Re: xlive: add initial stub dll (1/3)

2010-11-03 Thread Stefan Dösinger

Am 03.11.2010 um 06:00 schrieb Scott Ritchie:
 From a usability perspective we're not going to be doing much good
 unless this happens near-automatically with a typical Wine installation.
 That means either including them in Wine directly or including hooks to
 download them when needed (this could also be done in the packaging layer)
Another thing I wondered about: Can those DLLs be compiled outside the main 
wine tree? It'd seem nicer to me to have a project with just the DLL as opposed 
to a clone of the entire wine tree.

As for the server parts of xlive, I guess to get them working with a builtin 
DLL we'd have to build our own server infrastructure. Even then it would be an 
entirely different set of users and data than what you get on Windows.





Re: xlive: add initial stub dll (1/3)

2010-11-02 Thread Alexandre Julliard
Austin English austinengl...@gmail.com writes:

 This dll is not a core part of windows (at least, not yet), but I
 think it should be considered for inclusion in Wine. A bit of
 explanation is necessary:

 xlive.dll comes from Games for Windows (1,2), whose installer depends
 on .Net 3.5 (can be skipped with the /nodotnet parameter). Native
 fails on wine, however, unless a native msasn1.dll is provided,
 because xlive.dll is digitally signed (so implementing our own
 msasn1.dll won't help). As it currently stands, users can't play any
 'Game for Windows' that doesn't have a Windows license.

That's not a good enough reason, particularly considering how ugly the
resulting code is. And it seems unlikely that this is ever going to move
beyond the nasty hack stage, given the lack of documentation.

-- 
Alexandre Julliard
julli...@winehq.org




Re: xlive: add initial stub dll (1/3)

2010-11-02 Thread Austin English
On Tue, Nov 2, 2010 at 12:34 PM, Alexandre Julliard julli...@winehq.org wrote:
 Austin English austinengl...@gmail.com writes:

 This dll is not a core part of windows (at least, not yet), but I
 think it should be considered for inclusion in Wine. A bit of
 explanation is necessary:

 xlive.dll comes from Games for Windows (1,2), whose installer depends
 on .Net 3.5 (can be skipped with the /nodotnet parameter). Native
 fails on wine, however, unless a native msasn1.dll is provided,
 because xlive.dll is digitally signed (so implementing our own
 msasn1.dll won't help). As it currently stands, users can't play any
 'Game for Windows' that doesn't have a Windows license.

 That's not a good enough reason, particularly considering how ugly the
 resulting code is. And it seems unlikely that this is ever going to move
 beyond the nasty hack stage, given the lack of documentation.

Fair enough. You never know until you try and have the code in hand.

For those interested, I've put an initial fork at
http://github.com/austin987/wine/ . If you've got any games that need
xlive, please test against it and report any bugs to me directly or at
http://bugs.winehq.org/show_bug.cgi?id=23532. I plan to expand the
tests next, to try and get some documentation, so that it can possibly
be implemented in the future in a clean way (by someone else, if need
be).

-- 
-Austin




Re: xlive: add initial stub dll (1/3)

2010-11-02 Thread André Hentschel
Am 02.11.2010 20:46, schrieb Austin English:
 On Tue, Nov 2, 2010 at 12:34 PM, Alexandre Julliard julli...@winehq.org 
 wrote:
 Austin English austinengl...@gmail.com writes:

 This dll is not a core part of windows (at least, not yet), but I
 think it should be considered for inclusion in Wine. A bit of
 explanation is necessary:

 xlive.dll comes from Games for Windows (1,2), whose installer depends
 on .Net 3.5 (can be skipped with the /nodotnet parameter). Native
 fails on wine, however, unless a native msasn1.dll is provided,
 because xlive.dll is digitally signed (so implementing our own
 msasn1.dll won't help). As it currently stands, users can't play any
 'Game for Windows' that doesn't have a Windows license.

 That's not a good enough reason, particularly considering how ugly the
 resulting code is. And it seems unlikely that this is ever going to move
 beyond the nasty hack stage, given the lack of documentation.
 
 Fair enough. You never know until you try and have the code in hand.
 
 For those interested, I've put an initial fork at
 http://github.com/austin987/wine/ . If you've got any games that need
 xlive, please test against it and report any bugs to me directly or at
 http://bugs.winehq.org/show_bug.cgi?id=23532. I plan to expand the
 tests next, to try and get some documentation, so that it can possibly
 be implemented in the future in a clean way (by someone else, if need
 be).
 

@All:
There are already some Projects which reimplement non-windows dlls like e.g. 
cuda.
I totally see the reason to not integrate them into Wine, but maybe we could 
start a small Project hosting all that stuff at _one_ point and not spread over 
the www.
Maybe my upcoming CE dlls will also fall into that catagory...
Then we also could pack the .dll.so's up into one nsis installer, where you can 
select which subproject you want to install.
(Maybe 32-bit and 64-bit? And possibly in some other way also for ARM?)
Your opinions?

-- 

Best Regards, André Hentschel




Re: xlive: add initial stub dll (1/3)

2010-11-02 Thread Scott Ritchie
On 11/02/2010 01:34 PM, André Hentschel wrote:
 Am 02.11.2010 20:46, schrieb Austin English:
 On Tue, Nov 2, 2010 at 12:34 PM, Alexandre Julliard julli...@winehq.org 
 wrote:
 Austin English austinengl...@gmail.com writes:

 This dll is not a core part of windows (at least, not yet), but I
 think it should be considered for inclusion in Wine. A bit of
 explanation is necessary:

 xlive.dll comes from Games for Windows (1,2), whose installer depends
 on .Net 3.5 (can be skipped with the /nodotnet parameter). Native
 fails on wine, however, unless a native msasn1.dll is provided,
 because xlive.dll is digitally signed (so implementing our own
 msasn1.dll won't help). As it currently stands, users can't play any
 'Game for Windows' that doesn't have a Windows license.

 That's not a good enough reason, particularly considering how ugly the
 resulting code is. And it seems unlikely that this is ever going to move
 beyond the nasty hack stage, given the lack of documentation.

 Fair enough. You never know until you try and have the code in hand.

 For those interested, I've put an initial fork at
 http://github.com/austin987/wine/ . If you've got any games that need
 xlive, please test against it and report any bugs to me directly or at
 http://bugs.winehq.org/show_bug.cgi?id=23532. I plan to expand the
 tests next, to try and get some documentation, so that it can possibly
 be implemented in the future in a clean way (by someone else, if need
 be).

 
 @All:
 There are already some Projects which reimplement non-windows dlls like e.g. 
 cuda.
 I totally see the reason to not integrate them into Wine, but maybe we could 
 start a small Project hosting all that stuff at _one_ point and not spread 
 over the www.
 Maybe my upcoming CE dlls will also fall into that catagory...
 Then we also could pack the .dll.so's up into one nsis installer, where you 
 can select which subproject you want to install.
 (Maybe 32-bit and 64-bit? And possibly in some other way also for ARM?)
 Your opinions?
 

From a usability perspective we're not going to be doing much good
unless this happens near-automatically with a typical Wine installation.
 That means either including them in Wine directly or including hooks to
download them when needed (this could also be done in the packaging layer)

-Scott Ritchie