Voting for bugs (Was: Re: [Bug 20969])
Can/should voting for bugs be disabled if it is 'useless and does nothing except adding noise'? Thanks Tom On Tue, Nov 2, 2010 at 12:01 PM, wine-b...@winehq.org wrote: http://bugs.winehq.org/show_bug.cgi?id=20969 --- Comment #19 from Dmitry Timoshkov dmi...@codeweavers.com 2010-11-02 12:01:23 CDT --- (In reply to comment #18) Those of you with this issue, make sure you vote for this in bugzilla so that it can be confirmed, if you haven't already. Voting for bugs in Wine bugzilla is useless and does nothing except adding noise. -- Configure bugmail: http://bugs.winehq.org/userprefs.cgi?tab=email Do not reply to this email, post in Bugzilla using the above URL to reply. --- You are receiving this mail because: --- You are on the CC list for the bug.
Re: Linux kernel and game performance?
For the first reference, I have a WoW install. I haven't seen fps below 60 except in areas of the game where there are a lot of other players, but that was under OpenGL back before the latest patch. I haven't tried since Blizzard has disabled all high end settings in OpenGL mode for both Windows and Linux, and I've heard nothing but bad performance on D3D recently. I could give it a shot though. What is needed to make a kernel realtime other than adding CONFIG_SCHED_DEBUG and adding the sysctl settings? Thanks Tom On Mon, Oct 25, 2010 at 6:26 PM, Dan Kegel d...@kegel.com wrote: Hey folks, I've run into two web sites that claim that the Linux kernel causes performance problems in particular games (see below). Anybody know of others? And has anybody found concrete improvements in performance of a particular app (other than an audio workstation app) from using a realtime kernel? Thanks! - Dan First: http://wiki.archlinux.org/index.php/World_of_Warcraft#Kernel_Timing_Bug says in a section dated September 2008: If you are having problems with choppy video every 15 seconds or so, it is related to the kernel scheduler... to fix, add CONFIG_SCHED_DEBUG=y to your kernel config, then set kernel.sched_features=21 kernel.sched_batch_wakeup_granularity_ns=2500 kernel.sched_min_granularity_ns=400 in /etc/sysctl.cfg. Yikes. Any truth to that rumor? Second: http://hisouten.koumakan.jp/wiki/Linux_support#Resolved_bugs says The game runs too slowly Symptoms: Instead of running at about 60-62fps, like the game is supposed to, it'll run closer to 53fps. This is not ideal. The bug: This is a Linux timing issue. The game runs a secondary timing thread with THREAD_PRIORITY_TIME_CRITICAL, where it simply sleeps for 16ms and sends events to the main thread to tell it that a new frame is needed. On Linux the necessary timing accuracy is not available, so it wavers between 16ms and 20ms. The fix: I hacked around this by setting the timer period to 14ms. This leads to a steady 62-63fps. Which is close enough for use, really. For a constant 60fps turn on vsync in your video drivers.
Re: USB Device Support
On Wed, Sep 22, 2010 at 2:26 AM, Damjan Jovanovic damjan@gmail.comwrote: On Wed, Sep 22, 2010 at 1:52 AM, Tom Spear speeddy...@gmail.com wrote: On Tue, Sep 21, 2010 at 1:41 PM, Damjan Jovanovic damjan@gmail.com wrote: On Tue, Sep 21, 2010 at 8:07 PM, Tom Spear speeddy...@gmail.com wrote: On Tue, Sep 21, 2010 at 11:04 AM, Damjan Jovanovic damjan@gmail.com wrote: On Tue, Sep 21, 2010 at 5:04 PM, Tom Spear speeddy...@gmail.com wrote: Attached is the lsusb -v output, trimmed to only include the pedometer's info. I have many USB devices, so I didn't want to leave you to sort through a bunch of useless info. I don't have the webcam with me at the moment, but I will see if I can find it when I am at home soon. Thanks Tom On Tue, Sep 21, 2010 at 9:32 AM, Damjan Jovanovic damjan@gmail.com wrote: Please send the output of lsusb -v first so I can see if it's useful. Thank you for the offer Damjan On Tue, Sep 21, 2010 at 3:58 PM, Tom Spear speeddy...@gmail.com wrote: Now that I think about it, I have a webcam which the last supported windows version was XP. I'm not using it for anything since I have another one which is supported in 7 and linux, but I don't know if it's picked up in linux either. I could send it your way too tho. Thanks Tom On Tue, Sep 21, 2010 at 8:54 AM, Tom Spear speeddy...@gmail.com wrote: I have a USB pedometer that uploads the data to the internet. I could get another one and the driver software for you to play with. You have to be a registered member for a monthly fee to get one otherwise, but my job sponsors anyone that wants to get/stay in shape that works for them, so getting an extra pedometer is fine by me. I have been hoping for an opportunity to mention that it doesn't work, and this seems like as good as any. :-) Thanks Tom On Tue, Sep 21, 2010 at 5:03 AM, Damjan Jovanovic damjan@gmail.com wrote: On Wed, Sep 15, 2010 at 1:39 AM, Eric Durbin eadur...@gmail.com wrote: On Tue, Sep 14, 2010 at 10:48 AM, Damjan Jovanovic damjan@gmail.com wrote: When last I heard from Alexander Morozov (October 2009), he wasn't working on those patches much, and had no interest in sending them to wine-patches. I did some work on USB since then, and sent some patches starting from around March 2010 (too many attempts to list, search for them). Most were rejected. The USB story goes as follows: My libusb patch was rejected IIRC because the libusb situation was unclear. There's the old libusb-0.1 and the new more powerful libusb-1.0. IIRC each *nix hacked up its own specific variation of libusb that had to be detected specifically, and some *nixes didn't support the libusb-1.0 interface yet (libusb-1.0 itself only supports Linux and MacOS when last I checked, and they were doing a Windows port). The ntoskrnl that Wine currently emulates is total bogus: one process per driver, drivers all in separate processes from each other. On Windows there's a single address space for all drivers and they can communicate amongst themselves. I don't think inter-driver communication is that crucial initially, but it will be eventually (eg. last I heard, the iPod driver stacks on top of USBSTOR.SYS, and multi-function USB devices can use a different driver for each interface - these may communicate among themselves with private ioctl requests). The big problem with the multi process situation is hardware sharing: how do you set it up so each driver accesses its own and only its own hardware? Drivers either start on system startup (Wine starts those with the first process that starts), or get loaded on-demand as the hardware is plugged in. Most drivers should install themselves to be loaded on-demand. Who loads those and how? Windows uses USBHUB.SYS to do device I/O and load drivers on demand. Alexandre didn't want that dll because it exports nothing (all its features are accessible via internal ioctls), and suggested adding the features to USBD.SYS instead, which we already have and which has exports. Now USBD.SYS is linked to by most (but not all) USB drivers so (most of the time
Re: USB Device Support
I have a USB pedometer that uploads the data to the internet. I could get another one and the driver software for you to play with. You have to be a registered member for a monthly fee to get one otherwise, but my job sponsors anyone that wants to get/stay in shape that works for them, so getting an extra pedometer is fine by me. I have been hoping for an opportunity to mention that it doesn't work, and this seems like as good as any. :-) Thanks Tom On Tue, Sep 21, 2010 at 5:03 AM, Damjan Jovanovic damjan@gmail.comwrote: On Wed, Sep 15, 2010 at 1:39 AM, Eric Durbin eadur...@gmail.com wrote: On Tue, Sep 14, 2010 at 10:48 AM, Damjan Jovanovic damjan@gmail.com wrote: When last I heard from Alexander Morozov (October 2009), he wasn't working on those patches much, and had no interest in sending them to wine-patches. I did some work on USB since then, and sent some patches starting from around March 2010 (too many attempts to list, search for them). Most were rejected. The USB story goes as follows: My libusb patch was rejected IIRC because the libusb situation was unclear. There's the old libusb-0.1 and the new more powerful libusb-1.0. IIRC each *nix hacked up its own specific variation of libusb that had to be detected specifically, and some *nixes didn't support the libusb-1.0 interface yet (libusb-1.0 itself only supports Linux and MacOS when last I checked, and they were doing a Windows port). The ntoskrnl that Wine currently emulates is total bogus: one process per driver, drivers all in separate processes from each other. On Windows there's a single address space for all drivers and they can communicate amongst themselves. I don't think inter-driver communication is that crucial initially, but it will be eventually (eg. last I heard, the iPod driver stacks on top of USBSTOR.SYS, and multi-function USB devices can use a different driver for each interface - these may communicate among themselves with private ioctl requests). The big problem with the multi process situation is hardware sharing: how do you set it up so each driver accesses its own and only its own hardware? Drivers either start on system startup (Wine starts those with the first process that starts), or get loaded on-demand as the hardware is plugged in. Most drivers should install themselves to be loaded on-demand. Who loads those and how? Windows uses USBHUB.SYS to do device I/O and load drivers on demand. Alexandre didn't want that dll because it exports nothing (all its features are accessible via internal ioctls), and suggested adding the features to USBD.SYS instead, which we already have and which has exports. Now USBD.SYS is linked to by most (but not all) USB drivers so (most of the time) it automatically gets loaded into each one - great right? - but it has no idea which driver it got loaded with, nor a straightforward way to determine which device(s!) that driver wants to drive. Also, since most drivers only load on-demand, the driver will never load, and thus this won't work unless we load those drivers on startup instead. The other approach, which I tried, was to get Wine's mountmgr.sys to detect USB devices using HAL, then pass them to a loaded-on-startup instance of USBHUB.SYS using a Wine-private ioctl, which would detect the driver for the device and launch a new instance of itself that would make a device object and load the driver to attach to it. This was all a bit a hack (USBHUB.SYS uses environment variables to tell the child which device and driver to run) and Alexandre also didn't the the Wine-private ioctls. Alexander Morozov's patch did things the Windows way: all drivers in one ntoskrnl process - this won't work properly in Wine for years, if ever, since ntoskrnl is so incomplete and one bad driver will crash them all. Another possibility could be to keep drivers in separate processes, but allow inter-process communication, but I see serializing IRPs between processes as being complex and very slow. Driver installation is also quite a mission. Windows detects that the hardware doesn't have a driver installed, and then generates the device ID and compatible IDs and searches .INF files for one that can support it. Our setupapi needs to be substantially improved to be able to do the same, and some newdev.dll and manual INF parsing work to install the driver may also be necessary, and I can already think of cases where even class installers will be necessary too :-(. Wine only sends DeviceIoControl to drivers. For anything non-trivial, other file-related user-space functions (at least ReadFile, WriteFile) need to go to the driver too. The infrastructure for this does not even exist yet, and would probably affects wineserver as well. Regression tests for ntosnkrl.exe and kernel drivers don't exist, and are difficult to come up with, since we'd
Re: USB Device Support
Now that I think about it, I have a webcam which the last supported windows version was XP. I'm not using it for anything since I have another one which is supported in 7 and linux, but I don't know if it's picked up in linux either. I could send it your way too tho. Thanks Tom On Tue, Sep 21, 2010 at 8:54 AM, Tom Spear speeddy...@gmail.com wrote: I have a USB pedometer that uploads the data to the internet. I could get another one and the driver software for you to play with. You have to be a registered member for a monthly fee to get one otherwise, but my job sponsors anyone that wants to get/stay in shape that works for them, so getting an extra pedometer is fine by me. I have been hoping for an opportunity to mention that it doesn't work, and this seems like as good as any. :-) Thanks Tom On Tue, Sep 21, 2010 at 5:03 AM, Damjan Jovanovic damjan@gmail.comwrote: On Wed, Sep 15, 2010 at 1:39 AM, Eric Durbin eadur...@gmail.com wrote: On Tue, Sep 14, 2010 at 10:48 AM, Damjan Jovanovic damjan@gmail.com wrote: When last I heard from Alexander Morozov (October 2009), he wasn't working on those patches much, and had no interest in sending them to wine-patches. I did some work on USB since then, and sent some patches starting from around March 2010 (too many attempts to list, search for them). Most were rejected. The USB story goes as follows: My libusb patch was rejected IIRC because the libusb situation was unclear. There's the old libusb-0.1 and the new more powerful libusb-1.0. IIRC each *nix hacked up its own specific variation of libusb that had to be detected specifically, and some *nixes didn't support the libusb-1.0 interface yet (libusb-1.0 itself only supports Linux and MacOS when last I checked, and they were doing a Windows port). The ntoskrnl that Wine currently emulates is total bogus: one process per driver, drivers all in separate processes from each other. On Windows there's a single address space for all drivers and they can communicate amongst themselves. I don't think inter-driver communication is that crucial initially, but it will be eventually (eg. last I heard, the iPod driver stacks on top of USBSTOR.SYS, and multi-function USB devices can use a different driver for each interface - these may communicate among themselves with private ioctl requests). The big problem with the multi process situation is hardware sharing: how do you set it up so each driver accesses its own and only its own hardware? Drivers either start on system startup (Wine starts those with the first process that starts), or get loaded on-demand as the hardware is plugged in. Most drivers should install themselves to be loaded on-demand. Who loads those and how? Windows uses USBHUB.SYS to do device I/O and load drivers on demand. Alexandre didn't want that dll because it exports nothing (all its features are accessible via internal ioctls), and suggested adding the features to USBD.SYS instead, which we already have and which has exports. Now USBD.SYS is linked to by most (but not all) USB drivers so (most of the time) it automatically gets loaded into each one - great right? - but it has no idea which driver it got loaded with, nor a straightforward way to determine which device(s!) that driver wants to drive. Also, since most drivers only load on-demand, the driver will never load, and thus this won't work unless we load those drivers on startup instead. The other approach, which I tried, was to get Wine's mountmgr.sys to detect USB devices using HAL, then pass them to a loaded-on-startup instance of USBHUB.SYS using a Wine-private ioctl, which would detect the driver for the device and launch a new instance of itself that would make a device object and load the driver to attach to it. This was all a bit a hack (USBHUB.SYS uses environment variables to tell the child which device and driver to run) and Alexandre also didn't the the Wine-private ioctls. Alexander Morozov's patch did things the Windows way: all drivers in one ntoskrnl process - this won't work properly in Wine for years, if ever, since ntoskrnl is so incomplete and one bad driver will crash them all. Another possibility could be to keep drivers in separate processes, but allow inter-process communication, but I see serializing IRPs between processes as being complex and very slow. Driver installation is also quite a mission. Windows detects that the hardware doesn't have a driver installed, and then generates the device ID and compatible IDs and searches .INF files for one that can support it. Our setupapi needs to be substantially improved to be able to do the same, and some newdev.dll and manual INF parsing work to install the driver may also be necessary, and I can already think of cases where even class installers will be necessary too :-(. Wine only sends
Re: USB Device Support
Attached is the lsusb -v output, trimmed to only include the pedometer's info. I have many USB devices, so I didn't want to leave you to sort through a bunch of useless info. I don't have the webcam with me at the moment, but I will see if I can find it when I am at home soon. Thanks Tom On Tue, Sep 21, 2010 at 9:32 AM, Damjan Jovanovic damjan@gmail.comwrote: Please send the output of lsusb -v first so I can see if it's useful. Thank you for the offer Damjan On Tue, Sep 21, 2010 at 3:58 PM, Tom Spear speeddy...@gmail.com wrote: Now that I think about it, I have a webcam which the last supported windows version was XP. I'm not using it for anything since I have another one which is supported in 7 and linux, but I don't know if it's picked up in linux either. I could send it your way too tho. Thanks Tom On Tue, Sep 21, 2010 at 8:54 AM, Tom Spear speeddy...@gmail.com wrote: I have a USB pedometer that uploads the data to the internet. I could get another one and the driver software for you to play with. You have to be a registered member for a monthly fee to get one otherwise, but my job sponsors anyone that wants to get/stay in shape that works for them, so getting an extra pedometer is fine by me. I have been hoping for an opportunity to mention that it doesn't work, and this seems like as good as any. :-) Thanks Tom On Tue, Sep 21, 2010 at 5:03 AM, Damjan Jovanovic damjan@gmail.com wrote: On Wed, Sep 15, 2010 at 1:39 AM, Eric Durbin eadur...@gmail.com wrote: On Tue, Sep 14, 2010 at 10:48 AM, Damjan Jovanovic damjan@gmail.com wrote: When last I heard from Alexander Morozov (October 2009), he wasn't working on those patches much, and had no interest in sending them to wine-patches. I did some work on USB since then, and sent some patches starting from around March 2010 (too many attempts to list, search for them). Most were rejected. The USB story goes as follows: My libusb patch was rejected IIRC because the libusb situation was unclear. There's the old libusb-0.1 and the new more powerful libusb-1.0. IIRC each *nix hacked up its own specific variation of libusb that had to be detected specifically, and some *nixes didn't support the libusb-1.0 interface yet (libusb-1.0 itself only supports Linux and MacOS when last I checked, and they were doing a Windows port). The ntoskrnl that Wine currently emulates is total bogus: one process per driver, drivers all in separate processes from each other. On Windows there's a single address space for all drivers and they can communicate amongst themselves. I don't think inter-driver communication is that crucial initially, but it will be eventually (eg. last I heard, the iPod driver stacks on top of USBSTOR.SYS, and multi-function USB devices can use a different driver for each interface - these may communicate among themselves with private ioctl requests). The big problem with the multi process situation is hardware sharing: how do you set it up so each driver accesses its own and only its own hardware? Drivers either start on system startup (Wine starts those with the first process that starts), or get loaded on-demand as the hardware is plugged in. Most drivers should install themselves to be loaded on-demand. Who loads those and how? Windows uses USBHUB.SYS to do device I/O and load drivers on demand. Alexandre didn't want that dll because it exports nothing (all its features are accessible via internal ioctls), and suggested adding the features to USBD.SYS instead, which we already have and which has exports. Now USBD.SYS is linked to by most (but not all) USB drivers so (most of the time) it automatically gets loaded into each one - great right? - but it has no idea which driver it got loaded with, nor a straightforward way to determine which device(s!) that driver wants to drive. Also, since most drivers only load on-demand, the driver will never load, and thus this won't work unless we load those drivers on startup instead. The other approach, which I tried, was to get Wine's mountmgr.sys to detect USB devices using HAL, then pass them to a loaded-on-startup instance of USBHUB.SYS using a Wine-private ioctl, which would detect the driver for the device and launch a new instance of itself that would make a device object and load the driver to attach to it. This was all a bit a hack (USBHUB.SYS uses environment variables to tell the child which device and driver to run) and Alexandre also didn't the the Wine-private ioctls. Alexander Morozov's patch did things the Windows way: all drivers in one ntoskrnl process - this won't work properly in Wine for years, if ever, since ntoskrnl is so incomplete and one bad driver will crash them all. Another possibility could
Re: USB Device Support
On Tue, Sep 21, 2010 at 11:04 AM, Damjan Jovanovic damjan@gmail.comwrote: On Tue, Sep 21, 2010 at 5:04 PM, Tom Spear speeddy...@gmail.com wrote: Attached is the lsusb -v output, trimmed to only include the pedometer's info. I have many USB devices, so I didn't want to leave you to sort through a bunch of useless info. I don't have the webcam with me at the moment, but I will see if I can find it when I am at home soon. Thanks Tom On Tue, Sep 21, 2010 at 9:32 AM, Damjan Jovanovic damjan@gmail.com wrote: Please send the output of lsusb -v first so I can see if it's useful. Thank you for the offer Damjan On Tue, Sep 21, 2010 at 3:58 PM, Tom Spear speeddy...@gmail.com wrote: Now that I think about it, I have a webcam which the last supported windows version was XP. I'm not using it for anything since I have another one which is supported in 7 and linux, but I don't know if it's picked up in linux either. I could send it your way too tho. Thanks Tom On Tue, Sep 21, 2010 at 8:54 AM, Tom Spear speeddy...@gmail.com wrote: I have a USB pedometer that uploads the data to the internet. I could get another one and the driver software for you to play with. You have to be a registered member for a monthly fee to get one otherwise, but my job sponsors anyone that wants to get/stay in shape that works for them, so getting an extra pedometer is fine by me. I have been hoping for an opportunity to mention that it doesn't work, and this seems like as good as any. :-) Thanks Tom On Tue, Sep 21, 2010 at 5:03 AM, Damjan Jovanovic damjan@gmail.com wrote: On Wed, Sep 15, 2010 at 1:39 AM, Eric Durbin eadur...@gmail.com wrote: On Tue, Sep 14, 2010 at 10:48 AM, Damjan Jovanovic damjan@gmail.com wrote: When last I heard from Alexander Morozov (October 2009), he wasn't working on those patches much, and had no interest in sending them to wine-patches. I did some work on USB since then, and sent some patches starting from around March 2010 (too many attempts to list, search for them). Most were rejected. The USB story goes as follows: My libusb patch was rejected IIRC because the libusb situation was unclear. There's the old libusb-0.1 and the new more powerful libusb-1.0. IIRC each *nix hacked up its own specific variation of libusb that had to be detected specifically, and some *nixes didn't support the libusb-1.0 interface yet (libusb-1.0 itself only supports Linux and MacOS when last I checked, and they were doing a Windows port). The ntoskrnl that Wine currently emulates is total bogus: one process per driver, drivers all in separate processes from each other. On Windows there's a single address space for all drivers and they can communicate amongst themselves. I don't think inter-driver communication is that crucial initially, but it will be eventually (eg. last I heard, the iPod driver stacks on top of USBSTOR.SYS, and multi-function USB devices can use a different driver for each interface - these may communicate among themselves with private ioctl requests). The big problem with the multi process situation is hardware sharing: how do you set it up so each driver accesses its own and only its own hardware? Drivers either start on system startup (Wine starts those with the first process that starts), or get loaded on-demand as the hardware is plugged in. Most drivers should install themselves to be loaded on-demand. Who loads those and how? Windows uses USBHUB.SYS to do device I/O and load drivers on demand. Alexandre didn't want that dll because it exports nothing (all its features are accessible via internal ioctls), and suggested adding the features to USBD.SYS instead, which we already have and which has exports. Now USBD.SYS is linked to by most (but not all) USB drivers so (most of the time) it automatically gets loaded into each one - great right? - but it has no idea which driver it got loaded with, nor a straightforward way to determine which device(s!) that driver wants to drive. Also, since most drivers only load on-demand, the driver will never load, and thus this won't work unless we load those drivers on startup instead. The other approach, which I tried, was to get Wine's mountmgr.sys to detect USB devices using HAL, then pass them to a loaded-on-startup instance of USBHUB.SYS using a Wine-private ioctl, which would detect the driver for the device and launch a new instance of itself that would make a device object and load the driver to attach to it. This was all a bit a hack (USBHUB.SYS uses environment variables to tell
Re: USB Device Support
On Tue, Sep 21, 2010 at 1:41 PM, Damjan Jovanovic damjan@gmail.comwrote: On Tue, Sep 21, 2010 at 8:07 PM, Tom Spear speeddy...@gmail.com wrote: On Tue, Sep 21, 2010 at 11:04 AM, Damjan Jovanovic damjan@gmail.com wrote: On Tue, Sep 21, 2010 at 5:04 PM, Tom Spear speeddy...@gmail.com wrote: Attached is the lsusb -v output, trimmed to only include the pedometer's info. I have many USB devices, so I didn't want to leave you to sort through a bunch of useless info. I don't have the webcam with me at the moment, but I will see if I can find it when I am at home soon. Thanks Tom On Tue, Sep 21, 2010 at 9:32 AM, Damjan Jovanovic damjan@gmail.com wrote: Please send the output of lsusb -v first so I can see if it's useful. Thank you for the offer Damjan On Tue, Sep 21, 2010 at 3:58 PM, Tom Spear speeddy...@gmail.com wrote: Now that I think about it, I have a webcam which the last supported windows version was XP. I'm not using it for anything since I have another one which is supported in 7 and linux, but I don't know if it's picked up in linux either. I could send it your way too tho. Thanks Tom On Tue, Sep 21, 2010 at 8:54 AM, Tom Spear speeddy...@gmail.com wrote: I have a USB pedometer that uploads the data to the internet. I could get another one and the driver software for you to play with. You have to be a registered member for a monthly fee to get one otherwise, but my job sponsors anyone that wants to get/stay in shape that works for them, so getting an extra pedometer is fine by me. I have been hoping for an opportunity to mention that it doesn't work, and this seems like as good as any. :-) Thanks Tom On Tue, Sep 21, 2010 at 5:03 AM, Damjan Jovanovic damjan@gmail.com wrote: On Wed, Sep 15, 2010 at 1:39 AM, Eric Durbin eadur...@gmail.com wrote: On Tue, Sep 14, 2010 at 10:48 AM, Damjan Jovanovic damjan@gmail.com wrote: When last I heard from Alexander Morozov (October 2009), he wasn't working on those patches much, and had no interest in sending them to wine-patches. I did some work on USB since then, and sent some patches starting from around March 2010 (too many attempts to list, search for them). Most were rejected. The USB story goes as follows: My libusb patch was rejected IIRC because the libusb situation was unclear. There's the old libusb-0.1 and the new more powerful libusb-1.0. IIRC each *nix hacked up its own specific variation of libusb that had to be detected specifically, and some *nixes didn't support the libusb-1.0 interface yet (libusb-1.0 itself only supports Linux and MacOS when last I checked, and they were doing a Windows port). The ntoskrnl that Wine currently emulates is total bogus: one process per driver, drivers all in separate processes from each other. On Windows there's a single address space for all drivers and they can communicate amongst themselves. I don't think inter-driver communication is that crucial initially, but it will be eventually (eg. last I heard, the iPod driver stacks on top of USBSTOR.SYS, and multi-function USB devices can use a different driver for each interface - these may communicate among themselves with private ioctl requests). The big problem with the multi process situation is hardware sharing: how do you set it up so each driver accesses its own and only its own hardware? Drivers either start on system startup (Wine starts those with the first process that starts), or get loaded on-demand as the hardware is plugged in. Most drivers should install themselves to be loaded on-demand. Who loads those and how? Windows uses USBHUB.SYS to do device I/O and load drivers on demand. Alexandre didn't want that dll because it exports nothing (all its features are accessible via internal ioctls), and suggested adding the features to USBD.SYS instead, which we already have and which has exports. Now USBD.SYS is linked to by most (but not all) USB drivers so (most of the time) it automatically gets loaded into each one - great right? - but it has no idea which driver it got loaded with, nor a straightforward way to determine which device(s!) that driver wants to drive. Also, since most drivers only load on-demand, the driver will never load, and thus this won't work unless we load those drivers on startup instead. The other approach, which I tried
Re: Keeping people from trying iTunes in Wine?
Perhaps making a hash based on app name and version in the appdb, and then have wine reading the hash from the app to check against the appdb. If anyone uses Fedora, their ABRT tool generates hashes for different bugs and then searches their bugzilla before submitting the crash dump, to find if a report is already generated. If the report is already in bugz, then it appends to that bug.We could do something similar, but check against the appdb, and notify the user.. Maybe there could be a separate builtin app (like notepad and explorer) to read the appdb and check ratings, and allow access to the appdb without having to fire up the web browser? Thanks Tom On Thu, Sep 9, 2010 at 2:15 PM, Roderick Colenbrander thunderbir...@gmail.com wrote: On Thu, Sep 9, 2010 at 6:12 PM, Damjan Jovanovic damjan@gmail.com wrote: On Thu, Sep 9, 2010 at 5:47 PM, Austin English austinengl...@gmail.com wrote: On Thu, Sep 9, 2010 at 8:34 AM, Rosanne DiMesio dime...@earthlink.net wrote: On Thu, 9 Sep 2010 00:59:32 -0700 Dan Kegel d...@kegel.com wrote: Watching Twitter, one fairly frequently seems people trying and failing to run iTunes 10 and the like in Wine. Should we let them bash their heads against the wall like that? Maybe we should detect the top ten apps that don't work with Wine, and put up a warning dialog saying they are known not to work, and people shouldn't try. (Kind of like what Windows 7 does when you do something dangerous, e.g. try to look at the contents of drive C:.) Do you really want to prevent users from ever testing these apps in new versions of Wine, or trying to find workarounds? I do a fair amount of head-bashing myself, and I would find such a message patronizing and intrusive. Agreed. Wine doesn't make efforts to babysit users for most other things, I don't see why this should be any different. Also consider that if such a workaround were to go into wine, that code may long outlive the 'affected apps', and the list would quickly grow out of date. I suppose if a packager wanted to do something like this for their distro I wouldn't complain too much, unless users started asking about it in #winehq/the forums. But this _should not_ go into vanilla wine. -- -Austin The dialog could suggest upgrading Wine, that would prevent the affected app list from getting out of date. Damjan If we would want any application database stuff, perhaps appdb would be the place to store information like this. There should come some way to extract this information to an XML file or whatever format periodically. It could be packaged with a wine build or optionally downloaded when you run Wine or so. Roderick
Opinions on priority for an enhancement that Dan suggested some time ago?
Hi all, so it has been a long time since I have posted here. Much has changed; I now have 3 little ones. :-) Anyways, right to the point: I stumbled on an old bug of mine this evening while doing some long overdue mailbox maintenance. Bug 657: http://bugs.winehq.org/show_bug.cgi?id=657 Dan made a very good point back in 2007 that I believe now is the right time to possibly look at beginning an implementation of: [1] To summarize: Similar to the way that we now have a popup about wine-gecko missing, I believe it would be prudent to have some warning appear when certain native dll's are missing which required by the app being run, and which we do not implement builtins for. Similarly, I feel that it would be prudent for warnings to appear if certain linux shared objects are missing. More to the point is that there are quite a number of DLL's which won't be implemented soon, if ever, but yet we support running wine from the GUI both through shortcuts, and through binfmt_misc, thereby completely bypassing any hope of a user ever seeing some pretty critical errors that may occur. These errors could, if popped up on their screen, prevent a bunch of duplicates from appearing in Bugzilla, because the user can google for the error and be directed directly to a bug which someone else made, as opposed to not seeing anything and having no idea what to try googling for other than app x does this +wine As an example on the shared object side, on FC13 x86_64, using the rpm from the yum repo that RH provides, I can install wine, but since there is no RPM dependency currently for OpenAL (i686), I get no such file/directory warnings about the OpenAL shared objects when running certain programs (specifically, World of Warcraft and Starcraft II). Now granted the fact that OpenAL is not installed is, debatibly something to take up with the wine packager for RedHat and derivatives; I wanted to point out that with the way wine is at present, if I didn't run wine from a console, I would never have known that, and probably would still not have working sound, because there is nothing that says Hey! You're missing OpenAL! I'd like to hear opinions from others about what it might take, time-wise, if someone were to start now, to have something like this implemented. And also what kind of a priority this type of thing should have. Sure it is an enhancement, and so is naturally lower priority than, say a segfault in ntdll, but (imho) it's a pretty critical enhancement that is (again, imho) long overdue. Thoughts? Questions? [1] http://bugs.winehq.org/show_bug.cgi?id=657#c37 Also see comments of his below #37. Thanks Tom
Re: VST wrapper
On 6/25/07, Nathaniel Gray [EMAIL PROTECTED] wrote: By the way, please CC me with replies. I'm not set up to receive mail from the list. Might want to look into Codeweavers Crossover or Crossover Office. One of those handles IE plugins in linux firefox quite nicely. Should be a good starting point. Other people may have suggestions on how to achieve what you are looking for with standard wine, but I'm not sure. -- Thanks Tom
Re: Bugzilla mails and people replying to [EMAIL PROTECTED]
On 6/23/07, James Hawkins [EMAIL PROTECTED] wrote: Or how about adding some magic to the ML where any reply to wine-bugs gets added as a comment to the bug report? Web programming isn't my specialty, so I don't know if this if feasible or possible, but it seems like the best solution to the problem That would be great (we also have a message to attach logs and not paste, but most people don't read or follow that advice). I think the problem with that is that it isn't in bold, it kind of blends in with the page. I know that it had been there a while before I finally noticed it, although I had been attaching logs prior to that message being put there. I think that a notice in the email at the very top right before the link would be noticable enough to where people would see it in 90% of the cases when they would have hit reply otherwise. Another solution I thought of would be to just set wine-bugs@ to either bounce messages back to the sender with undeliverable address, or redirect it to /dev/null, although neither of those would be as helpful as James' or my own ideas. Last solution, although I personally don't like it would be to have reply-to set to wine-devel@ -- Thanks Tom
Re: msi: Fix use of uninitialized variable (Coverity) (Try 2)
On 6/23/07, James Hawkins [EMAIL PROTECTED] wrote: On 6/23/07, Andrew Talbot [EMAIL PROTECTED] wrote: This patch should fix Coverity bug CID-562. -- Andy. --- Changelog: msi: Fix use of uninitialized variable (Coverity). diff -urN a/dlls/msi/action.c b/dlls/msi/action.c --- a/dlls/msi/action.c 2007-06-18 17:52:27.0 +0100 +++ b/dlls/msi/action.c 2007-06-23 17:08:55.0 +0100 @@ -4648,7 +4648,7 @@ LPWSTR deformatted, ptr; DWORD flags, type, size; LONG res; -HKEY env, root = HKEY_CURRENT_USER; +HKEY env = NULL, root = HKEY_CURRENT_USER; static const WCHAR environment[] = {'S','y','s','t','e','m','\\', @@ -4759,7 +4759,7 @@ res = RegSetValueExW(env, name, 0, type, (LPVOID)newval, size); done: -RegCloseKey(env); +if (env) RegCloseKey(env); msi_free(deformatted); msi_free(data); msi_free(newval); Please don't check env for NULL; RegCloseKey will just fail harmlessly if env is NULL. We spent a lot of time removing such checks. Perhaps a conformance test is in order, because from what I have read, RegCloseKey returns a nonzero error code on any error, including a NULL variable being passed to it, which screws up obtaining the original error if x program or api tries to use GetLastError. We don't want bugs in wine, unless it is duplicating a windows bug, and what I see if we don't check the way Andrew has, is a bug that I am pretty sure does not exist in windows. RegCloseKey documentation (our RegCloseKey function matches this): http://msdn2.microsoft.com/en-us/library/ms724837.aspx -- Thanks Tom
Re: wine menus
I actually did some testing behind .lnk functionality (unscientific testing, that is) with wine a few months ago. Xorg actually fully supports Windows' .lnk files with 2 minor annoyances.. The .lnk files could theoretically be used as the links in the menus if these 2 annoyances could be fixed (probably by Xorg, not wine).. 1) The .lnk extension shows up even in the menu 2) The icon is just a picture of a window with a white background and a MS flag. If that icon could be changed to, say the icon specified inside the .lnk file, it would be perfect. Other than that, you can double click on a .lnk file and the program will be executed with wine (assuming BINFMT_MISC is setup properly on your system), including whatever command line options are specified to the program, in the lnk file. Now with all that being said, this still deviates from Freedesktop, so I kinda need to backtrack and say that I am not recommending that we USE those files as the menu entries. Quite the contrary! I think if we can find out what license the xorg code to parse the .lnk files is under, then we could potentially use that code to help us with a parser, or if it is not LGPL'ed then we might be able to get the code author's permission.. One other amendment to what I said above, because I haven't fully checked this.. It may just be KDE that supports the lnk files, and not Xorg.. As KDE is what I use, I don't have time to install/test with GNOME, or others. Someone please feel free to give it a shot with something other than KDE and post results back.. Hope that helps Tom P.S. While we are at it, can someone look into bug 3548 please? On 6/22/07, Damjan Jovanovic [EMAIL PROTECTED] wrote: Hi At the moment, wine builds fd.o menus by calling winemenubuilder and wineshelllink when shell32's IPersistFile_fnSave is invoked. This both misses menus copied directly into the menu directory without calling IPersistFile_fnSave, and provides no way to remove the menus when the app is uninstalled. Wine also keeps .lnk files in ~/.wine/drive_c/windows/profiles/user/Start Menu/ just like Windows does, and those .lnk files seem like a much better representation of the menus that should appear on the system, because they are also deleted when an app is uninstalled. Is it possible to add a new utility, or patch an existing utility like wineboot, to synchronize between wine's .lnk files in the windows directory and the fd.o menus, instead of using winemenubuilder/wineshelllink? Would a patch that does that be okay? Thank you Damjan -- Thanks Tom
Re: RegOpenKeyExW() Question
On 6/22/07, Andrew Talbot [EMAIL PROTECTED] wrote: Hi, I want to patch msi:action.c:ITERATE_WriteEnvironmentString() so that it only calls RegCloseKey(env), in the cleanup, if env has been initialized (to fix Coverity report CID-562). I can bypass the call to RegCloseKey() for any early exit that occurs prior to calling RegOpenKeyExW(), and I can include the call to RegCloseKey() for exits that occur after a successful call to RegOpenKeyExW(). However, the problem, for me, is what to do if the call to RegOpenKeyExW() fails: does the occurrence of this initialization depend on the reason for failure. In other words, do I need something like the following code to cater for the reason failure occurred? res = RegOpenKeyExW(root, environment, 0, KEY_ALL_ACCESS, env); if (res != ERROR_SUCCESS) if (res == ERROR_INVALID_HANDLE || res == RtlNtStatusToDosError(STATUS_BUFFER_OVERFLOW || res == RtlNtStatusToDosError(STATUS_INVALID_PARAMETER)) goto done2; else goto done1; ... done1: RegCloseKey(env); done2: ... Or will the following code suffice? res = RegOpenKeyExW(root, environment, 0, KEY_ALL_ACCESS, env); if (res != ERROR_SUCCESS) goto done2; ... done1: RegCloseKey(env); done2: ... At the risk of looking like an idiot, I'll take a stab at this.. If you RegCloseKey on env, and it is not initialized, then RegCloseKey will generate ERROR_INVALID_HANDLE, so I personally think that something like the first approach would be the better one to go with. env should only be initialized (according to what I have seen) if RegOpenKeyExW returns ERROR_SUCCESS. But I haven't seen every case, so there may be something that causes it to be initialized when the return is not ERROR_SUCCESS. -- Thanks Tom
Bugzilla mails and people replying to [EMAIL PROTECTED]
Just curious how trivial it would be to add a note to the bugzilla mail template saying to post replies to comments in the bug? Something like the following: A bug you are watching, Bug #(bug number), has a new comment. To reply to this comment, click (bug link). Then the rest of the email as it currently appears. I'm asking because I have received several replies to the list recently, and once wine reaches 1.0, I'm sure many more people will join the list and do the same. I don't want comments to be lost due to them not being posted in the proper place, as everyone's input on a bug is important. -- Thanks Tom
Re: NULL ptr dereferences found with Calysto static checker
On 6/21/07, James Hawkins [EMAIL PROTECTED] wrote: On 6/21/07, Michael Stefaniuc [EMAIL PROTECTED] wrote: The problem with NULL pointer dereferences is that those may be valid. If a function crashes in Windows on a NULL pointer dereference Wine should/has do the same. Why? Several apps depend on these 'bugs'. This is true, many applications are written to work around bugs in windows. If we fix these bugs, then the apps dont have anything to work around, and therefore they do unexpected things Wine is meant to run software written for windows. I doubt that many people use Wine as a platform to write software intended for windows. So, if you run all the applications the Windows run, just better, what's wrong with that? If we don't conform to the API, it's not better. Like I said before, we've run into bugs where an app is expecting an exception, and it won't work if it doesn't get it. See above -- Thanks Tom
Re: Uninstaller
On 6/14/07, Misha Koshelev [EMAIL PROTECTED] wrote: Actually that brings up a good point, does anything ever get written to Software\Microsoft\Windows\CurrentVersion in HKEY_CURRENT_USER in windows? I have no such key on my (rather plain) XP install, and from what I can tell wine MSI never writes to this key (see dlls/msi/registry.c, function MSIREG_OpenUninstallKey line 378) in HKEY_CURRENT_USER. Is this done by some specific application, or if not what was the original motivation for the patches? From what I can tell this was not in the emails you linked. Btw, I think it would be a good idea to include this kind of info when you submit the patches. The original try of the emails (I think towards the end of May), had it. There is a bug filed for it on bugzilla (search for IMVU). Some applications (IMVU instant messenger, for one) do use current user by default, and some installers ask where you want start menu entries and application settings stored, all users, or current user, where all users as far as the registry is concerned would be HKLM.. It was brought up in private after I sent the linked emails, that I should include that info in every try, so I will do so in the future. As for the currentversion key not being there in your xp install, it may be just because it's a plain xp install. I use XP on a daily basis and frequently access the registry, and I've never noticed a time when it was not there with something under it, although I never checked immediately after a clean install. -- Thanks Tom
Uninstaller
It's been a while since I brought anything up on this, so I figured I'd ask what was wrong with the last set of patches I sent.. The patches are at [1] and [2], the discussion generated by them are at [3], [4], [5], and [6]... Summary, I did everything I was told in private by a developer (who isn't alexandre, i know), and I got told by another developer, who also isn't alexandre, to fix some things that the first developer said didn't need fixing.. So should I fix those things, or is there some other reason it wasn't committed? [1] http://www.winehq.org/pipermail/wine-patches/2007-June/039911.html [2] http://www.winehq.org/pipermail/wine-patches/2007-June/039912.html [3] http://www.winehq.org/pipermail/wine-devel/2007-June/057253.html [4] http://www.winehq.org/pipermail/wine-devel/2007-June/057254.html [5] http://www.winehq.org/pipermail/wine-devel/2007-June/057255.html [6] http://www.winehq.org/pipermail/wine-devel/2007-June/057258.html Someone please let me know what is wrong (I'd love to hear from The Man, if he has time to take a look), but anyone's responses are welcome (hopefully I get more than 2 reponses this time lol) -- Thanks Tom
Re: How to determine if something is debuggable ... ?
On 6/1/07, Marc G. Fournier [EMAIL PROTECTED] wrote: Another thing you can _try_, since it looks like it messing up in msvcrt or possibly even dbghelp, is to get msvcrt.dll and dbghelp.dll from a windows install, place them in your wine's c:\windows\system32 folder (usually ~/.wine/drive_c/windows/system32), and then use winecfg to set both dll's to run (native, builtin) for that application.. See if that helps, no guarantees tho, especially since it is FreeBSD. -- Thanks Tom
Re: Bug triage, or spam?
On 5/31/07, Jan Zerebecki [EMAIL PROTECTED] wrote: I'm not sure there is a agreement what some things here mean. The following is my understanding of things, please correct me or state differing understanding: triage bugs: Make sure the bug is properly filed, has enough information and possibly uncover the cause (e.g. regression testing, finding where a NULL that causes a crash inside the application comes from). This also includes marking a bug resolved,fixed or closed or whatever, but the prior thing is more important because it makes it easier to fix. Agree resolved,fixed: I only mark bugs where I'm confident that they are really fixed as this. So if I need to ask the reporter or some user if it now works for them I do this before resolving it. I think I never closed a bug. Agree To detect e.g. resolved bugs with new comments (e.g. requesting reopen) I run a query for changed bugs (where I made a comment) since last date up to which I queried this (I noted that down) and e.g. yesterday. Closing bugs doesn't help here either as they could be closed in error, so someone would still want to request those to be reopened. True, bugs could be closed in error as well, which is why I don't close a bug, unless I was working with the reporter, or it is over (eg) 6 months since the last comment, and is already resolved. So is someone really using the closed status (not in the sense that they set it but e.g. use it in queries)? No, but that was my point. I search thru resolved bugs, to double check that users are satisfied with the result, and it does me no good to do that if I am searching thru bugs that are 6months old. If it is closed, a query for resolved bugs will not find those that are closed On Thu, May 31, 2007 at 05:36:45PM -0500, Tom Spear wrote: So I was closing bugs that were invalid/abandoned/dupe/worksforme so that they wouldnt show in the lists of resolved bugs, so its less I have to sort thru Does closed convey any more meaning than resolved as invalid/abandoned/dupe/worksforme? I mean who would mark a bug as resolved if that is not the conclusion and not reopen when that was done in error? So isn't closing perhaps something we _really_ want to avoid doing too prematurely? Perhaps something we only do every major release ( like 0.9 ). Otherwise it looses it's meaning as this is something we never ever need to look at. See above, however you are onto something with the only closing stale bugs at major releases.. According to James, the standard for marking a bug abandoned is 6 months from the request of more information without any response from the reporter or someone else having the same issue, and also not reproducible via a download. Perhaps we could do something similar to what Jonathan did before 0.9, say ping every bug at the 1.0 code freeze, and then resolve AND close any with no response, or that the reporter replied saying it is not an issue anymore.. With that in mind, once 1.0 goes live, will we still be doing monthly stable 1.0x releases, or will the release cycle be more of any x.0x releases are development and the stables will be x.x or x.5? However it is done, I think it would be a good idea to do pings of bugs prior to any STABLE release, and during development release periods, just ping when more than 6mos old like we currently are doing. -- Thanks Tom
Re: [1/2] uninstaller: add ability to scan HKCU for uninstall entries (Try 2)
On 6/1/07, James Hawkins [EMAIL PROTECTED] wrote: On 6/1/07, Tom Spear [EMAIL PROTECTED] wrote: Here is a 2nd attempt at the last one I sent. This time the root field is initialized, and this part of the patch is written against current git, not against changes to my own tree. Oh and its onl;y a 2-parter This will make wine's programs/uninstaller look thru not only HKEY_LOCAL_MACHINE for uninstall entries, but also HKEY_CURRENT_USER. +/* Loop thru HKCU first, then thru HKLM */ +for (iRootKey=0; iRootKeysizeof(rootKeys) / sizeof(rootKeys[0]); ++iRootKey) +{ +/* If there is no uninstall info in a specific root key, + * finish this run and go to the next */ +if (RegOpenKeyExW(rootKeys[iRootKey], PathUninstallW, 0, + KEY_READ, hkeyUninst) != ERROR_SUCCESS) +continue; Indentation? It was recommended to separate the indentation from the rest of the patch by Dan, because it makes it easier to see what lines actually changed -- Thanks Tom
Re: [2/2] uninstaller: indentation cleanups (try 2)
On 6/1/07, James Hawkins [EMAIL PROTECTED] wrote: On 6/1/07, Tom Spear [EMAIL PROTECTED] wrote: Here is part 2 of try 2. Just cleans up the indentation of my changes in the previous patch in this series. I think you misunderstood what Dan meant about indentation changes. You shouldn't have a patch that doesn't properly indent if you add a new block, and then come in with another patch to fix the indentation. That is exactly what he told me to do, and I even submitted it to him like this and he said to send it in. -- Thanks Tom
Re: [1/2] uninstaller: add ability to scan HKCU for uninstall entries (Try 2)
On 6/1/07, Christoph Frick [EMAIL PROTECTED] wrote: please excuse my bold mail on top of plain ignorance of the windows api - but i would consider finding a key in either the user or the global branch of the registry a task that many developers had to face (even in wine) and i bet there is an API function for that? would it be a Good Thing to add such a thing if not? I looked, but was unable to find any. -- Thanks Tom
Bug triage, or spam?
Hi all, I just have a quick question.. I got a message from someone last night asking me to stop closing bugs because I'm spamming the list, however I also received a message from someone a couple of nights ago thanking me for doing bug triage.. Which is it, and if it is both, then what draws the line between triage and spam, and how do I close stale bugs without it being considered spam? -- Thanks Tom
Whitespace changes in an indentation patch, is it acceptable?
Hi all, just curious, in my work on uninstaller, I am writing my patches to where when indentation is changed, due to adding a for loop, it is done in a separate patch file. I was wondering if it is acceptable to make whitespace changes to other parts of the file in that same patch.. For example, there is: int somevariable; do something; (there are 2 spaces in the blank line). Would it be acceptable to remove those spaces in a patch that changes the following code (which is introduced by the first patch in the series): for blah { do something; for something else { do another thing; } } to: for blah { do something; for something else { do another thing; } } or should I just leave those extra whitespaces in the blank line alone? -- Thanks Tom
Re: Bug triage, or spam?
On 5/31/07, Ben Hodgetts [EMAIL PROTECTED] wrote: Can someone explain to me the point of closing bugs please? If it's resolved one way or another then surely that's enough? Just seems like a waste of effort to be honest. Well, I can tell you this about it, it's supposed to work like this: - User reports bug - Developer looks into bug, and eventually fixes it, at which point he marks it resolved, and asks user to verify - User says yes it is resolved and marks it verified, or says no it is not, at which point developer reopens bug, and the process repeats - Once marked verified, developer then goes back and closes bug. It hardly ever works that way, unfortunately, but I prefer to see a bunch of closed bugs than ones that are sitting marked resolved as invalid, or resolved as works for me, etc, since invalid isnt really resolved, and neither is works for me.. Thats why I go thru and close the bugs that are just left resolved for more than a month or 2. -- Thanks Tom
Re: Bug triage, or spam?
On 5/31/07, James Hawkins [EMAIL PROTECTED] wrote: Marking bugs as closed has nothing to do with bug triage. Triaging bugs would be a really helpful thing, but mass-closing bugs does nothing but give subscribers a whole lot of emails to delete. We don't keep track of stats like other projects, so there's really no point to blindly close all bugs that are resolved, instead of closing bugs on a case by case basis. I'm not blindly closing all bugs marked resolved.. It seems to me that you didnt real thru all of the emails, but I have been posting comments, etc before closing, and there are some that are marked resolved that I did not close, or that I even reopened based on the most recent commentS that indicate the bug might not be fixed, or might actully be a real wine bug.. since invalid isnt really resolved, and neither is works for me But closing these bugs somehow makes them resolved for you? It's an organizational thing, if they are closed, then we know for sure that they are resolved. Since everyone is prone to marking a bug as invalid when it is really a bug, or works for me when it really should be open because some users are experiencing it but not others, leaving it as resolved just says that the person who resolved it was too lazy to close it, and that they didn't check back to make sure that some other users didnt have the same problem.. Case in point: bug 3889, has been resolved as worksforme (i did it over a year ago), when it is really a bug, and has never truely been resolved in wine. I just looked back thru the conversation that took place on wine-devel and saw no definitive response saying that the bug was fixed by an alexandre patch, and saw nothing that says the bug is not a bug, so I am about to reopen it and ask if the issue is fixed in current wine. -- Thanks Tom
Re: Whitespace changes in an indentation patch, is it acceptable?
On 5/31/07, Dan Kegel [EMAIL PROTECTED] wrote: Now that you've fixed a couple bugs in your uninstall patch, I think you should post the very simplest form of it possible, without any other change mixed in. In this case, I think that means you should do the whitespace changes in a second patch. Right, which is the way the patch is already done, I just wanted to see if making changes to the whitespace in other parts of the file (via the 2nd patch, which changes the indentation of my additions) would be considered random whitespace changes, or if it would be helpful, because it reduces filesize.. I recall a few years back (2003?) There was a discussion about removing extra whitespace at the end of lines, and someone came up with a bash/sed script to look thru the entire wine tree, strip trailing whitespace, and then somehow commit it (either with a really large diff, or by it being run on the machine that the upstream tree is on, I cant remember which). Once that was done, the tarballs shrank by at least 1-2 megs in download size, and even more uncompressed. IMHO it may be time to do that again... -- Thanks Tom
Re: Bug triage, or spam?
On 5/31/07, James Hawkins [EMAIL PROTECTED] wrote: I do read through all of your bug emails, which is exactly the problem, because I don't trust that you make the right decision on every bug, and in some cases I've had to go back and correct it. That is the issue. It has been a small number of cases.. The statement still stands. Resolved isn't resolved enough for you, but closed is...If a bug is marked resolved, I'm pretty sure it's resolved. Obviously someone that commits patches to the upstream bugzilla tree agrees with me, because otherwise, there either wouldnt be a resolved option, OR there would be a close option on the bugs that are in any state of open, which is something I have suggested before, to cut down on list traffic. No one has problems with people correcting bugs that are marked incorrectly, but of the 30 or so bugs a day that you change, this case is a small amount. If you *only* changed this class of bugs, then that would be fine. I honestly dont see why it is such an inconvenience to you for me to close the bugs that are marked resolved. You read thru every one of my emails because you dont trust my decisions, understandable, but if you are so worried about my marking a bug wrongly, then maybe you should stop developing, and strictly focus on helping clean up bugzilla, or the inverse could be true, maybe you could just trust that if I make a status change, it is the correct thing to do, and if not, then let someone else, who does have more time on their hands, catch it. Like you said, it's a small amount of bugs that are marked incorrectly. Of that percentage, I represent an even smaller amount, and like I said, I dont blindly close bugs, I read the majority of the comments, especially the ones toward the end, and the initial report, and then make a change only when I am confident that it is the right change to make.. -- Thanks Tom
Re: Bug triage, or spam?
On 5/31/07, James Hawkins [EMAIL PROTECTED] wrote: Obviously someone that commits patches to the upstream bugzilla tree agrees with me, because otherwise, there either wouldnt be a resolved option, OR there would be a close option on the bugs that are in any state of open, which is something I have suggested before, to cut down on list traffic. We are our own project, and we use bugzilla differently, just like every other project out there has their own practices. Ok. I honestly dont see why it is such an inconvenience to you for me to close the bugs that are marked resolved. You read thru every one of my emails because you dont trust my decisions, understandable, but if you are so worried about my marking a bug wrongly, then maybe you should stop developing, and strictly focus on helping clean up bugzilla, huh? That doesn't make any sense. You're saying I should take even more time away from development to check your changes. No, I'm saying that if someone higher up in the heirarchy were to do this as well, then I wouldnt have as many bugs to mess with, and therefore I wouldnt be able to make as many bad changes. I'm not the only one that reads each change, and even if I were, and someone else took over, it's still wasted time on their part. I know you arent the only one who reads each change. What I'm saying is that someone who has more free time (someone who's time it wouldnt be wasting, because they can't do anything atm) should join the fun. You continue to miss the point from the very beginning: you're trying to fix a problem that doesn't exist. There's nothing wrong with old bugs being left as resolved and marking new bugs closed on a case by case basis. It's a problem to me, it may not be a problem to you, but that doesn't make it an invalid point. Marcus and Dan have both said to keep going, I'm sure others here (I'm not trying to speak for anyone, so someone else feel free to correct me if I am wrong) dont have any problem with my doing that either. And you miss the point as well, most new bugs that are marked resolved dont end up being closed on a case-by-case basis, which is why I am going back and doing that! Case in point: http://bugs.winehq.org/show_bug.cgi?id=7885 has been resolved for over a month, why was it not closed? http://bugs.winehq.org/show_bug.cgi?id=8373 has been resolved for 2 weeks, same question applies. -- Thanks Tom
Re: Bug triage, or spam?
On 5/31/07, Jesse Allen [EMAIL PROTECTED] wrote: Tom, This is how I finalize bugs: * When a bugs has decisively been fixed, by a merged patch, with test cases, or reported by user been fixed, then I close it. If it's decisively not a wine bug close invalid. * When a bug is rumored to be fixed, forgotten, or simply doesn't appear on your end, then probably resolve fixed, abandoned, or worksforme appropriately. * When the bug has been only set to resolved, and continues to have erroneous activity (i.e. commenting from random visitors that don't understand the report), then close it to discourage use of the bug. * If it's a bug that I don't know anything about, I shouldn't touch it. So the reason is, only resolved could maybe get revisited, and closed I never want to see again. Other than that, it makes no sense to me to close bugs unless there is some activity related to it (i.e. it is proved that an uncertain fix has really been fixed and the issue is done). Simply closing bugs worries me as does James. It really does need to be case-by-case, so what we know you are doing is right. Thanks for the clarification. So if it is already resolved as anything other than fixed, just leave it alone unless it continues to get activity. I still don't like it, but as with anything, majority rules, so I will stop. Sorry for the spam everyone. -- Thanks Tom
Re: Bug triage, or spam?
On 5/31/07, Jesse Allen [EMAIL PROTECTED] wrote: Remember that this is only my opinion. Other people handle things a little different. It is probably true that the two recent bugs you mentioned I would have closed, but it was only resolved. It's just how it was handled. If you want to close bugs that are clearly done with like that, it's fine by me. And I do close bug reports with other status too, if it's one I really don't want to see again. If it's clear why you are closing bugs, it's all fine by me, whatever it is. Just to clarify my reasoning, because I just now thought about the _real_ reasoning lol, I often look thru lists of bugs that are resolved as fixed and (although i havent done it in a few years, i was going to start again), post a message to ask if the reporter can confirm if it is fixed. So I was closing bugs that were invalid/abandoned/dupe/worksforme so that they wouldnt show in the lists of resolved bugs, so its less I have to sort thru -- Thanks Tom
Re: Bug triage, or spam?
On 5/31/07, Jesse Allen [EMAIL PROTECTED] wrote: Bug triage is a good idea so that stuff like this gets cleaned up. I'm not sure what everyone wants though. I guess you can figure that out :P See, my opinion of what triage is isnt the same as everyone else's.. Either way, I'll just leave resolved bugs resolved, until they stop me from finding a specific bug I'm looking for. Allow me to get a consensus. Random Closing of Resolved bugs that have no activity is not deemed helpful by everyone, even if it does help me, so would it be acceptable for me to close a few (less than 20) a day? That way I'm not spamming the list for any more than 30 mins, and I can still kind of make it easier to find the bugs that need a followup. -- Thanks Tom
Re: Pretty Build System (Take 3)
On 5/29/07, Robert Shearman [EMAIL PROTECTED] wrote: Hi, Attached is a patch that reduces the verbosity of the build system by printing the build commands in one of two formats: COMMAND output_file or COMMAND input_file - output_file I hope it meets Alexandre's requirements of not reducing portability of Wine to other platforms (including Windows). Well, after applying the patch and doing like Kirill instructed me, it looks pretty good. I just have a couple of nits, and they really can be worked out later on. 1) make depend/make install/make uninstall/make test, etc does not appear pretty 2) is there any way to prettify the entering/leaving directory messages, I think the 2.6 kernel has it that way, but I'm not 100% sure -- Thanks Tom
uninstaller: Shorten variable name length.. Whats wrong?
Just curious to know if there was something wrong with this patch, or if maybe it wasnt committed due to being a cosmetic change? If it's the latter, I'll just resend the other 2 parts of the patch (2and 3 of 3) against the code the way it is instead of changing the lines.. On 5/23/07, Tom Spear [EMAIL PROTECTED] wrote: This patch shortens the length of some lines in FetchUninstallInformation, as suggested by Peter Beutner -- Thanks Tom -- Thanks Tom 0002-uninstaller-shorten-variable-and-attribute-names-fo.patch Description: Binary data
Re: Pretty Build System (Take 3)
On 5/29/07, Robert Shearman [EMAIL PROTECTED] wrote: Hi, Attached is a patch that reduces the verbosity of the build system by printing the build commands in one of two formats: COMMAND output_file or COMMAND input_file - output_file I hope it meets Alexandre's requirements of not reducing portability of Wine to other platforms (including Windows). I get the following (on Slackware 11) while running the make depend portion of ./configure make depend make sudo make install: make[1]: Entering directory `/home/speeddy/wine-git/tools' ../tools/makedep -C. -S.. -T.. -I/usr/include/freetype2 bin2res.c fnt2bdf.c fnt2fon.c make_ctests.c makedep.c relpath.c sfnt2fnt.c make[1]: Leaving directory `/home/speeddy/wine-git/tools' make[1]: Entering directory `/home/speeddy/wine-git/tools' make[1]: `makedep' is up to date. make[1]: Leaving directory `/home/speeddy/wine-git/tools' make[1]: Entering directory `/home/speeddy/wine-git/dlls' make[2]: Entering directory `/home/speeddy/wine-git/dlls/acledit' ../../tools/makedep -C. -S../.. -T../.. main.c make[2]: Leaving directory `/home/speeddy/wine-git/dlls/acledit' make[2]: Entering directory `/home/speeddy/wine-git/dlls/acledit' make[2]: MAKERULE_MAKEDEP@: Command not found make[2]: *** [depend] Error 127 make[2]: Leaving directory `/home/speeddy/wine-git/dlls/acledit' make[1]: *** [acledit/__depend__] Error 2 make[1]: Leaving directory `/home/speeddy/wine-git/dlls' make: *** [dlls/__depend__] Error 2 -- Thanks Tom
Re: GIT access fails
On 5/24/07, Kirill K. Smirnov [EMAIL PROTECTED] wrote: Hi, Something strange happened to git - fetch fails if uses git:// protocol. In tcpdump I can see my machine and wine.codeweavers.com chating. Thus connection is OK. When I changed protocol to http:// in .git/remotes/origin, I managed to update local git repo. This is very strange - git:// does not work, but http:// does! I did not update git utility on my system. git version is 1.4.2.1. No firewalls too. Any ideas? Did you happen to update curl/libcurl? -- Thanks Tom
Re: Bugzilla outcome?
On 5/22/07, Jan Zerebecki [EMAIL PROTECTED] wrote: I'll ask what the progress is with our Bugzilla upgrade when the one who offered this comes online again. Thanks I would consider neither of hang, stuck, 100% cpu usage, freeze to be a crash. Out of curiosity, why not? I think that regardless of the result of the bug it might be useful to check that it gets debugged. So we might distinguish two boolean variables for bugs: - (1) Has it all the information that can be provided (e.g. crash log and back-trace with debug symbols)? This would also contain checking that all the fields are properly set. - (2) Do we know the cause for the bug? ( E.g. for a segmentation fault AKA crash in the application, do we know what in Wine caused it? ) Should we include in (1) if it can be reproduced independently? That would be helpful to you guys, so I see no problem with it. -- Thanks Tom
Re: Bugzilla outcome?
On 5/23/07, Jeremy Newman [EMAIL PROTECTED] wrote: On Wed, 2007-05-23 at 04:36 +0200, Jan Zerebecki wrote: I'll ask what the progress is with our Bugzilla upgrade when the one who offered this comes online again. While you are asking, make sure they are going to Bugzilla 3.0. Since that just went stable not very long ago. I may make a suggestion for upgrading. It may be easier to start a new DB for Bugzilla, and write a script that imports all the old data to the new tables. I think I would even prefer that.. Maybe it could be set to not import anything that had the abandoned or invalid flags so that we have a leaner meaner db as well? -- Thanks Tom
Re: Bugzilla outcome?
I sent this to the wrong list before, sorry for that.. Just curious to see if anything was ever done with the offer to upgrade bugzilla? On a side note, it would be nice if there was a crash keyword to make finding crashes easier -- Thanks Tom
Re: Bugzilla outcome?
On 5/22/07, Robert Shearman [EMAIL PROTECTED] wrote: Searching for Unhandled exception is too hard? Not all bugs have a proper trace, not all crashes result in the unhandled exception message (100% cpu usage being one case), not all bugs say crash (some say hang, or stuck, 100% cpu usage, freeze, etc). So If we had a crash keyword, then as people look at the new bugs, it can be added. Then people who are looking for crashes to fix or at least debug further (as I do), can find those bugs easier. -- Thanks Tom
Re: Bugzilla outcome?
On 5/22/07, James Hawkins [EMAIL PROTECTED] wrote: I don't see how a 'crash' keyword will add any utility to bugzilla. Is someone planning on focusing solely on apps that crash, and if so, why? Well, when I look thru bugzilla, that is my main focus is find bugs with crashes, and try to debug them as far as I can so that all the info is there to write a fix without a real developer having to sit there and hold a user's hands thru the dirty part. -- Thanks Tom
Re: msi ole automation: where to next?
On 5/20/07, Misha Koshelev [EMAIL PROTECTED] wrote: Hi everybody, Hope you are all having a nice Sunday. Since MSI OLE automation is a bit overwhelming, I have been using installers that have JScript/VBScript actions to guide me as to what specific functions/objects to implement next. I started out by implementing all the functions used by the Vector NTI installer (which is how I got into wine programming at all; and which was actually dependent on one of these actions for installing at the time), and for the past few weeks I have been following the functions required by the iTunes 7 installer. However, now that I have implemented all the functions used by both of these (yay), I am having trouble finding another installer that uses a lot of scripting, so I was wondering if anyone had any suggestions? It really helps to have something like that to guide/motivate me. (I googled UNHANDLED ACTION TYPE and tried the installers I found but all the installers I actually tried out didn't seem to use scripting or maybe just unimplmeneted scripting functions anymore). Alternately, is there any other use for MSI OLE automation besides installers with scriptable actions? If so, what functions would be important for that use? (I know Mike McCormack talked at one point about using these interfaces to run custom actions in a separate process, but I am not really clear as to why automation is really necessary for this and, if so, how exactly it would be used, since all the automation functions are just wrappers around existing functions anyway.) I'm not sure if this is OLE or not, but AutoIt installers use a lot of scripting and can be wrapped around msi.. I would post locations of where to find quite a few here on the list, but I'm afraid of being flamed for spamming again, so look for my next email. -- Thanks Tom
Re: Direct X 10 game demo
On 5/17/07, Michael Lothian [EMAIL PROTECTED] wrote: In my opinion games will check for DX version on Windows version. Eventually once enough people have vista they will release DX10 on XP. Probably about a month before the wine version is ready for XP. What do you guys think? Am I just being cynical? I think so. According to everything I have read DX10 is going to be tied in with the low level stuff even more so than DX9 was, and theres a lot of new low-level stuff for it to tie into that was not there with XP.. They would have to backport all of that stuff from Vista to XP, release it as either a Service Pack 3 or as XP Second Edition (groan), and then release DX10 for XP.. Now what they could/might do is offer a scaled down version of DX10 for XP, but there wouldnt be much real benefit over DX9.. Just SM4.0 etc -- Thanks Tom
Re: Direct X 10 game demo
On 5/15/07, EA Durbin [EMAIL PROTECTED] wrote: The lost planet demo utilizing Direct X 10 is out, for those of you interested in developing/testing Direct X 10. I'm curious. When we do start work on DX10, are we going to set it up so that it will only work if the user uses winver ~= vista or greater, or will it be available for all winvers that are capable of supporting it? I'd like to see (a few years from now obviously), if it would be possible to use wine's directx to play a dx10 game on winxp, since Ms is not going to make dx10 for xp.. -- Thanks Tom
Re: Direct X 10 game demo
On 5/16/07, Stefan Dösinger [EMAIL PROTECTED] wrote: snip Past that I don't see any reason why they would code additional version checks after that. All the games that I've seen don't do that. snip I don't see a reason why it needs* to check the version of windows. Those are windows apps, keep that in mind ;-) Never expect sane behavior from them :-) So true ;-) Unfortunately, like I said MS is not making DX10 compatible with XP, which really sucks because imo Vista is the biggest piece (of junk) ever and I personally _probably_ wont 'upgrade' my windows machine to it even after Service Pack 2 comes out. So since MS is going to do that, I expect to see a lot of games (at least at the beginning) that have compatibility code, where it checks the windows version at start to determine whether to run in dx10 mode, or not, due to windows users who upgrade and dont reinstall their apps. -- Thanks Tom
mentoring
Hi all, just wanted to know if anyone would _mind_ helping me with my (small) todo list. I know I've been somewhat of an annoyance lately, and for that I sincerely apologize. I'd really like to get the wine uninstaller program up to speed with the patch I have been working on, but it is not going as I had hoped. One person offered to review my patches, but I am admittedly impatient sometimes (I haven't been able to get a response from him since Thursday, and I had hoped for one by Friday night), and all in all, I just dont like having something so small take so long. I am making changes to code I didn't intend to change, which I don't mind doing, but then when I make the change, I get told to do it 3 different ways, and none of the ways I submit gets committed, even after being told that it looks ready to go in. I'd like someone who is really in touch with what Alexandre is looking for to help me. I only have 4 patches, all less than 50 lines each to get this thing in, and I will be done submitting patches for a while, until I learn more about the code (maybe months or even years).. I will take whatever advice is given me (unless it is to not bother with the patch), and really appreciate the help. If you dont mind helping me, please post a note on my bug about the patch http://bugs.winehq.org/show_bug.cgi?id=8364 -- Thanks Tom
Re: mentoring
On 5/14/07, Stefan Dösinger [EMAIL PROTECTED] wrote: Hi Tom! I am afraid you have come accross an unmaintained part of wine. There is maybe nobody who can competently help you. Nobody minds helping you, but nobody can help you. What happens then is that you write a mail and get no answers. You get frustrated(not your fault), write a mail in which you complain nobody has answered. 3 persons who do no know the part of wine you're modifying reply 3 different answers. Discussions start, and then AJ stops applying patches until the discussions have come to a conclusion. Ok, thank you very much. That was exactly what I needed to hear. Thats just what I think has happened, but I don't know the uninstaller eiter. So count this as a 4th answer to why does nobody answer me :-) It wasnt so much a question of why does nobody answer me.. I have been getting responses, it's just that I was not getting anywhere with the responses I was getting. I have all of the information I need I think to be able to finally get this handled now.. As for finding out what aj disliked about your patches my experience is that it is best to ask him on irc(#winehackers on freenode). But I also know that aj stops looking at patches if there is way too much dispute about them. In this case you should send single patches to wine-patches starting that you want to start the process over. Of course the patches should have the suggestions implemented that were made before, or if not a good reason why you have chosen to do otherwise(Not all suggestions are good). My last question is this: What is an acceptable amount of time I should wait before asking AJ what is wrong with a patch? I see that patches get submitted and then the next day they are committed oftentimes. However I would like I am pressuring him if I submit a patch, it didnt get committed the next day, and then I shortly after ask him what was wrong. So, should I wait until day 2 or so? Unfortunately my company firewall blocks all of the IRC ports, so I can't get onto freenode from work, and by the time I get home, most times #winehackers (well #winehq anyways) is dead. I'll check in winehackers next time.. Thank you very much for putting my fears to rest. Now for the arduous task of starting over heheh. -- Thanks Tom
Re: Internet explorer not detected
On 5/12/07, Stefan Dösinger [EMAIL PROTECTED] wrote: really %SYSTEMDRIVE%\Program Files\Internet Explorer, right? I don't really feel sorry for apps that assume everything is running Program Files isn't hardcoded either, like on my German locale its Programme. Which would be done as %PROGRAMFILES%, anyways, not %SYSTEMDRIVE%\Program Files -- Thanks Tom
Re: posting advertisements
On 5/10/07, Stefan Dösinger [EMAIL PROTECTED] wrote: It's a gmail account. Does gmail automatically append advertisements? As far as I know, gmail does not. This mail will tell us... Like I said I thought I might have spyware, I have since managed to get internet access (as opposed to ethernet only access) for my linux machine here at work, so no more adverts.. Sorry for the spam -- Thanks Tom
Re: [PATCH] Uninstaller: Use a define'd length instead of a magic number
On 5/10/07, James Hawkins [EMAIL PROTECTED] wrote: Yes, you should make a new define. MAX_STRING_LEN is the maximum length for a string from the resource file, which is totally unrelated to the limit for the length of a subkey name. Should the new define be put in resource.h, main.c, or somewhere else entirely? -- Thanks Tom Check out this new 3D Instant Messenger called IMVU. It's the best I have seen yet! http://imvu.com/catalog/web_invitation.php?userId=1547373from=power-email
Re: blizzard conference
On 5/10/07, Scott Ritchie [EMAIL PROTECTED] wrote: Does anyone have statistics we can present about how many Wine users playing Blizzard games there are out there? The closest thing I've ever seen was a steam-based hardware survey Valve released a couple of years ago. Even then, there were tens of thousands of Wine users playing Steam games. With Wine more popular now and World of Warcraft often working better than in Windows, it seems pretty reasonable that there would be even more Wine-based WoW players. Perhaps we could ask Blizzard to just include a simple test for Wine's registry key in their software and then keep statistics on it. I dont know about every other WoW player, but given Blizzard's history of phoning home with previous games, I dont know how comfortable I would be with even a check for a wine registry key that phones home every time I play WoW. -- Thanks Tom
Re: [PATCH] Uninstaller: Change hardcoded value for maximum subkey name length to a const.
On 5/8/07, Dan Hipschman [EMAIL PROTECTED] wrote: I'm not sure what version of C wine is strictly trying to conform to, but it's usually best to go with the lowest common denominator. This is not C89. The const keyword doesn't actually make the variable constant (I know it's not the greatest feature of C). The array size won't be evaluated until runtime, so technically this is a variable length array which weren't added until C99. You can make it C89 simply by using #define instead of declaring it as a variable. Hi. Thanks for the comment. I looked thru resource.h in the same directory, and see MAX_STRING_LEN is defined at 255. Would using that be acceptable, or should I add a new define, something like #define MAX_SUBKEY_LEN 255 ? If I should add a new define, should I add it to resource.h or directly to main.c? -- Thanks Tom Check out this new 3D Instant Messenger called IMVU. It's the best I have seen yet! http://imvu.com/catalog/web_invitation.php?userId=1547373from=power-email
Drop-in replacement of windows dll's?
I just remembered reading somewhere (wiki, perhaps?) that the ultimate goal of wine (aside from documenting the windows api, etc) was to provide open-source drop-in replacements for windows' core dll's. If that is the case, do we have a framework setup for building the wine dlls as windows .dll files, so that we could test the actual status of this? I think it would be interesting to see how much of windows will run with wine's dlls in their current state. Obviously it would be bad to test this with something like user32.dll, etc, but something like riched32 and maybe some of the common controls dlls should be ok to test with? -- Thanks Tom Check out this new 3D Instant Messenger called IMVU. It's the best I have seen yet! http://imvu.com/catalog/web_invitation.php?userId=1547373from=power-email
Re: blizzard conference
On 5/8/07, Jesse Allen [EMAIL PROTECTED] wrote: On 5/7/07, Kai Blin [EMAIL PROTECTED] wrote: On Tuesday 08 May 2007 05:34, Dan Kegel wrote: Hey, that's only an hour away, maybe I / Lei / Nigel can drop by. But what's the itchy /etc/hosts bug? Is that http://bugs.winehq.org/show_bug.cgi?id=7929 ? I would assume so. I don't know if telling them is going to fix anything, though. For new games, I hope they're not using gethostbyname(gethostname()) anyway. For the other issues that are hitting us with that bug, all seem to be valid winsock programming. I'd rather have someone talk to them about better overall wine compatibility. Right, Diablo 2 does use gethostbyname in a bad way. But I think distros should just manage the /etc/hosts file better. I kinda tend to agree, but the question on their end would be something along the lines of with most users using dynamic ip's, how do you propose that we link the hostname to the current ip? One idea would be to setup a cron job that checks ifconfig for the ip of ethx, and then changes /etc/hosts accordingly. But then how would you handle users that either a) dont have cron installed, or b) dont use an ethernet card (i.e. they use wireless, or are hooked to the cable modem via usb cable, or are using dialup).. Maybe inetd/xinetd need to be rewritten from the ground up? lol I dont think that will happen any time soon. The best bet is for wine to work with what it has and build in support somehow for each scenario as we come to it. -- Thanks Tom Check out this new 3D Instant Messenger called IMVU. It's the best I have seen yet! http://imvu.com/catalog/web_invitation.php?userId=1547373from=power-email
Re: Drop-in replacement of windows dll's?
On 5/8/07, Steven Edwards [EMAIL PROTECTED] wrote: On 5/8/07, Tom Spear [EMAIL PROTECTED] wrote: I just remembered reading somewhere (wiki, perhaps?) that the ultimate goal of wine (aside from documenting the windows api, etc) was to provide open-source drop-in replacements for windows' core dll's. If that is the case, do we have a framework setup for building the wine dlls as windows .dll files, so that we could test the actual status of this? Google around for posts by me and other about build wine with mingw. I've been meaning to document it for years but never gotten around to it. You either need to build on windows with a mingw or have two trees on Linux, one with the host tools and the other configured for the cross-compiler. That would be the framework. I appreciate the response. Is this something that either now or in the future will be officially supported by wine, or even a true goal of wine, or is it more of a side project for you and some others? -- Thanks Tom Check out this new 3D Instant Messenger called IMVU. It's the best I have seen yet! http://imvu.com/catalog/web_invitation.php?userId=1547373from=power-email
(no subject) (WAS: Re: Drop-in replacement of windows dll's?)
On 5/8/07, Brian Vincent [EMAIL PROTECTED] wrote: On 5/8/07, Tom Spear [EMAIL PROTECTED] wrote: That would be the framework. I appreciate the response. Is this something that either now or in the future will be officially supported by wine, or even a true goal of wine, or is it more of a side project for you and some others? It's not an explicit goal, but it does work, and it is a configuration that we want to keep working. It's just not a configuration that gets tested every day, so there might be some little issues. Previously ReactOS was the big consumer of Wine DLL's as PE. Ok, thanks for the input. Thanks to all of the developers for their hard work, and dedication. Unfortunately I am going to have to mute myself on the list (I can hear the sighs of relief now), as I dont think I am getting anywhere trying to get even trivial patches committed. I sent a patch over a week ago to correct a trace message (not the one for winecfg, but one for uninstaller), and it never got committed. Then I get told to send all of my patches to a developer privately for review before I send them to the patches list, and when I send them to that developer, I _basically_ get told that I should bother submitting them. I know I'm not good enough to even write conformance tests right now, which is why I dont work on the major API's or even on the conformance tests, but sheesh, what do I have to do to get one problematic trace fixed? I'll send that patch one more time, just to make sure it got to the list, and I would appreciate someone offering some comment if there is a reason that it should not be committed. -- Thanks Tom Check out this new 3D Instant Messenger called IMVU. It's the best I have seen yet! http://imvu.com/catalog/web_invitation.php?userId=1547373from=power-email
Re: Drop-in replacement of windows dll's?
On 5/8/07, Thorsten Kani [EMAIL PROTECTED] wrote: Hi Tom Please read through wwn244 and 245. If i remenber correctly, at that time common controls from wine wont work out of the box. I had to apply reactos patches to them and had to binary patch sfc.dll on windows. Thanks for the info. I have sfc completely removed from my windows system (as in sfc.dll doesnt even exist) via a nifty little windows component remover. I didnt expect for the common controls to work out of the box, but it would be interesting to see how close we come. I'll try the various things mentioned here when I get time (dont know when that will be), and post results on the wiki, unless someone beats me to the punch. -- Thanks Tom Check out this new 3D Instant Messenger called IMVU. It's the best I have seen yet! http://imvu.com/catalog/web_invitation.php?userId=1547373from=power-email
Re: Installer already running
On 5/7/07, Pavel Troller [EMAIL PROTECTED] wrote: Dear Vitaliy, I'm very sorry for my total stupidity, ignorance and maybe even debility, but I can't find nothing common between Bug# 219 and my question. The PhysX installer doesn't check for any *ICE components present in the system, doesn't output any strings to wine console, which are described in the bug desc, and the message The installer is already running, which is typical for my problem, is not mentioned in the bug description at all. So please be patient with me and explain me in the deeper detail, why are you thinking that I'm asking for #219. With regards, Pavel Troller I agree, this is not an issue with Copy Protection, this is an issue where the installer thinks that the installer is already running. A couple of questions -Have you tried running the PhysX installer manually, after the game is installed? -Have you tried running wineboot to make sure any run entries are cleared and then running the PhysX installer manually? -Any idea what brand of installer it is (InstallShield, MSI, WISE, or some other)? If you aren't sure, I can always take a look at the installer, if you will bzip and send to me (less than 10mb). -- Thanks Tom Check out this new 3D Instant Messenger called IMVU. It's the best I have seen yet! http://imvu.com/catalog/web_invitation.php?userId=1547373from=power-email
Re: Installer already running
On 5/7/07, Vitaliy Margolen [EMAIL PROTECTED] wrote: Pavel Troller wrote: it chokes on SECDRV.SYS). That is all you need to see to know this is SafeDisk. (for anybody who would like to try this too, the game itself has to be patched by nocd crack first, otherwise it chokes on SECDRV.SYS). If you read that part again, it says the game itself needs the patch. He is writing about an installer issue with the installer for the PhysX card. -- Thanks Tom Check out this new 3D Instant Messenger called IMVU. It's the best I have seen yet! http://imvu.com/catalog/web_invitation.php?userId=1547373from=power-email
Re: Installer already running
On 5/7/07, Pavel Troller [EMAIL PROTECTED] wrote: Hi Tom, there is the installer file. With regards, Pavel Troller Hi Pavel, I looked into the file. It is a 7-zip compressed Nullsoft installer. What that means is that if you decide to file a bug on bugzilla (highly recommended), make sure to include that information in the summary or body of the bug somewhere, so that others with the same problem with this type of installer can find your bug easily, and so that developers who are looking for this type of installer to test patches against can contact you to get it... With that being said: I downloaded the file on my linux box, and was able to run it all the way up to the point of hitting next to actually install, without any problems. Here are a couple of things you can try: - Run wineboot to make sure all runonce entries are cleared from the registry, and then restart the computer, to make sure that all running copies of wine are cleared. - If that does not work, rename ~/.wine to something like ~/.wine.bak and then run winecfg and re-setup all of your options. Then try running the PhysX installer again. If that does not fix it, you will probably have to do a fully clean install of wine, where you need to delete all copies of wine on your system, and everything that you have installed in wine (I'm trying to avoid you having to do this, for obvious reasons), and then reinstall it all, starting with JTF2 and the PhysX drivers. If that still does not fix the problem, then there is a misconfig somewhere else in the system, or you somehow have a stale wineserver process hanging around.. Hope that helps, and maybe someone else has some other things for you to try..? -- Thanks Tom Check out this new 3D Instant Messenger called IMVU. It's the best I have seen yet! http://imvu.com/catalog/web_invitation.php?userId=1547373from=power-email
Re: blizzard conference
On 5/7/07, Bryan Haskins [EMAIL PROTECTED] wrote: Oi, If only I could go =[ I would like to point out to any players who would go there on the behald of the wine project wink wink nudge nudge, that they generally give away some sick in game prizes! If you go and arent a player, you could easily sell the codes they give out (at least last time) for a heap of cash to go towards the project in some form. I'm definitely going to try to go, but I can't promise anything as yet, due to that being a 2-day drive and on a Friday and Saturday, and also due to having to schedule it with work first. Once (and if) I confirm that I am going, could someone fill me in on the details of the /etc/hosts bug, or point me to a link that has all of the information I'll need, and I'll be happy to mention it. If you would rather someone more experienced with developing bring it up, just let me know, and I'll keep my mouth shut. -- Thanks Tom Check out this new 3D Instant Messenger called IMVU. It's the best I have seen yet! http://imvu.com/catalog/web_invitation.php?userId=1547373from=power-email
Re: USB device support in wine
On 5/3/07, Jon Burgess [EMAIL PROTECTED] wrote: Hi, I have found some talk of implementing USB device support in wine in this list some time ago (2005), but as far as I know, nothing ever came of it. I would perhaps be interested in getting this going again, as I have an application (Serato Scratch Live: http://www.rane.com/scratch.html) for which the software appears to run ok under wine (not that I am able to test much of its functionality on the other hand), but is utterly useless without support for its associated USB hardware device. One thing is that you need to make sure that the hardware is detected and usable by Linux prior to trying to implement support for it in wine. If linux can't see it, then wine won't be able to, just like if you dont have the drivers installed in windows, then you wont be able to use it within the application on windows. -- Thanks Tom Check out this new 3D Instant Messenger called IMVU. It's the best I have seen yet! http://imvu.com/catalog/web_invitation.php?userId=1547373from=power-email
Re: Linuxworld San Francisco?
On 5/4/07, Dan Kegel [EMAIL PROTECTED] wrote: I was thinking of volunteering, too. On 5/4/07, Scott Ritchie [EMAIL PROTECTED] wrote: Oooh, I'll do it! Now I just need to come up with a specific topic other than Wine Thanks, Scott Ritchie On Fri, 2007-05-04 at 10:30 -0500, Jeremy White wrote: Well, we'll be there with a booth. I think I've given enough Wine talks through the years at LinuxWorld that they're tired of me :-/. But, to be honest, I don't think I submitted a proposal this year. I think if someone else wanted to give the talk, they'd probably appreciate a fresh face. Cheers, Jeremy -Original Message- From: [EMAIL PROTECTED] [mailto:[EMAIL PROTECTED] On Behalf Of Dan Kegel Sent: Friday, May 04, 2007 10:13 AM To: wine-devel@winehq.org Subject: Linuxworld San Francisco? Is anyone going to Linuxworld, August 6-9 in San Francisco? A guy working on a possible large migration said he'd be there, and was wondering if there's going to be a Wine BOF. I might go just to chat with him. Also, I see no Wine-related talks on the program. http://www.linuxworldexpo.com/live/12/events/12SFO07A/conferen ce/tracks If all the great stuff I expect actually happens this summer, there's going to be a lot to talk about, seems like a good year for a Wine talk there...? -- Wine for Windows ISVs: http://kegel.com/wine/isv I'd like to try to be there to listen to the talk, but it depends on finances and if work would let me off.. -- Thanks Tom Check out this new 3D Instant Messenger called IMVU. It's the best I have seen yet! http://imvu.com/catalog/web_invitation.php?userId=1547373from=power-email
Unsecured API functions
I was writing up a Hello World with input program for a demonstration for a non-developer coworker last week, and used the unsecured getch() and got the standard warning about how it was unsecured and dangerous to use that. That prompted me to look up the basic secured functions on the MS website, and compare to wine code. According to MSDN, things like gets have been replaced with gets_s. However, as far as I can tell, wine still only implements gets for Windows programs to use.. Do we implement secured versions of other functions, and if not, how come? -- Thanks Tom Check out this new 3D Instant Messenger called IMVU. It's the best I have seen yet! http://imvu.com/catalog/web_invitation.php?userId=1547373from=power-email
Re: Unsecured API functions
On 5/3/07, Robert Shearman [EMAIL PROTECTED] wrote: Tom Spear wrote: I was writing up a Hello World with input program for a demonstration for a non-developer coworker last week, and used the unsecured getch() and got the standard warning about how it was unsecured and dangerous to use that. That prompted me to look up the basic secured functions on the MS website, and compare to wine code. According to MSDN, things like gets have been replaced with gets_s. However, as far as I can tell, wine still only implements gets for Windows programs to use.. Do we implement secured versions of other functions, and if not, how come? Q: Why doesn't Wine implement X? A: Because not many programs use it and no-one has felt interested in implementing it for fun. So in other words, most programs use insecure functions (like gets) instead of using secure functions (like gets_s), leaving themselves vulnerable to all sorts of buffer overflows? I wonder if microsoft doesn't silently convert gets calls to gets_s calls, then, and maybe didn't document that? Otherwise I assume there would be thousands of buffer overflows that (malicious) people would exploit. I understand that most programs dont use either of those functions, but there are others that are used by nearly every program that ms deprecated in favor of secure versions. -- Thanks Tom Check out this new 3D Instant Messenger called IMVU. It's the best I have seen yet! http://imvu.com/catalog/web_invitation.php?userId=1547373from=power-email
Re: Unsecured API functions
On 5/3/07, Marcus Meissner [EMAIL PROTECTED] wrote: wine is not using gets() at all, insofar there is no risk from it. That much I knew, however we do use strcpy (especially in msi), and that is another one that has been deprecated (banned).. See http://msdn2.microsoft.com/en-us/library/bb288454.aspx for the complete list.. It would be quite hard to convert gets - gets_s by magic ;) hmm, I thought so, and re-reading the page, it appears that it is actually more of a proposal, than a list of api's that have actually already been deprecated, however if msdn has an article from the sdl that pushes for the deprecation of non-strsafe functions, I think we should take that seriously, and at least investigate the difficulty (I'm not pushing for it to be replaced anywhere in the code right now, because we are already spread too thin). -- Thanks Tom Check out this new 3D Instant Messenger called IMVU. It's the best I have seen yet! http://imvu.com/catalog/web_invitation.php?userId=1547373from=power-email
Re: Unsecured API functions
On 5/3/07, Kai Blin [EMAIL PROTECTED] wrote: On Thursday 03 May 2007 23:16, Tom Spear wrote: Noone should use gets(). There are lots of better alternatives. For the other deprecated functions, there are ways to check that the input is valid before calling it, iirc. I agree that nobody uses it, and that there are better alternatives, I was just using it as an example. You are correct, though about there being ways to check the input. As far as imnplementing the secured functions in Wine, I have yet to see a program that's failing because it tries to use one of them. I do agree that we shouldnt implement it until a program fails. I know I worded the email wrong. I was actually just wondering if we had any plans to implement them in the future and if we had the resources to do that. -- Thanks Tom Check out this new 3D Instant Messenger called IMVU. It's the best I have seen yet! http://imvu.com/catalog/web_invitation.php?userId=1547373from=power-email
Re: Any way to get the current code besides git?
On 5/2/07, Alexandre Julliard [EMAIL PROTECTED] wrote: Tom Spear [EMAIL PROTECTED] writes: Haven't killed it yet and been waiting for 2 hours.. I'm on a T-3 afaik. Last line that prints before it just sits is: walk 4eea356e2d39f1a958afb4d8f5b54381e8972ecf This turns out to be a bug in libcurl. It's fixed in libcurl = 7.16. -- Alexandre Julliard [EMAIL PROTECTED] Thanks, I just got the new curl/libcurl, and it just went straight past the walk 4eea I've just added a note to the GitWine wiki page about it. Just so I know, what is the longest I should expect to wait between messages before deciding that it has hung, because it seems to have hung 2 lines down now: walk 4eea356e2d39f1a958afb4d8f5b54381e8972ecf Getting pack 03d06cfa82414ad54614fcb3c633633bf3bfed49 which contains 7a837e5386cbc23fbadf279073222423a57b4afb But I haven't killed it, and I've only waited about 5 minutes so far. -- Thanks Tom Check out this new 3D Instant Messenger called IMVU. It's the best I have seen yet! http://imvu.com/catalog/web_invitation.php?userId=1547373from=power-email
Patch question
Just curious if there was something wrong with the regedit patch I sent on Friday, or if maybe there is a particular reason that the /C was left out of the code in the first place, that I am missing, and would help me to understand. http://www.winehq.org/pipermail/wine-patches/2007-April/038652.html -- Thanks Tom Check out this new 3D Instant Messenger called IMVU. It's the best I have seen yet! http://imvu.com/catalog/web_invitation.php?userId=1547373from=power-email
Re: Any way to get the current code besides git?
On 5/2/07, Tom Spear [EMAIL PROTECTED] wrote: On 5/2/07, Alexandre Julliard [EMAIL PROTECTED] wrote: Tom Spear [EMAIL PROTECTED] writes: Haven't killed it yet and been waiting for 2 hours.. I'm on a T-3 afaik. Last line that prints before it just sits is: walk 4eea356e2d39f1a958afb4d8f5b54381e8972ecf This turns out to be a bug in libcurl. It's fixed in libcurl = 7.16. -- Alexandre Julliard [EMAIL PROTECTED] Thanks, I just got the new curl/libcurl, and it just went straight past the walk 4eea I've just added a note to the GitWine wiki page about it. Just so I know, what is the longest I should expect to wait between messages before deciding that it has hung, because it seems to have hung 2 lines down now: walk 4eea356e2d39f1a958afb4d8f5b54381e8972ecf Getting pack 03d06cfa82414ad54614fcb3c633633bf3bfed49 which contains 7a837e5386cbc23fbadf279073222423a57b4afb But I haven't killed it, and I've only waited about 5 minutes so far. Well I waited a while longer, and went about working, and just now checked. It finished. Thanks Alexandre, I never would have figured out it was a race condition, nor that it was libcurl that needed to be upgraded. I just dont have the time to subscribe to another devel list to where I could have reported the problem back when it started. -- Thanks Tom Check out this new 3D Instant Messenger called IMVU. It's the best I have seen yet! http://imvu.com/catalog/web_invitation.php?userId=1547373from=power-email
Re: http access ot GIT?
On 5/1/07, Ben Taylor [EMAIL PROTECTED] wrote: Found myself behind an http-proxy and can't seem to find the magic to use the http proxy to update my git archive. Is there a way to use an http proxy (squid) with wine's git? What error are you getting? If no error, then if you to try to clone the whole repository, does it get to a point and then just sit there? -- Thanks Tom Check out this new 3D Instant Messenger called IMVU. It's the best I have seen yet! http://imvu.com/catalog/web_invitation.php?userId=1547373from=power-email
Re: http access ot GIT?
On 5/1/07, Ben Taylor [EMAIL PROTECTED] wrote: -- Original message -- From: Tom Spear [EMAIL PROTECTED] On 5/1/07, Ben Taylor [EMAIL PROTECTED] wrote: Found myself behind an http-proxy and can't seem to find the magic to use the http proxy to update my git archive. Is there a way to use an http proxy (squid) with wine's git? What error are you getting? If no error, then if you to try to clone the whole repository, does it get to a point and then just sit there? fatal: Unable to look up source.winehq.org (port 9418) (Name or service not konwn) Cannot get the repository state from git://source.winehq.org/git/wine.git I've setup an http_proxy according to the documentation I've found on the web. There's just not a lot out there that is specific to this conifguration. Did you try changing the git:// to http:// ? That is supposed to work thru port 80 with or without squid. YMMV -- Thanks Tom Check out this new 3D Instant Messenger called IMVU. It's the best I have seen yet! http://imvu.com/catalog/web_invitation.php?userId=1547373from=power-email
Any way to get the current code besides git?
I'm having an issue using git at work, where it will start downloading the packs, and then just stalls out at the same pack every time.. This has occurred since I first started working here Maybe someone knows the problem, and a solution? If not, is there a way to get git to show file names instead of md5 sums, like cvs did, back when we used it, so I can at least know what file is causing the problem? If there is no solution, then maybe someone does know where I can get current code aside from git? -- Thanks Tom Check out this new 3D Instant Messenger called IMVU. It's the best I have seen yet! http://imvu.com/catalog/web_invitation.php?userId=1547373from=power-email
Re: Any way to get the current code besides git?
On 5/1/07, Alexandre Julliard [EMAIL PROTECTED] wrote: Tom Spear [EMAIL PROTECTED] writes: I'm having an issue using git at work, where it will start downloading the packs, and then just stalls out at the same pack every time.. This has occurred since I first started working here One of the packs is pretty large, most likely you didn't wait long enough. Haven't killed it yet and been waiting for 2 hours.. I'm on a T-3 afaik. Last line that prints before it just sits is: walk 4eea356e2d39f1a958afb4d8f5b54381e8972ecf Just to dispel any idea that there is a size limit, I can download the entire dvd iso of Slackware 11 in about 20 minutes, with no problems, but that is from an http/ftp server. -- Thanks Tom Check out this new 3D Instant Messenger called IMVU. It's the best I have seen yet! http://imvu.com/catalog/web_invitation.php?userId=1547373from=power-email
Re: quicken 2007 ungodly slow
On 4/29/07, Dan Kegel [EMAIL PROTECTED] wrote: (vmlinuz won't do, nor do I know how to uncompress it.) It's a bzImage, unless ubuntu changed the default make rules. So, (just a guess), bunzip2 might do the trick. -- Thanks Tom Check out this new 3D Instant Messenger called IMVU. It's the best I have seen yet! http://imvu.com/catalog/web_invitation.php?userId=1547373from=power-email
Re: [programs/regedit] Fix command line processing for /? patch
On 4/27/07, Alexandre Julliard [EMAIL PROTECTED] wrote: You don't want to do that, even if you print usage, an invalid switch needs to cause an error. So should we fprintf the usage statement and exit(1); or should we print both the usage, and the error for the invalid switch. Unfortunately I don't have a copy of win98's regedit, and winxp's regedit does not accept command line switches (I have tried), so I can't check (easily) how the native regedit that ours is supposed to be command-line compatible with, does it. -- Thanks Tom Check out this new 3D Instant Messenger called IMVU. It's the best I have seen yet! http://imvu.com/catalog/web_invitation.php?userId=1547373from=power-email
Re: [programs/regedit] Fix command line processing for /? patch
On 4/27/07, Tom Spear [EMAIL PROTECTED] wrote: On 4/27/07, Alexandre Julliard [EMAIL PROTECTED] wrote: You don't want to do that, even if you print usage, an invalid switch needs to cause an error. So should we fprintf the usage statement and exit(1); or should we print both the usage, and the error for the invalid switch. Unfortunately I don't have a copy of win98's regedit, and winxp's regedit does not accept command line switches (I have tried), so I can't check (easily) how the native regedit that ours is supposed to be command-line compatible with, does it. Neither. I just figured out why /? is not working. I can't believe that I did not think of this before. The shell processes an unescaped ?. In order for it to work, I have to use /\? .. If I just do /? then s ends up being /c /d and so ch ends up being c, which makes chu C (which is missing from the ignored switches if statement).. With the old code: regedit /c returns regedit: Undefined switch /C! regedit /? returns regedit: Undefined switch /C! regedit /C returns regedit: Undefined switch /C! The usage shows /c and /C as being valid switches. So.. I changed the line if (chu == 'S' || chu == 'V') { to if (chu == 'S' || chu == 'V' || chu == 'C') { and so now: regedit /c opens regedit regedit /? shows the usage statement ! regedit /C opens regedit Is this a proper fix? Can I submit it? The diff is attached.. -- Thanks Tom Check out this new 3D Instant Messenger called IMVU. It's the best I have seen yet! http://imvu.com/catalog/web_invitation.php?userId=1547373from=power-email regedit.cmdline.patch Description: Binary data
Re: [programs/regedit] Fix command line processing for /? patch
On 4/27/07, Tom Spear [EMAIL PROTECTED] wrote: On 4/27/07, Tom Spear [EMAIL PROTECTED] wrote: On 4/27/07, Alexandre Julliard [EMAIL PROTECTED] wrote: You don't want to do that, even if you print usage, an invalid switch needs to cause an error. So should we fprintf the usage statement and exit(1); or should we print both the usage, and the error for the invalid switch. Unfortunately I don't have a copy of win98's regedit, and winxp's regedit does not accept command line switches (I have tried), so I can't check (easily) how the native regedit that ours is supposed to be command-line compatible with, does it. Neither. I just figured out why /? is not working. I can't believe that I did not think of this before. The shell processes an unescaped ?. In order for it to work, I have to use /\? .. If I just do /? then s ends up being /c /d and so ch ends up being c, which makes chu C (which is missing from the ignored switches if statement).. With the old code: regedit /c returns regedit: Undefined switch /C! regedit /? returns regedit: Undefined switch /C! regedit /C returns regedit: Undefined switch /C! The usage shows /c and /C as being valid switches. So.. I changed the line if (chu == 'S' || chu == 'V') { to if (chu == 'S' || chu == 'V' || chu == 'C') { and so now: regedit /c opens regedit regedit /? shows the usage statement ! regedit /C opens regedit Is this a proper fix? Can I submit it? The diff is attached.. Further testing shows apparently not, because s is being shifted to the /d now. Any ideas on how to properly capture /? vs having to escape it like /\? ?? -- Thanks Tom Check out this new 3D Instant Messenger called IMVU. It's the best I have seen yet! http://imvu.com/catalog/web_invitation.php?userId=1547373from=power-email
Re: [programs/regedit] Fix command line processing for /? patch
On 4/27/07, Tom Spear [EMAIL PROTECTED] wrote: On 4/27/07, Tom Spear [EMAIL PROTECTED] wrote: On 4/27/07, Tom Spear [EMAIL PROTECTED] wrote: On 4/27/07, Alexandre Julliard [EMAIL PROTECTED] wrote: You don't want to do that, even if you print usage, an invalid switch needs to cause an error. So should we fprintf the usage statement and exit(1); or should we print both the usage, and the error for the invalid switch. Unfortunately I don't have a copy of win98's regedit, and winxp's regedit does not accept command line switches (I have tried), so I can't check (easily) how the native regedit that ours is supposed to be command-line compatible with, does it. Neither. I just figured out why /? is not working. I can't believe that I did not think of this before. The shell processes an unescaped ?. In order for it to work, I have to use /\? .. If I just do /? then s ends up being /c /d and so ch ends up being c, which makes chu C (which is missing from the ignored switches if statement).. With the old code: regedit /c returns regedit: Undefined switch /C! regedit /? returns regedit: Undefined switch /C! regedit /C returns regedit: Undefined switch /C! The usage shows /c and /C as being valid switches. So.. I changed the line if (chu == 'S' || chu == 'V') { to if (chu == 'S' || chu == 'V' || chu == 'C') { and so now: regedit /c opens regedit regedit /? shows the usage statement ! regedit /C opens regedit Is this a proper fix? Can I submit it? The diff is attached.. Further testing shows apparently not, because s is being shifted to the /d now. Any ideas on how to properly capture /? vs having to escape it like /\? ?? And even more further testing shows that my distro is wonky. echo /? for me echos /c /d to my terminal Guess it's time to email Pat Volkerding to find out if he can fix it for the next slack release. -- Thanks Tom Check out this new 3D Instant Messenger called IMVU. It's the best I have seen yet! http://imvu.com/catalog/web_invitation.php?userId=1547373from=power-email
Re: [programs/regedit] Fix command line processing for /? patch
On 4/27/07, Carl-Daniel Hailfinger [EMAIL PROTECTED] wrote: On 27.04.2007 21:58, Tom Spear wrote: On 4/27/07, Tom Spear [EMAIL PROTECTED] wrote: On 4/27/07, Tom Spear [EMAIL PROTECTED] wrote: On 4/27/07, Tom Spear [EMAIL PROTECTED] wrote: On 4/27/07, Alexandre Julliard [EMAIL PROTECTED] wrote: You don't want to do that, even if you print usage, an invalid switch needs to cause an error. So should we fprintf the usage statement and exit(1); or should we print both the usage, and the error for the invalid switch. Unfortunately I don't have a copy of win98's regedit, and winxp's regedit does not accept command line switches (I have tried), so I can't check (easily) how the native regedit that ours is supposed to be command-line compatible with, does it. Neither. I just figured out why /? is not working. I can't believe that I did not think of this before. The shell processes an unescaped ?. In order for it to work, I have to use /\? .. If I just do /? then s ends up being /c /d and so ch ends up being c, which makes chu C (which is missing from the ignored switches if statement).. With the old code: regedit /c returns regedit: Undefined switch /C! regedit /? returns regedit: Undefined switch /C! regedit /C returns regedit: Undefined switch /C! The usage shows /c and /C as being valid switches. So.. I changed the line if (chu == 'S' || chu == 'V') { to if (chu == 'S' || chu == 'V' || chu == 'C') { and so now: regedit /c opens regedit regedit /? shows the usage statement ! regedit /C opens regedit Is this a proper fix? Can I submit it? The diff is attached.. Further testing shows apparently not, because s is being shifted to the /d now. Any ideas on how to properly capture /? vs having to escape it like /\? ?? And even more further testing shows that my distro is wonky. echo /? for me echos /c /d to my terminal Guess it's time to email Pat Volkerding to find out if he can fix it for the next slack release. That is expected if you have one-letter directories in your root folder. Try echo /? Wonderful. Thanks for that. Now I have renamed my symlinks to ~/.wine/drive_c and mnt/d so that they are no longer one letter, and regedit works as it should! All this trouble over a stupid directory name.. Although I do understand why and how it works, now.. That is extremely annoying, but since regedit is the only wine program that I know of that uses /? it's not too big of a deal.. -- Thanks Tom Check out this new 3D Instant Messenger called IMVU. It's the best I have seen yet! http://imvu.com/catalog/web_invitation.php?userId=1547373from=power-email
Re: wine killing X?
On 4/27/07, Juan Lang [EMAIL PROTECTED] wrote: I just upgraded to the latest wine version and tried wineprefixcreate (no .wine directory), and it kills X. Nvidia drivers 1.0.9755. Worked fine for me. But I'm not using nvidia drivers or an nvidia card. -- Thanks Tom Check out this new 3D Instant Messenger called IMVU. It's the best I have seen yet! http://imvu.com/catalog/web_invitation.php?userId=1547373from=power-email
{Resend} [programs/uninstaller (Try 6)] Scan HKCU
Scan HKCU for uninstall entries. Separated out the trace fix, moved the HeapAlloc for entries outside the for loop, realigned the 2nd line of code for a trace. This patch obsoletes all previous ones. -- Thanks Tom Check out this new 3D Instant Messenger called IMVU. It's the best I have seen yet! http://imvu.com/catalog/web_invitation.php?userId=1547373from=power-email uninstaller.patch Description: Binary data
Trace fix
Just curious if Alexandre missed my patches for regedit and my trace fix for uninstaller, or if there is a backlog of commits to be done, or if he even plans to commit them, and if not, why.. http://www.winehq.org/pipermail/wine-patches/2007-April/038566.html - Spelling fixes for regedit http://www.winehq.org/pipermail/wine-patches/2007-April/038568.html - Correct parsing of /? and unknown parameters on command line for regedit http://www.winehq.org/pipermail/wine-patches/2007-April/038569.html - Fix for a trace in uninstaller (Make sure it shows the current selection, not the old one) I am aware that my attempt 6 to scan hkcu was not attached to the original email for it, so I resent that, but expect a day or 2 delay for it to be committed, but the others, I figured would have been by now. -- Thanks Tom Check out this new 3D Instant Messenger called IMVU. It's the best I have seen yet! http://imvu.com/catalog/web_invitation.php?userId=1547373from=power-email
Re: Trace fix
On 4/26/07, Alexandre Julliard [EMAIL PROTECTED] wrote: That's particularly true for cosmetic patches like typo fixes; if I have to spend more than 10 seconds on such a patch, chances are it will end up in /dev/null. Ok, so then if I resend them now, will you take a look? I believe I have fixed every problem with each one, as I havent gotten any more comments about something being borked. -- Thanks Tom Check out this new 3D Instant Messenger called IMVU. It's the best I have seen yet! http://imvu.com/catalog/web_invitation.php?userId=1547373from=power-email
msvcp60 and bug 7679
http://bugs.winehq.org/show_bug.cgi?id=7679 - IMVU 3D Avatar Chat client crashes Just curious, since we don't have a builtin msvcp60.dll and I seriously doubt that we will implement it any time soon (if ever), should we try to support native, like we do with mfc, or would this bug be a wontfix? If we should try to at least get native working, I would like to _try_ to at least figure out what is wrong that causes the app to crash in msvcp60.dll and could use some pointers if someone doesn't mind. -- Thanks Tom Check out this new 3D Instant Messenger called IMVU. It's the best I have seen yet! http://imvu.com/catalog/web_invitation.php?userId=1547373from=power-email
Re: msvcp60 and bug 7679
On 4/26/07, Tom Spear [EMAIL PROTECTED] wrote: http://bugs.winehq.org/show_bug.cgi?id=7679 - IMVU 3D Avatar Chat client crashes Just curious, since we don't have a builtin msvcp60.dll and I seriously doubt that we will implement it any time soon (if ever), should we try to support native, like we do with mfc, or would this bug be a wontfix? If we should try to at least get native working, I would like to _try_ to at least figure out what is wrong that causes the app to crash in msvcp60.dll and could use some pointers if someone doesn't mind. I ran the program thru winedbg, and I keep hitting a wall. It may just be my inability to really use a debugger the way it is meant to, but I keep getting page fault on write access in the debugger in kernel32/virtual.c line 656. That line is in IsBadWritePtr, which (from what I can tell) is supposed to generate a page fault and catch it if there is no write access to a particular memory block. Shouldn't winedbg skip over this then? Is this a problem with that line, or is it working properly, and if it is working properly, then how do I get past that exception in winedbg, so I can get to where I need to be? -- Thanks Tom Check out this new 3D Instant Messenger called IMVU. It's the best I have seen yet! http://imvu.com/catalog/web_invitation.php?userId=1547373from=power-email
Re: msvcp60 and bug 7679
On 4/26/07, Eric Pouech [EMAIL PROTECTED] wrote: use the pass command in winedbg instead of c (cont) (or rtfm) Haha, but I already tried that, which is why I said I'm hitting a wall.. I know pass is supposed to go around it, but when I do pass, it just prints the same thing it printed before. I even tried pass about 40 times in a row to make sure it wasnt just rerunning that function, and still no luck. I'll see what a +dbghelp trace gives me tho since Louis was able to get it working with native dbghelp. -- Thanks Tom Check out this new 3D Instant Messenger called IMVU. It's the best I have seen yet! http://imvu.com/catalog/web_invitation.php?userId=1547373from=power-email
Re: msvcp60 and bug 7679
On 4/26/07, Louis Lenders [EMAIL PROTECTED] wrote: I gave the application a try, and looks like there's a problem with builtin dbghelp.dll. When i use native dbghelp.dll , the login window comes up just fine, and i'm able to login as well, and get the client running. didn't test any further. BTW, could someone please add these categories like dbghelp, urlmon, quartz etc to bugzilla? has been requested a few times but no response yet :( Thanks for looking at this Louis. At least for now, I can put a note on appdb to use native. I made a +seh,+dbghelp log, but am bashing my head again because I dont see any problems with it... I'll do the rest of my commenting on this program on the bug -- Thanks Tom Check out this new 3D Instant Messenger called IMVU. It's the best I have seen yet! http://imvu.com/catalog/web_invitation.php?userId=1547373from=power-email
Re: Trace fix
On 4/26/07, Alexandre Julliard [EMAIL PROTECTED] wrote: Tom Spear [EMAIL PROTECTED] writes: Ok, so then if I resend them now, will you take a look? I believe I have fixed every problem with each one, as I havent gotten any more comments about something being borked. No, resend just one, making absolutely sure that you got everything right, and wait for it to be committed before sending the next one. You have to make sure you can walk before you try to run. Ok. Any particular order? I'll do a regedit one for now.. -- Thanks Tom Check out this new 3D Instant Messenger called IMVU. It's the best I have seen yet! http://imvu.com/catalog/web_invitation.php?userId=1547373from=power-email
Re: [Try 3] [programs/uninstaller] Check HKCU for uninstall entries
On 4/25/07, Peter Beutner [EMAIL PROTECTED] wrote: Tom Spear schrieb: Imo it's more common to use all uppercase names only for macros. And i think it is mostly done that way in wine. In fact you have to look quite hard to find an all uppercase variable name ;) I agree, which is why I had it with certain letters capitalized, originally. But are there any reasons why making that a variable at all? #define MAX_SUBKEY_LEN 255 No particular reason other than trying to keep the with the same format as the rest of the file. Undid the random whitespace change. Any others? it is still there in try4 - + RemoveSpecificProgram( argv[i++] ); there is one more -if(count != 0) +if (count != 0) Fixed. + WINE_TRACE(allocated entry #%d: %s (%s), %s\n, +numentries, wine_dbgstr_w(entries[numentries-1].key), wine_dbgstr_w(entries[numentries-1].descr), wine_dbgstr_w(entries[numentries-1].command)); You mixed tabs and spaces, and that line is too long. Who originally wrote this damn code and how did it get by you in the first place? If it's too long, please give more info on how I can make it shorter, while still getting the same output. Especially since I didnt write that line. As a suggestion: uninst_entry *entry; ... entry = entries[numentries - 1]; Now you only need to write entry-xxx instead of entries[numentries - 1].xxx in that whole block and get slightly shorter lines. That is very helpful, thanks! Hopefully here is the last try, coming up in my next email. -- Thanks Tom Check out this new 3D Instant Messenger called IMVU. It's the best I have seen yet! http://imvu.com/catalog/web_invitation.php?userId=1547373from=power-email
[Try 5] Check HKCU for uninstall entries
Hopefully this will be the last try. Instead of just checking HKLM for uninstall entries, check HKCU as well. Converted all instances of entries[numentries-1].xxx to entry-xxx Fixed a bug with a trace. Ran the file thru kwrite to make SURE there are NO tabs! Double checked the diff to make sure no whitespaces were removed by adding new lines. Any comments? This is sent to wine-patches, so if there are no problems with this patch, it is ready for commit. -- Thanks Tom Check out this new 3D Instant Messenger called IMVU. It's the best I have seen yet! http://imvu.com/catalog/web_invitation.php?userId=1547373from=power-email uninstaller.patch Description: Binary data
[programs/uninstaller (Try 6)] Scan HKCU
Scan HKCU for uninstall entries. Separated out the trace fix, moved the HeapAlloc for entries outside the for loop, realigned the 2nd line of code for a trace. This patch obsoletes all previous ones. -- Thanks Tom Check out this new 3D Instant Messenger called IMVU. It's the best I have seen yet! http://imvu.com/catalog/web_invitation.php?userId=1547373from=power-email
Uninstaller: 2 questions
Why do we show a dialog if there are no uninstall entries found in the registry? Windows does not do that, and I think we shouldn't either. However, on that same note, I think we should, since this uninstaller is not designed to mimic Windows' Add/Remove Programs, catch when a program's uninstaller does not remove it's uninstall entry from the registry, after the process exits, and go ahead and remove the entry. I already have a patch for the 2nd question ready to be committed once my patch to check HKCU is committed, however I am waiting for comment before I actually send it. The first question I will wait to write a patch for, however it is only like a 4 line patch.. -- Thanks Tom Check out this new 3D Instant Messenger called IMVU. It's the best I have seen yet! http://imvu.com/catalog/web_invitation.php?userId=1547373from=power-email
Re: Uninstaller: 2 questions
On 4/25/07, Steven Edwards [EMAIL PROTECTED] wrote: Hi Tom, I've watched your discussions for a while and have been meaning to comment but have been super busy. The uninstaller was one of the first things I looked at when getting in to wine development and it confused me to no end due to a bug that only showed up when running it under windows. I assume, then, that you have since fixed it? If not, I'd like to take a stab. The whole windows uninstall process was borked for a long time until the advent of msi. I think we should follow the behavior of newer versions of windows and if the process fails to remove the entry, the next time the user tries to uninstall the application, it should prompt to remove the offending entry. I know it seems dumb, if the uninstall really did work but was just being stupid in not removing the entry, why not remove it? I just worry we will get in to a situation where the uninstall partly worked and uninstall.exe goes ahead, removes the entry and then the user is left with old files floating around. Better to leave the entry and then the next time they try to run it, prompt them to just remove the offending entry. Ok, then that is the current behavior, so I will leave it alone. -- Thanks Tom Check out this new 3D Instant Messenger called IMVU. It's the best I have seen yet! http://imvu.com/catalog/web_invitation.php?userId=1547373from=power-email
Re: Uninstaller: 2 questions
On 4/25/07, Frank Richter [EMAIL PROTECTED] wrote: So a user starts the uninstall app but doesn't see a dialog... and probably thinks it's a bug. On the other hand, just showing a dialog with an empty list makes it clear that there's nothing to uninstall and will probably not produce false bug reports. Good point. I will leave it alone. I just worry we will get in to a situation where the uninstall partly worked and uninstall.exe goes ahead, removes the entry and then the user is left with old files floating around. Better to leave the entry and then the next time they try to run it, prompt them to just remove the offending entry. Well, if we remove the entry, we need to let the user know what is going on. Windows does it this way, and the reason is because (at least with some of the smarter users), it will make them realize that they should check to see if there are any stale files hanging around. -- Thanks Tom Check out this new 3D Instant Messenger called IMVU. It's the best I have seen yet! http://imvu.com/catalog/web_invitation.php?userId=1547373from=power-email
Re: Uninstaller: 2 questions
On 4/25/07, Robert Shearman [EMAIL PROTECTED] wrote: What happens if you cancel the uninstall? Won't the uninstaller remove the entry with the program still installed? That is true too. I was originally thinking along the lines of checking the exit status of the uninstaller, but whether it is cancelled, or completes successfully, or hell, crashes for that matter, it's not going to give you a usable (in this context) exit status. So I decided to scrap that idea. -- Thanks Tom Check out this new 3D Instant Messenger called IMVU. It's the best I have seen yet! http://imvu.com/catalog/web_invitation.php?userId=1547373from=power-email
Typo fix
Just curious. I sent a patch to wine-patches last week to fix a typo in a trace for regedit, but it was not committed. Should I go ahead and resend it? -- Thanks Tom Check out this new 3D Instant Messenger called IMVU. It's the best I have seen yet! http://imvu.com/catalog/web_invitation.php?userId=1547373from=power-email
Should we handle this?
While testing my uninstaller patch, I ran into the issue that was a problem in Win95, where the uninstaller did not handle deletion of it's uninstall registry key. What I did was install a program that I new put it's registry key under HKLM. Then I copied that key to HKCU. Then I installed a program that I knew put its key under HKCU and copied that key to HKLM. Then I uninstalled each one with the copied key. The program's uninstaller removed the key that it was instructed to take out, but left the other one behind. My question is this. If an uninstaller (outside of my test environment) does not remove it's registry key before the process exits, should our uninstaller check for that after the process exits, and go ahead and remove the key, or should we make the user press the uninstall button again and ask them if they want to remove the key (like we already do, in the event they manually removed the program, by deleting the folder)? -- Thanks Tom Check out this new 3D Instant Messenger called IMVU. It's the best I have seen yet! http://imvu.com/catalog/web_invitation.php?userId=1547373from=power-email