On Thursday 29 June 2006 07:49, you wrote:
> what is the gain of autorun, while i dont have wine running? would not
> the common user expect, once she knows, that wine support autorun.inf,
> that she can insert a cd and away it goes? while the explorer.exe is
> not running permanently - e.g. via xinitrc - then i still would have to
> deal with starting my setup.exe by hand. i have my doubts, that wine is
> the correct place to implement this - but something like hotplugd or
> whatevernowadaysdoesthetrickwithcdroms should handle this?
One could think of Wine as a service, and start it with X if they wanted, so
they could then run Windows programs whenever they wanted with potentially
faster startup times, and with Windows "features" like autorun, and with a
persistant Windows state (eg. instead of being shut down when all Windows
programs close, it would stay around until Wine itself is explicitly
stopped). Or, like me currently, they could run a Windows desktop/taskbar
replacement in a Wine virtual desktop with managed windows off on a seperate
X desktop and pretend to have Windows running along side X.
I'd think if Wine were trying to "emulate" Windows as much as possible, it'd
be backwards to not have autorun capabilities. As I said before, autorun.inf
needs to be parsed anyway since it can define an icon to use for the drive,
so if you're going to automatically parse it to load an icon, why not also
ask (one time only) if the user wants to turn autorun on or off to run the
defined program?
Why not replace the require_once() with requires instead of putting it on the
todo?
We don't really need the $aClean TODO item since thats high on the priority
list at the moment and we'll just end up removing it in a day or two right?
The move of show_note() into its class is on my todo list for class cleanups
today or tomorrow as well.
Basically I think we can do without the todo notes for now since most of it is
in the process of being fixed. The rest of the comments look good though.
Chris
On Thursday 29 June 2006 9:43 am, Jonathan Ernst wrote:
> You might need to apply the previous patches first.
>
> Changelog:
> - update the comments in some files
(FYI: I have no internet connection at the moment, so I say 'hi' to
all Wine developers!)
Wine is nortorious for NOT supporting Autorun. Sure, it's stupid to
load the autorun program right after the disk is inserted, but who
wants to have to parse the autorun.inf by hand?
So I wrote this little program. You can scan a CD for viruses first,
then run this to emulate the Windows autorun feature. All you tell it
is the drive letter of the CD, and it parses autorun.inf, and runs the
listed program. It also check to make sure the file is a valid
autorun.inf.
It is not complete, and it is pure standard C. No Win32-specific
functions are used. It uses several functions that aren't in the
msvcrt, so I have included the portable snprintf library (version 2.1)
and some portablized code from glibc. The tarball includes a prebuilt
binary for your convience, but i strongly suggest you recompile the
program from source (emerge xmingw on gentoo, and you can use my
Makefile unmodified!)
On Friday 30 June 2006 07:36, James Hawkins wrote:
> + lstrcatW(cache_path, backslash);
> + lstrcatW(cache_path, content);
> +
> + if (!GetTempFileNameW(cache_path, prefix, 0, cache_path))
> + return E_FAIL;
> +
> + if (lstrlenW(cache_path) > dwBufLength)
> + return E_INVALIDARG;
> +
> + lstrcpyW(szFileName, cache_path);
> +
> + return URLDownloadToFileW(lpUnkCaller, szURL, szFileName, 0, pBSC);
This is not the correct behaviour - doing this will result in the cache
directories filling up with files that are not mentioned in the cache index.
You should be calling CreateUrlCacheEntryW to get the file name for the new
file.
--
Troy Rollo - wine(a)troy.rollo.name
Patch looks good and although it is large like you said it isn't something we
can do much about at this point.
It unfortunately doesn't apply cleanly against cvs head.
If you could submit all of the smaller cleanup patches ahead of big patches
like this that would be the best. Its more likely that larger patches like
this will require changes than simple renames or refactoring and getting
those out of the way means neither of us have to track which patches have
gone in vs. need to be rediffed against cvs head.
I'll usually make a single change in a single checked out appdb directory.
That way I'm always diffing against cvs head and I can avoid having patches
that depend on each other. I'm concerned now if you've made further changes
to the directory where this patch was generated that some fixes will be
missed or unwanted changes will sneak in.
Lets get all of the rest of the isolated cleanups that you've submitted into
the appdb and then lets give this patch another shot.
Chris
On Thursday 29 June 2006 4:59 am, Jonathan Ernst wrote:
> Before you ask, there was no possibility to split this one up because
> everything is interdependant.
>
> Changelog:
> - rename all Get/Post/Cookies variables using type prefixes so that we can
> later use the autofiltering function
Patch doesn't apply against CVS head.
Chris
On Thursday 29 June 2006 7:05 am, Jonathan Ernst wrote:
> You might need to apply the previous patches first.
>
> Changelog:
> - factor exit calls after redirect
Hi,
On 6/28/06, Maarten Lankhorst <M.B.Lankhorst(a)gmail.com> wrote:
>
> --- tools/wine.inf.ori 2006-06-28 18:32:28.000000000 +0200
> +++ tools/wine.inf 2006-06-28 18:35:22.000000000 +0200
> @@ -2467,3 +2467,6 @@
> HKLM,%Control%\ProductOptions,"ProductType",2,"WinNT"
> HKLM,%Control%\Windows,"CSDVersion",0x10003,0x400
> HKLM,%Control%\Session Manager\Environment,"OS",2,"Windows_NT"
> +HKLM,"Software\Microsoft\Internet Explorer","Version",,"6.0.2900.2180"
> +HKLM,"Software\Microsoft\Internet Explorer","W2kVersion",,"6.0.2900.2180"
> +HKLM,"Software\Microsoft\Internet Exlporer","Build",,"62900"
Does installing IE6 in win98 mode still work after this? I think its
going to have problems. We really should report IE 5 so the IE 5.5 and
6 installers dont get confused.
--
Steven Edwards
"There is one thing stronger than all the armies in the world, and
that is an idea whose time has come." - Victor Hugo
We'll want to perform this move of error_page() and the renaming of calls to
it in a single patch.
Chris
On Wednesday 28 June 2006 1:21 pm, Jonathan Ernst wrote:
> Changelog:
> - new util_show_error_page that will replace errorpage