apologies - there was a typo in that patch - should be looking for "://" rather than "::/" . Otherwise my questions are the same.
--- On Mon, 7/4/08, Hin-Tak Leung htl10@users.sourceforge.net wrote:
From: Hin-Tak Leung htl10@users.sourceforge.net Subject: a proposed fix for bug 11690 concerning hhctrl.ocx To: jacek@codeweavers.com, truiken@gmail.com Cc: wine-devel@winehq.org Date: Monday, 7 April, 2008, 7:02 PM Hi Jacek and James - according to git, you two are the ones last working on the part of code which I have a possible fix for bug 11690 (see the bug report for somebody tested okay something like this), so I'd like to see what you think of this patch, specially (and others can chip in as well...) Here is what I think:
NavigateToUrl() in hhctrl.ocx/help.c is only used by the HH_HELP_CONTEXT case in HtmpHelpW() in hhctrl.ocx/hhctrl.c
- I think in some circumstances, it may return a full url,
but in the case of bug 11690, it returns a "relative" one inside the same CHM, so the code in NavigateToUrl() should drop through to look up in chm. The top part of NavigateToUrl() is really wrong because it assumes a IWebbrowser_navigate would fail on a relative url and drop through, but the way I see it, Iwebbrower_navigate should return the status of whether talking to IE/gecko is successful, not whether they gets "404 not found".
I have a few questions for you two:
- is there any known case where HH_HELP_CONTEXT returns a
fully formed URL? (i.e. if this change is committed, it is going to break something else?)
- Since NavigateToUrl is only used once (and it needs to
drop through to NavigateToChm() in most cases, pending your answer question 1), wouldn't it make sense to just remove it?
Thanks, Hin-Tak
Yahoo! For Good helps you make a difference
___________________________________________________________ Yahoo! For Good helps you make a difference