298622 - [bfcache on] find fails on page with url typed

Status: RESOLVED FIXED

(Core :: DOM: Navigation, defect)

Description

[Peter van der Woude [:Peter6]]

Mozilla/5.0 (Windows; U; Windows NT 5.0; en-US; rv:1.8b2) Gecko/20050623
Firefox/1.0+ ID:2005062312

repro:
1.Open Firefox
2.Load a random page in the first tab (via bookmarks or type in locationbar)
3.Now type (! important) any random address , say http://www.mozilla.org/ in the
locationbar.
4.press ctrl+f and type a word which you seen on that page in the findbar.

result:
nothing can be found

I only have been able to reproduce using 1 tab in a window.
With multiple tabs open it allways worked

Comment 1

[Peter van der Woude [:Peter6]]

in step 3 you need to press enter to load the page

tested with old profile / -safe-mode / new profile+dir same result 

Comment 2

[Jorge Villalobos [:jorgev] (he/him)]

Mozilla/5.0 (Windows; U; Windows NT 5.1; en-US; rv:1.8b2) Gecko/20050623
Firefox/1.0+
WFM

Comment 3

[Chase Phillips]

FYI I filed bug 298639 which may be describing the same issue, though I see this
problem with about:blank being the first page Firefox opens to.

Peter, does this bug (bug 298622) occur for you if about:blank is the first page
the browser opens to?  If so, we should remove the "if first page was not
about:blank" conditional from the summary.

Also, did you mean to type "find fails on page with url typed" instead of "find
fails one page with url typed"?

Comment 4

[Peter van der Woude [:Peter6]]

*** Bug 298639 has been marked as a duplicate of this bug. ***

Comment 5

[Peter van der Woude [:Peter6]]

(In reply to comment #3)
> FYI I filed bug 298639 which may be describing the same issue, though I see this
> problem with about:blank being the first page Firefox opens to.
> 
> Peter, does this bug (298622) occur for you if about:blank is the first page
> the browser opens to? 

- The browser opens with about:blank.
- If I type about:blank for step 2 it does not appear

> If so, we should remove the "if first page was not
> about:blank" conditional from the summary.
- Yep

> Also, did you mean to type "find fails on page with url typed" instead of "find
> fails one page with url typed"?
- Yep

Comment 6

[Ria Klaassen (not reading all bugmail)]

I tried this (anxious as always to find a cause and a cure for everything) with
the addresses www.mozilla.org and www.mozillazine.org and at least here it seems
to depend on the pref: user_pref("browser.sessionhistory.max_viewers", 5);

Comment 7

[Peter van der Woude [:Peter6]]

(In reply to comment #6)
> I tried this (anxious as always to find a cause and a cure for everything) with
> the addresses www.mozilla.org and www.mozillazine.org and at least here it seems
> to depend on the pref: user_pref("browser.sessionhistory.max_viewers", 5);
> 
Good catch
confirming
the bug appears with browser.sessionhistory.max_viewers !=0
the bug doesn't appear with browser.sessionhistory.max_viewers =0

->Core
->history session
cc bryner

Comment 8

[Brian Ryner (not reading)]

Does anyone know if this regressed recently, or has it always been a problem
with fastback turned on?

Comment 9

[Peter van der Woude [:Peter6]]

(In reply to comment #8)
> Does anyone know if this regressed recently, or has it always been a problem
> with fastback turned on?

works in Mozilla/5.0 (Windows; U; Windows NT 5.0; en-US; rv:1.8b2)
Gecko/20050620 Firefox/1.0+ ID:2005062014 (release 1431pdt)

fails in Mozilla/5.0 (Windows; U; Windows NT 5.0; en-US; rv:1.8b2)
Gecko/20050620 Firefox/1.0+ ID:2005062022 (release 2319pdt)

checkins approximately :
http://bonsai.mozilla.org/cvsquery.cgi?treeid=default&module=PhoenixTinderbox&branch=HEAD&branchtype=match&filetype=match&whotype=match&sortby=Date&hours=2&date=explicit&mindate=2005-06-20+13%3A30%3A00&maxdate=2005-06-20+23%3A00%3A00&cvsroot=%2Fcvsroot

Comment 10

[Brian Ryner (not reading)]

Attachment: patch

This is a regression from bug 297173.  Typeahead find relies on the pres shell
destruction as a signal that it should no longer be used for find operations. 
With the cache on, of course, the pres shell stays around after you leave the
page.  I made it check that the shell is actually hooked up to a DOM window as
well.

Comment 11

[David Baron :dbaron: (⌚️UTC-4, no longer working on Mozilla)]

Comment on attachment 187456 [details] [diff] [review]
patch

>-      *aResult = hasWrapped ? FIND_WRAPPED : FIND_FOUND;
>+      if (hasWrapped)
>+        *aResult = FIND_WRAPPED;
>+      else
>+        *aResult = FIND_FOUND;

Well, I like it the way it was :-)

r+sr=dbaron if you've tested that TAF is OK when you've fast-backed to a page
as well (both if you have used it before leaving the first time and if you
haven't)

Comment 12

[Brian Ryner (not reading)]

(In reply to comment #11)
> (From update of attachment 187456 [details] [diff] [review] [edit])
> >-      *aResult = hasWrapped ? FIND_WRAPPED : FIND_FOUND;
> >+      if (hasWrapped)
> >+        *aResult = FIND_WRAPPED;
> >+      else
> >+        *aResult = FIND_FOUND;
> 
> Well, I like it the way it was :-)

Twas a warning fix.

> 
> r+sr=dbaron if you've tested that TAF is OK when you've fast-backed to a page
> as well (both if you have used it before leaving the first time and if you
> haven't)
> 

Comment 13

[Brian Ryner (not reading)]

Comment on attachment 187456 [details] [diff] [review]
patch

requesting approval

Comment 14

[Chase Phillips]

Comment on attachment 187456 [details] [diff] [review]
patch

a=chase

Comment 15

[Brian Ryner (not reading)]

checked in

Comment 16

[Jonathan Griffin (:jgriffin)]

Attachment: mochitest test case

Comment 17

[Boris Zbarsky [:bzbarsky]]

Comment on attachment 387563 [details] [diff] [review]
mochitest test case

This doesn't seem to be testing what this bug was about, which was find on the page that's loaded when you load your dummy page...

It's probably best to test find on both pages in this test, just to cover all the bases.

Comment 18

[Jonathan Griffin (:jgriffin)]

Attachment: mochitest test case

Thanks for the review Boris, and for catching the fact that the test was testing Find on the wrong page.  It now tests Find on both pages just for good measure.

Comment 19

[Boris Zbarsky [:bzbarsky]]

Pushed test as http://hg.mozilla.org/mozilla-central/rev/00bd4f026a53