jaclaz

I guess you will need something newer if your intention is to confuse (maybe you meant surprise?) me. That Sprites Mod hack is what ? 2013? , and anyway is very little confusing, as a matter of fact (like BTW most of the things written by Jeroen Domburg) it is extremely simple and can be followed relatively easily. jaclaz

Ok. Attached ETD_XP2.inf (70 KB, slightly smaller than the original 1940 or so ) I set it to: DriverDate = "03/05/2013" DriverVersion = "11.0.0.31" to reflect the actual data of the used ETD.sys. ETD_XP2.inf , renamed to ETD.inf and copied to the expanded \tp118w7\{app}\Elantech\11.4.14.1\X86\ thus replacing the existing ETD.inf, should install fine the "basic" driver and functions. Try it after having cleaned/checked the Registry (just in case of leftovers from previous tests). See if it works. I concur that the only "pretty" control panel that makes sense is the ETDSimpleUI.exe, I will make a separate .inf (that of course will need the ETDSimpleUI.exe file from the Samsung driver, as - seemingly - all the .avi's used by ETDSimpleUI.exe are already in the 11.14.4.1). jaclaz ETD_XP_2.zip

Good , which is actually called "track". And right when you think you got the concepts right let me confuse you with geometry and the way it is "shielded" by the disk hardware: https://blog.stuffedcow.net/2019/09/hard-disk-geometry-microbenchmarking/ jaclaz

WHICH circle? Almost a cluster is a bunch of sectors addressed as the smallest unit of storage in a file system. What the file system driver does may depend on a number of factors, do not assume too much. Not really a block ( or sector) is the minimal addressable unit on device, it is not arbitrary at all, it is hardcoded in either the device or in its driver, for all it matters its size is normally 512 bytes but can be also 4096 bytes (newer disks). Again, do not assume too much, you can write a single byte to disk just fine. This is irrelevant (the two together). The stripe size is always a multiple of clusters, whilst stripe size may (marginally) matter for performance in a RAID 0, 2, 3, 4. 5, etc. i.e. in "striped" sets, (and it has no influence on Raid 1), it is a separate parameter from cluster size, which should be chosen as the "best possible" for the data to be managed independently from stripe size. About stripe size: https://www.tomshardware.com/reviews/RAID-SCALING-CHARTS,1735-4.html The rule of the thumb is however that the larger the stripe size, the faster the data transfer is, but this may also depend on the specific (software or hardware) setup. About cluster size, more or less the same rule of the thumb applies, the larger the cluster size, the faster the data transfer will be BUT in the case of smaller files there will be more "slack" space (wasted space), so you need a compromise. Since you are limited to FAT32 AND you have volume >32 GB you are "forced" to have cluster size 32K (and no, don't think that making two volumes with cluster size 16KB will change anything in a noticeable amount, it will be only very slightly slower). Then for stripe size, use the largest you are allowed to. jaclaz

Yep , that could be a good alternative, and - like the one above - could be an option, for the people that are not satisfied by the "standard"with ONLY files from 11.4.14.1. Summing it up: 1) new "base" ETD.inf that installs (hopefully) a working driver from 11.4.14.1 ONLY 2) optional .inf's that adds (hopefully) any among these (renamed to OEMElan.exe): Asus_UI.exe -> Asus_UI.inf ETDAniConf.exe -> ETDAniConf.inf Lenovo.exe -> Lenovo.inf DellTpad.exe -> DellTpad.exe ETDSimpleUI.exe ->ETDSimpleUI.exe ToshibaTp.exe ->ToshibaTp.inf and makes the added Control Panel item to launch it (basically it is several copies of the same inf with just the name of the .exe changed AND the relative .avi's [1]) 3) optional ETDUI.inf that adds (hopefully) ETDUI.cpl and and makes the added Control Panel item to launch it Back to work, what about: jaclaz [1] this (finding which .avi "belongs" to which .exe) might be the most work, I managed (somehow) to have the ETDSimpleUI.exe run on my machine, and I already found that in the ETD_XP.inf three needed .avi's are missing (Edge_Scroll.avi, Edge_inertial.avi and Scroll_No_FreeScroll.avi), so (hopefully) the "final" list for ETDSimpleUI.exe is: Button_ClickPad.avi DnD01.avi DnD02.avi DnD03.avi Edge_inertial.avi Edge_Scroll.avi Rotation_Twist.avi Scroll_Inertial.avi Scroll_No_FreeScroll.avi Swipe_3F_Down_SwitchWnd.avi Swipe_3F_LR_PageupDown.avi Swipe_3F_Up_Computer.avi Tapping_1F.avi Tapping_2F_Middle_Button.avi Tapping_2F_Popup.avi Zooming.avi The list for Asus_UI.exe should be the one in the ETD.inf from Z11566 (since that is the 10.0.26 version that "works best" for you (no files in common with the above): 1F_Activate.avi 1F_Drag.avi 1F_Tap.avi 2F_Popup.avi 2F_Rotation.avi 2F_Scroll.avi 2F_Zoom.avi 3F_ReturnDesktop.avi 3F_SwitchPage.avi 3F_SwitchWnd.avi 1F_appbar_top.avi 1F_appbar_down.avi 1F_charmbar.avi 1F_changepage.avi 1F_applist.avi 1F_snapview.avi 3F_Changepage_tab.avi 2F_Zoom_all.avi The list for Lenovo.exe should be the one in the ETD.inf from tp118w7, but there are overlaps.

NO ideas. But I think we have determined that ETDAniconf.exe sucks (and sucks big) and that what we really want is the "built-in" "complex" control panel in ETDmcpl.dll, and all in all I find natural enough that clicking on the tray Icon brings up the Mouse Control Panel and from its "Elan" tab, pressing the Options brings up the settings. It is two or three clicks more, but it isn't that bad IMHO. The {3EA22083-D5BF-411b-A6D6-A27541794266} Control Panel namespace could be managed by a separate .inf or ,reg that might be useful to those that want to use either the Asus_UI.exe (BTW have you noticed how in the filemon log *every* .exe is looked for BUT the Asus_UI.exe) or the ETDSimpleUI.exe (or *whatever*). From what I understand, one could (optionally) add to the Elantech directory *any* of: Asus_UI.exe ETDAniConf.exe Lenovo.exe DellTpad.exe ETDSimpleUI.exe ToshibaTp.exe renamed to - say - ElanCPL.exe and import the .inf or .reg (pointing to the new name) to have one of them as an alternative (via the Cotrol Panel icon), not as a replacement, of the ETDmcpl.dll "complex" settings (best of both worlds). jaclaz

I don't know. The original .inf settings are for a .exe (Asus_UI.exe or ETDSimple_UI.exe). Try tracing with Filemon what happens in 3) : 1) open the mouse control panel 2) choose the "Elan" tab 3) click on the "Options" button Maybe a parameter is passed to the .dll? jaclaz

Good. We can then remove, besides the unused "control panel" files also the stupid .avi files. Yep, almost. But WHICH Registry entry? ONLY the UI_ThemeOption? I.e. do you have now: CustomerDefinition=5 UI_ThemeOption=0 If yes, try setting to 0 also the CustomerDefinition (only to see if it actually changes something). As a side note I suspect that a reboot is not needed, but stopping (Task Manager) and re-starting ETDCtrl.exe should do the same. And - in order to make a (hopefully) final ETD_XP.inf we need to take a standing in favour (or against) attmpting removing: 1) anything that is not "ETD_STD_Inst" <- this should make it (hopefully) functional (at least installable) for ETD 000,101-102,105-109,10A-114 2) anything related to pointstick 3) the [ETD_APOptimize.AddReg] section (whole (3a) or leaving some "key" entries, like Normal Desktop and Windows Explorer (3b)) 4) the [ETD_GestureAPHotKey.AddReg] (whole (4a) or leaving some "key" entries, like Desktop and Windows Explorer (4b) ) Personally I propend for 1/2/3b/4b. jaclaz

Sure it doesn't. Actually the whole HKLM\CurrentControSet hive does not exist outside the running system CurrentControlSet is a "volatile" path created at boot by the OS and it is a copy of either ControlSet001 or ControlSet002 (typically, but I have seen systems with ControlSet up to 025). To know which "real" hive is in use (or will be in use at next boot, check the HKEY_LOCAL_MACHINE\SYSTEM\Select key and its contents). Normally you only have two ControlSet's, 001 and 002 and you want to have EnableBigLba=1 on both. jaclaz

Yep. Speaking of names, I know a man with a wooden leg named Smith ... jaclaz

Well, you make it sound like it was from before the war . ETDSimpleUI.exe dated 08/10/2012 version 10.0.0.31 in an archive with ETD.sys version 10.0.0.197 Asus_UI.exe dated 03/01/2013 version 10.0.0.26 in an archive with ETD.sys version 11.0.0.6 (and both used with an actual ETD.sys dated 05/03/2013 version 11.0.0.31 Everything is within a 6 months period. jaclaz

Probably in Cheshire: https://en.wikipedia.org/wiki/Ashley,_Cheshire and here we have a rare image of the Cheshire cat: jaclaz

UNexpected or not, that "full" control panel is EITHER: 1) exactly the same as the one that comes up with UI_ThemeOption=0 (which should mean that ETDUI.cpl has been "phased out" becsause its functionalities have been included in ETDmcpl.dll OR: 2) an even slighlty different one Yep, and this was actually "expected". The point is whether changing the one, or the other or both the: CustomerDefinition=5 UI_ThemeOption=5 to either: CustomerDefinition=0 UI_ThemeOption=0 or: CustomerDefinition=1 UI_ThemeOption=1 or possibly: CustomerDefinition=4 UI_ThemeOption=4 changes the way the settings are read by the driver. jaclaz

I don't know. It is possible that (like seemingly half the files in the install) there are several layers of "cruft" and that simply the driver doesn't "talk" with this (or that) control panel, or maybe the issue is with the "missing" ETDCtrlHelper.dll and/or ETDService.exe. Setting "0" : the ETDmcpl.dll is definitely BOTH the added tab in "Mouse" AND a Control panel of sorts, possibly what you call the "complex" control panel. Setting "1": and definitely the ETDaniconf.exe is another control panel, possibly the "simple" one that is "inserted" before the above (if the file is present and setting is 0) Setting "2": Dell Setting "3": Dual mode Setting "4": Needs some other file, very likely ETDUISimple.exe (the one we found only on the Samsung driver. Setting "5": Asus_UI.exe or similar .NET based control panel Then there remains the (possibly the oldest) ETDUI.cpl <- can you try this? that may be nto working at all. There is another key that RainyShadow pointed out earlier: https://msfn.org/board/topic/181362-xp-driver-for-amd-hd8180-graphics-and-elan-ps2-touchpad/page/17/?tab=comments#comment-1180835 maybe that one must be changed also - in "sync" with UI_Theme one. In the .inf of "Asus" drivers we have: CustomerDefinition=5 UI_ThemeOption=5 in the .inf of "Samsung" drivers we have: CustomerDefinition=4 UI_ThemeOption=4 in the .inf of "Lenovo" driver we have: CustomerDefinition=0 (and also have a different "definition" of the values) UI_ThemeOption= various values, 1, 6, 3, 8 in the .inf of "Toshiba" drivers we have: CustomerDefinition=7 UI_ThemeOption=7 in the .inf of "Dell" drivers we have: CustomerDefinition=2 UI_ThemeOption=2 in the .inf of "Medion" and "Gigabyte" those keys are nowhere to be found, I am assuming that to be equivalent to have both of them 0. You could try using Regmon to see which keys the: ETDmcpl.dll ETDAniconf.exe Asus_UI.exe ETDUISimple.exe try to access when you change a single setting through them. I suspect that the Asus_UI.exe (and the ETDSimpleUI.exe) use different paths from the other ones, and there must be a setting (possibly the CustomerDefinition key) that changes this behaviour. If we take the "Medion" and "Gigabyte" drivers (which are the "simpler" ones and have no "fancy" OEM branded settings tools included in the .inf) they must *somehow* work on the supported hardware, and they miss both these keys, maybe setting them both to 0 makes the ETDmcpl and/or ETDAniconf keep the settings. jaclaz

But, once you have it, what happens when you click on it? I mean, does it "produce" something if not better, different from Asus_UI.exe or it does nothing useful? From what I understand (but I may well be wrong, and I am now confused by the reports) the famous key settings are: 0 = Mcpl UI <- uses *somehow* ETDmcpl.dll 1 = Animation UI <- uses *somehow* ETDaniconf.exe 2 = Dell UI <- uses *somehow* DellTpad.exe (or something like that) 3 = Dual Mode <- ?? but from what you report it is something that uses *something* for the trackpoint and ETDAniconf.exe for the smartpad 4 = Simple UI <- uses *somehow* ETDSimpleUI.exe 5 = ASUS UI <- uses *somehow* Asus_UI.exe Excluding setting 2 (DELL) and setting 5 (Asus) which are fine (the first doesn't work - probably beacuse it wants the DellTpad.exe), the second does work, though without animations, what happens with the other ones? In theory: 0 should work as long as ETDmcpl.dll is in the folder 1 should work as long as ETDAniconf.exe (and possibly a number of stupid .avi files) is in the folder 3 Whatever, we don't want/need it 4 should work as long as ETDSimpleUI.exe is in the folder (but to get the ETDSimpleUI.exe file one would need to additionally get the Samsung driver) Can you re-check? Back to *needed* files: ASUS_UI.exe 10.0.0.26 -> from Z11566 ETDCtrl.exe 11.0.0.13 -> from tp118w7 ETDUn_inst.exe 10.0.0.9 -> from tp118w7 ETDApix.dll 11.0.0.8 -> from tp118w7 ETDCmds.dll 11.0.0.8 -> from tp118w7 ETDFavorite.dll 11.0.0.7 -> from tp118w7 ETDMcpl.dll 11.0.0.10 -> from tp118w7 <- actually optional if Asus_UI.exe is used ETDUninst.dll 10.0.0.1 -> from tp118w7 Optional files: Aniconf.exe 11.0.0.7 -> from tp118w7 ETDSimpleUI.exe 10.0.0.31 -> from Samsung 4.9.0.0 Correct? Then, to install (i.e. in the source files) there are: Setup.exe 11.0.0.4 -> from tp118w7 Setup.ini as follows: (should work also with CheckHWID=1 unREMmed) ETD.sys 11.0.0.31 -> from tp118w7 ETD,inf <- the actual edited ETD_XP.inf renamed to ETD.inf Correct? So, all files either come from tp118w7 or are home-made, exception made for the ASUS_UI.exe. In order to replicate your setup, one would need to: 1) download the tp118w7.exe from Lenovo: https://pcsupport.lenovo.com/hn/en/products/laptops-and-netbooks/lenovo-g-series-laptops/lenovo-g505s-notebook/downloads/ds035770 and expand it with Universal Extractor 2) download the Z11566 from either: http://dlcdnet.asus.com/pub/ASUS/nb/DriversForWin8/Touchpad/Touchpad_Elantech_Win7_8_32_VER11566.zip http://ftp.tekwind.co.jp/pub/asustw/nb/DriversForWin8/Touchpad/Touchpad_Elantech_Win7_8_32_VER11566.zip and get from it just the Asus_UI.exe. So I am assuming that the "\Elantech\UIFunctionOption,UI_Theme" should be set in the .inf as 5. Now the big question (same as the one at the beginning of the post, re-worded), if anyone only gets the tp118w7.exe (and doesn't want the Asus_UI.exe) is there any other setting 0,1,2,3 for the "\Elantech\UIFunctionOption,UI_Theme" key that produce a (even if simplified) control settings change tool (that saves and keeps settings)? And the other big question, what does the ETDUI.cpl (coming from either the Samsung driver or from the Z11542 or from some other files) does? This file is a mistery as it is in the Samsung driver but there is no reference to it in the .inf. (in the Z11542 and in earlier drivers there is a reference to it but it is remmed out). jaclaz

No, thanks. jaclaz

Surely someone does, now if you could actually ask a proper question, explaining what the issue is, then those people may be put in the condition to understand what the problem is. All I can see that may be vaguely relevant is this oneliner on RyanVM: jaclaz

I am confused. Anyway, maybe, without the ToshibaTp.exe, the current set of files (but with the ETD.sys from 11.4.14.1) work "better" with the key set to 0,1,2,3,4? I guess that you are ready for the final verdict? (so that we can have a "final" configuration for the .inf) ETDSimpleUI.exe or Asus_UI.exe? And with which setting of the \Elantech\UIFunctionOption",UI_Theme, key? Which final version of each file do you want to use? Which exact files do you want to use? (or which files are not needed?) Do you need/want the added Control Panel item as in: https://msfn.org/board/topic/181362-xp-driver-for-amd-hd8180-graphics-and-elan-ps2-touchpad/?do=findComment&comment=1182136 What do you think, can we remove the whole "[ETD_APOptimize.AddReg]" from the .inf? jaclaz

Well, all you will obtain will be an updated definition , something like "some selfish teenage indian dude who watches anime". The "teenage" becoming (already or soon) inaccurate, time passes faster than one would expect ... Haters are gonna hate anyway. jaclaz

The PackardBell one doesn't seem anyway to offer *any* advantage, whilst the Toshiba one may. Using the Toshiba 11.8.16.2 files, restore only the (working) 11.4.14.1 ETD.sys and try only running the Toshiba settings tool, only to see if it is actually more responsive than the Asus one. If the newer files still don't work restore also the other files from 11.4.14.1 and try again. And of course you need to change: HKLM,"%ServiceRoot%\Elantech\UIFunctionOption",UI_Theme,%REG_DWORD%,7 (it has to be seen if the option 7 is connected to the ETD.SYS or - more likely - to the ETDCtrl.exe version) BTW this option 7 should imply that also an option 6 exists ... jaclaz

Using two separate partitions/volumes, 1 for Vista and 1 for 8.1 would be needed. The Vista install (unlike 7 and later) does not create the (second - third in this case) separate "boot" volume (What MS calls "System") on unpartitioned devices which actually, in this case might be useful . http://www.multibooters.co.uk/system.html The possible issue (and a decision to be made) is whether: 1) having Vista on C: and Windows 8.1 on D: (and have each system "see" the other volume) and with the BOOTMGR and \boot\BCD on C: or 2) having Vista on C: and Windows 8.1 on C: hiding the "other" partition to the "other" OS and with separate BOOTMGR and \boot\BCD on each volume, using a third party boot manager such as grubdos to hide the other volume or 3) having either of the above and a third volume with a "common" BOOTMGR and \boot\BCD, with or without a drive letter normally assigned to it. jaclaz

Yep, right now the "ETD_XP.inf" works just fine, it is only a matter of choosing which files (and which versions of files) to add in the source directory, how to set the two or three Registry Keys we have seen as being the "mode selection" and which ones to REM (or to UNREM) in the .inf according to what the "final UI" is desired. And yes, you can use the "ETD_XP.inf" with the files from those two latest drivers and see if they are in any way "better" (maybe - and if actually works/runs - the "Toshiba" UI is better than the Asus_UI or than the ETDSimpleUI). jaclaz

Maybe that is a more general issue with XP and .avi files (and the mechanism that the Asus_UI.exe uses to show them, or the relatively poor support of .Net in XP)? Anyway, now you have an almost complete panoramic view of the various options, and - right when you think it's over - what if I told you about later driver versions seemingly XP compatible (5.1)? 11.8.16.2 actual ETD.sys version 11.98.8.3 dated 26/03/2013 And that has in its .inf: https://drivers.softpedia.com/get/KEYBOARD-and-MOUSE/Elantech/Toshiba-Elantech-TouchPad-Driver-118162.shtml and 11.6.28.201 actual ETD.sys version 11.116.0.0 dated 02/10/2013 And that has in its .inf: http://global-download.packardbell.com/GDFiles/Driver/TouchPad/TouchPad_ELANTECH_11.6.28.201_W81x64_A.zip?acerid=635170560607154373 I remain curious about the set of dices the good guys at Elantech use for assigning versions to their drivers ... jaclaz

Yep, maybe something else has been changed in all the tests. jaclaz

I don't know, try with 4. jaclaz