When I echo'd the line before the initially
observed error I can view both the source html and the rendered version.
The row of the array is correct, but the sent source is truncated.

The error does not occur on a development machine (PIII700 w/ 128MB) running
RH7/php4/apache, but does occur on the production server PI133 w/ 64 MB)
RH5.2/php4/AOLServer.

Can anyone suggest the cause of this bug and a possible fix?  Has anyone
heard of a AOLServer truncating data from php?  Does HTTP sometimes truncate
packets in weird ways?

Eric Nielsen


-- 
PHP General Mailing List (http://www.php.net/)
To unsubscribe, e-mail: [EMAIL PROTECTED]
For additional commands, e-mail: [EMAIL PROTECTED]
To contact the list administrators, e-mail: [EMAIL PROTECTED]

Reply via email to