X11 expertise on ancient HW sought... (4-plane visual (overlay) via X-server on MS-WIndows)
[also posted to comp.graphics.x today] Hello Folks, I'm trying to get an application that currently uses a local display on an ancient DEC Alpha workstation with a (for the time) mid-to-high-end graphics controller (ZLX-E2) to instead use an X-server running under MS-Windows. The application is complaining that it cannot find a "4/5-bit visual". It almost certainly wants to use this visual for an overlay, as the application displays moving objects superimposed on a map. On the original hardware, xdpyinfo tells me: > [...] > supported pixmap formats: > [...] > depth 4, bits_per_pixel 8, scanline_pad 32 > [...] > screen #0: >[...] > depths (4):8, 12, 24, 4 > [...] > number of visuals:21 > [lots of other visuals here, but no 4-plane except for the following] > visual: > visual id:0x36 > class:PseudoColor > depth:4 planes > available colormap entries:16 > red, green, blue masks:0x0, 0x0, 0x0 > significant bits in color specification:4 bits and "xprop -root" tells me: > [...] > SERVER_OVERLAY_VISUALS(SERVER_OVERLAY_VISUALS) = 0x36, 0x1, 0x0, 0x1 As you can see, there seems to be exactly one overlay, whose visual id (0x36) corresponds to the single 4-plane visual listed by xdpyinfo. When I use the above commands to retrieve the capabilities of the MS-Windows X-server (Exceed, in this case), xdpyinfo does not list a 4-plane visual at all. "xprop" lists lots of overlays, 24 in total, all of them 8-plane visuals. The MS-Windows box is running Windows-7 (64bit) and has a Nvidia Quadro 400 GPU. I used the "Nvidia Control Panel" to set "Enable overlay" to "on" in the "Manage 3D settings" section. Also, in the Exceed X-server configuration I enabled "OpenGL", and within that enabled "Overlay Support" and "GLX 1.3 Support". I conclude that the MS-Windows SW/HW system (X-server, MS-Win GPU driver, GPU) cannot offer 4-plane visuals. However, I don't know what system component(s) is/are the cause the problem. I have tested VcXsrv, Reflection-X, Exceed (with 3D option), X-Win32 and even the ancient DEC Pathworks X-server eXcursion with no success. I'm working on getting an evaluation copy of PTC's MKSTools X/Server. Of the X-servers I've tested, Exceed seems to offer the most configuration parameters. I'm not even sure the Quadro 400 can handle 4bpp "visuals", or whatever MS-Windows calls them. In fact, I wonder if any modern hardware offers 4bpp capability. On my Linux box with a GeForce GT 430 I don't have any 4-plane visuals, and xprop doesn't mention any overlays either. I'm somewhat confused about where overlays fit into the X scheme. I have seen lots of references to overlays in an OpenGL context, however the Alpha seems not to have any OpenGL capability: GLX is not in the list of extentions printed by xdpyinfo. Can someone clear this up for me? Am I correct to assume that the GPU must support 4bpp in order for it even to be possible for the X-server to propagate a 4-plane visual to a client? If yes, how can I determine if a GPU supports 4bpp? Nvidia is very sparing with the information in their specs for the Quadro 400 GPU. Assuming I can find a GPU that supports/offers 4bpp, does anyone know an X-server product/project that can provide 4-plane overlays? thanks, Rob
Re: X11 expertise on ancient HW sought... (4-plane visual (overlay) via X-server on MS-WIndows)
On Thu, 4 Jun 2015, IMAP List Administration wrote: > On the original hardware, xdpyinfo tells me: > > [...] > > supported pixmap formats: > > [...] > > depth 4, bits_per_pixel 8, scanline_pad 32 > > [...] > > screen #0: > >[...] > > depths (4):8, 12, 24, 4 I can muddy the water for you. On my VMS 8.3 box (a not new Alphastation 500) I get this: richardlo $ xdpyinfo name of display:_WSA2: version number:11.0 vendor string:DECWINDOWS Hewlett-Packard Development Company OpenVMS vendor release number:8003 maximum request size: 65535 longwords (262140 bytes) motion buffer size: 0 bitmap unit, bit order, padding:32, LSBFirst, 32 image byte order:LSBFirst number of supported pixmap formats:2 supported pixmap formats: depth 1, bits_per_pixel 1, scanline_pad 32 depth 8, bits_per_pixel 8, scanline_pad 32 So I don't have a pixmap format with a depth of 4 either. But my Redhat Linux EL6.6 box does, it reports support for 1, 4, 8, 15, 16, 24, and 32. -- Richard Loken VE6BSV, Unix System Administrator : "Anybody can be a father Athabasca University: but you have to earn Athabasca, Alberta Canada : the title of 'daddy'" ** richar...@admin.athabascau.ca ** : - Lynn Johnston
Re: X11 expertise on ancient HW sought... (4-plane visual (overlay) via X-server on MS-WIndows)
Richard Loken wrote: > > On Thu, 4 Jun 2015, IMAP List Administration wrote: > > > On the original hardware, xdpyinfo tells me: > > > [...] > > > supported pixmap formats: > > > [...] > > > depth 4, bits_per_pixel 8, scanline_pad 32 > > > [...] > > > screen #0: > > >[...] > > > depths (4):8, 12, 24, 4 > > I can muddy the water for you. On my VMS 8.3 box (a not new Alphastation > 500) I get this: > > richardlo $ xdpyinfo > name of display:_WSA2: > version number:11.0 > vendor string:DECWINDOWS Hewlett-Packard Development Company OpenVMS > vendor release number:8003 > maximum request size: 65535 longwords (262140 bytes) > motion buffer size: 0 > bitmap unit, bit order, padding:32, LSBFirst, 32 > image byte order:LSBFirst > number of supported pixmap formats:2 > supported pixmap formats: > depth 1, bits_per_pixel 1, scanline_pad 32 > depth 8, bits_per_pixel 8, scanline_pad 32 > > So I don't have a pixmap format with a depth of 4 either. But my Redhat > Linux EL6.6 box does, it reports support for 1, 4, 8, 15, 16, 24, and 32. > If further muddying is required, my about 20 year old Alphaserver 800 running VMS 8.2 with a Powerstorm something (maybe 3D30 or 4D20?) graphics card and my PWS 500a(u) running VMS 8.3 and some slightly lesser capable graphics card that I cannot recall the name of at all both report this: visual: visual id:0x36 class:PseudoColor depth:4 planes size of colormap:16 entries red, green, blue masks:0x0, 0x0, 0x0 significant bits in color specification:4 bits and this: SERVER_OVERLAY_VISUALS(SERVER_OVERLAY_VISUALS) = 0x36, 0x1, 0x0, 0x1 Regards, Peter Coghlan.
Re: X11 expertise on ancient HW sought... (4-plane visual (overlay) via X-server on MS-WIndows)
On 4 June 2015 at 19:24, IMAP List Administration wrote: > [also posted to comp.graphics.x today] > > Hello Folks, > > I'm trying to get an application that currently uses a local display on an > ancient DEC Alpha workstation with a (for the time) mid-to-high-end graphics > controller (ZLX-E2) to instead use an X-server running under MS-Windows. > > The application is complaining that it cannot find a "4/5-bit visual". It > almost > certainly wants to use this visual for an overlay, as the application displays > moving objects superimposed on a map. > > On the original hardware, xdpyinfo tells me: >> [...] >> supported pixmap formats: >> [...] >> depth 4, bits_per_pixel 8, scanline_pad 32 >> [...] >> screen #0: >>[...] >> depths (4):8, 12, 24, 4 >> [...] >> number of visuals:21 >> [lots of other visuals here, but no 4-plane except for the following] >> visual: >> visual id:0x36 >> class:PseudoColor >> depth:4 planes >> available colormap entries:16 >> red, green, blue masks:0x0, 0x0, 0x0 >> significant bits in color specification:4 bits > > and "xprop -root" tells me: >> [...] >> SERVER_OVERLAY_VISUALS(SERVER_OVERLAY_VISUALS) = 0x36, 0x1, 0x0, 0x1 > > As you can see, there seems to be exactly one overlay, whose visual id (0x36) > corresponds to the single 4-plane visual listed by xdpyinfo. > > When I use the above commands to retrieve the capabilities of the MS-Windows > X-server (Exceed, in this case), xdpyinfo does not list a 4-plane visual at > all. > "xprop" lists lots of overlays, 24 in total, all of them 8-plane visuals. > > The MS-Windows box is running Windows-7 (64bit) and has a Nvidia Quadro 400 > GPU. > I used the "Nvidia Control Panel" to set "Enable overlay" to "on" in the > "Manage > 3D settings" section. Also, in the Exceed X-server configuration I enabled > "OpenGL", and within that enabled "Overlay Support" and "GLX 1.3 Support". > > I conclude that the MS-Windows SW/HW system (X-server, MS-Win GPU driver, GPU) > cannot offer 4-plane visuals. However, I don't know what system component(s) > is/are the cause the problem. > > I have tested VcXsrv, Reflection-X, Exceed (with 3D option), X-Win32 and even > the ancient DEC Pathworks X-server eXcursion with no success. I'm working on > getting an evaluation copy of PTC's MKSTools X/Server. Of the X-servers I've > tested, Exceed seems to offer the most configuration parameters. > > I'm not even sure the Quadro 400 can handle 4bpp "visuals", or whatever > MS-Windows calls them. In fact, I wonder if any modern hardware offers 4bpp > capability. On my Linux box with a GeForce GT 430 I don't have any 4-plane > visuals, and xprop doesn't mention any overlays either. > > I'm somewhat confused about where overlays fit into the X scheme. I have seen > lots of references to overlays in an OpenGL context, however the Alpha seems > not > to have any OpenGL capability: GLX is not in the list of extentions printed by > xdpyinfo. Can someone clear this up for me? > > Am I correct to assume that the GPU must support 4bpp in order for it even to > be > possible for the X-server to propagate a 4-plane visual to a client? If yes, > how > can I determine if a GPU supports 4bpp? Nvidia is very sparing with the > information in their specs for the Quadro 400 GPU. > > Assuming I can find a GPU that supports/offers 4bpp, does anyone know an > X-server product/project that can provide 4-plane overlays? Have you tried MobaXterm? On my Thinkpad T420s (Intel gfx) booting into Windows 7, then running my NetBSD install in Virtual box and firing up an xterm using Moba as the X server reports: depths (7): 24, 1, 4, 8, 15, 16, 32 The Personal version of MobaXterm is free :)
Re: X11 expertise on ancient HW sought... (4-plane visual (overlay) via X-server on MS-WIndows)
On 06/05/2015 12:19 AM, David Brownlee wrote: > Have you tried MobaXterm? > > On my Thinkpad T420s (Intel gfx) booting into Windows 7, then running > my NetBSD install in Virtual box and firing up an xterm using Moba as > the X server reports: > > depths (7): 24, 1, 4, 8, 15, 16, 32 > > The Personal version of MobaXterm is free :) I have learned from trying the various MS-Win X-servers that that list of depths is not particularly meaningful. What does "xprop -root" say about overlays? Do you see a 4-plane visual, and also a 4-plane overlay whose ID matches the 4-plane visual? Rob
Re: X11 expertise on ancient HW sought... (4-plane visual (overlay) via X-server on MS-WIndows)
On 5 June 2015 at 16:15, IMAP List Administration wrote: > > On 06/05/2015 12:19 AM, David Brownlee wrote: >> Have you tried MobaXterm? >> >> On my Thinkpad T420s (Intel gfx) booting into Windows 7, then running >> my NetBSD install in Virtual box and firing up an xterm using Moba as >> the X server reports: >> >> depths (7): 24, 1, 4, 8, 15, 16, 32 >> >> The Personal version of MobaXterm is free :) > I have learned from trying the various MS-Win X-servers that that list of > depths > is not particularly meaningful. > > What does "xprop -root" say about overlays? Do you see a 4-plane visual, and > also a 4-plane overlay whose ID matches the 4-plane visual? Not very much: _NET_DESKTOP_NAMES(UTF8_STRING) = "Desktop" _NET_NUMBER_OF_DESKTOPS(CARDINAL) = 1 _NET_CURRENT_DESKTOP(CARDINAL) = 0 WM_ICON_SIZE(WM_ICON_SIZE): minimum icon size: 16 by 16 maximum icon size: 48 by 48 incremental size change: 16 by 16 _XKB_RULES_NAMES(STRING) = "base", "pc105", "gb", "", "" Full xdpyinfo below in case it helps... name of display:localhost:10.0 version number:11.0 vendor string:Moba/X vendor release number:11603000 maximum request size: 16777212 bytes motion buffer size: 256 bitmap unit, bit order, padding:32, LSBFirst, 32 image byte order:LSBFirst number of supported pixmap formats:7 supported pixmap formats: depth 1, bits_per_pixel 1, scanline_pad 32 depth 4, bits_per_pixel 8, scanline_pad 32 depth 8, bits_per_pixel 8, scanline_pad 32 depth 15, bits_per_pixel 16, scanline_pad 32 depth 16, bits_per_pixel 16, scanline_pad 32 depth 24, bits_per_pixel 32, scanline_pad 32 depth 32, bits_per_pixel 32, scanline_pad 32 keycode range:minimum 8, maximum 255 focus: window 0x800075, revert to Parent number of extensions:22 BIG-REQUESTS Composite DAMAGE DOUBLE-BUFFER GLX Generic Event Extension Present RANDR RECORD RENDER SGI-GLX SHAPE SYNC Windows-DRI X-Resource XC-MISC XFIXES XFree86-Bigfont XINERAMA XInputExtension XKEYBOARD XTEST default screen number:0 number of screens:1 screen #0: dimensions:1600x900 pixels (423x238 millimeters) resolution:96x96 dots per inch depths (7):24, 1, 4, 8, 15, 16, 32 root window id:0x101 depth of root window:24 planes number of colormaps:minimum 1, maximum 1 default colormap:0x20 default number of colormap cells:256 preallocated pixels:black 0, white 16777215 options:backing-store WHEN MAPPED, save-unders NO largest cursor:32x32 current input event mask:0x4a0004 ButtonPressMask StructureNotifyMask SubstructureNotifyMask PropertyChangeMask number of visuals:64 default visual id: 0x21 visual: visual id:0x21 class:TrueColor depth:24 planes available colormap entries:256 per subfield red, green, blue masks:0xff, 0xff00, 0xff significant bits in color specification:8 bits visual: visual id:0xc2 class:TrueColor depth:24 planes available colormap entries:256 per subfield red, green, blue masks:0xff, 0xff00, 0xff significant bits in color specification:8 bits visual: visual id:0xc3 class:TrueColor depth:24 planes available colormap entries:256 per subfield red, green, blue masks:0xff, 0xff00, 0xff significant bits in color specification:8 bits visual: visual id:0xc4 class:TrueColor depth:24 planes available colormap entries:256 per subfield red, green, blue masks:0xff, 0xff00, 0xff significant bits in color specification:8 bits visual: visual id:0xc5 class:TrueColor depth:24 planes available colormap entries:256 per subfield red, green, blue masks:0xff, 0xff00, 0xff significant bits in color specification:8 bits visual: visual id:0xc6 class:TrueColor depth:24 planes available colormap entries:256 per subfield red, green, blue masks:0xff, 0xff00, 0xff significant bits in color specification:8 bits visual: visual id:0xc7 class:TrueColor depth:24 planes available colormap entries:256 per subfield red, green, blue masks:0xff, 0xff00, 0xff significant bits in color specification:8 bits visual: visual id:0xc8 class:TrueColor depth:24 planes available colormap entries:256 per subfield red, green, blue masks:0xff, 0xff00, 0xff significant bits in color specification:8 bits visual: visual id:0xc9 class:TrueColor depth:24 planes available colormap entries:256 per subfield red, green, blue masks:0xff, 0xff00, 0xff significant bits in color specification:8 bits visual:
Re: X11 expertise on ancient HW sought... (4-plane visual (overlay) via X-server on MS-WIndows)
> I'm trying to get an application that currently uses a local display > on an ancient DEC Alpha workstation with a (for the time) > mid-to-high-end graphics controller (ZLX-E2) to instead use an > X-server running under MS-Windows. > The application is complaining that it cannot find a "4/5-bit > visual". [...] > I conclude that the MS-Windows SW/HW system (X-server, MS-Win GPU > driver, GPU) cannot offer 4-plane visuals. Only for values of "cannot" that are more like "do not bother to include the software to", I would say. > I'm not even sure the Quadro 400 can handle 4bpp "visuals", or > whatever MS-Windows calls them. "Visual" is a technical term in X. I don't know whether Windows has a corresponding abstraction, so I don't know whether there is anything it calls them. > Am I correct to assume that the GPU must support 4bpp in order for it > even to be possible for the X-server to propagate a 4-plane visual to > a client? No. However, it is substantially more difficult for an X server to present a visual that the hardware doesn't support, which is probably why the server you have doesn't do it. Well-behaved clients must be prepared to handle whatever capabilities the server presents; the problem here is that the client you have handles the server you have by complaining and dying. Depending on what other capabilities the client is using, you might be able to get the overlay effect using colourmap hackery with comparatively small code changes. What PseudoColor or DirectColor visuals are available? > If yes, how can I determine if a GPU supports 4bpp? Read the documentation, of course, or contact the manufacturer's support department. > Nvidia is very sparing with the information in their specs for the > Quadro 400 GPU. You may be out of luck without repalcing the hardware, then. (That's one of the prices of buying undocumented hardware) > Assuming I can find a GPU that supports/offers 4bpp, does anyone know > an X-server product/project that can provide 4-plane overlays? You probably do not need a GPU. The era when 4-bit visuals were common was full of dumb memory-mapped framebuffers; modern CPUs are fast enough that they can probably fake up a 4-bit overlay visual and still run at least as fast as the hardware your client software was designed to run against. I'm not sure how hard it would be to do. My X server hackery has never included faking something the hardware doesn't support, so my experience is rather limited in that direction. But I've done DDX layers for at least three widely disparate framebuffers, and I feel reasonably confident what you want could be done. /~\ The ASCII Mouse \ / Ribbon Campaign X Against HTMLmo...@rodents-montreal.org / \ Email! 7D C8 61 52 5D E7 2D 39 4E F1 31 3E E8 B3 27 4B
Re: X11 expertise on ancient HW sought... (4-plane visual (overlay) via X-server on MS-WIndows)
Mouse, thanks very much for taking the time to comment. On 06/06/2015 02:35 PM, Mouse wrote: >> Am I correct to assume that the GPU must support 4bpp in order for it >> even to be possible for the X-server to propagate a 4-plane visual to >> a client? > No. However, it is substantially more difficult for an X server to > present a visual that the hardware doesn't support, which is probably > why the server you have doesn't do it. Well-behaved clients must be > prepared to handle whatever capabilities the server presents; the > problem here is that the client you have handles the server you have by > complaining and dying. > > Depending on what other capabilities the client is using, you might be > able to get the overlay effect using colourmap hackery with > comparatively small code changes. What PseudoColor or DirectColor > visuals are available? If you are talking about hacking the x-client, no chance. At least, that would probably entail unacceptable cost for the customer, who does not have the source. >> If yes, how can I determine if a GPU supports 4bpp? > Read the documentation, of course, or contact the manufacturer's > support department. Kind of hopeless in the case of manufacturers of mass-market hardware... >> Nvidia is very sparing with the information in their specs for the >> Quadro 400 GPU. > You may be out of luck without repalcing the hardware, then. (That's > one of the prices of buying undocumented hardware) The Quadro 400 is coincidence. The test hardware was randomly chosen before all this 4-plane stuff was known. The solution can be anything reasonable. >> Assuming I can find a GPU that supports/offers 4bpp, does anyone know >> an X-server product/project that can provide 4-plane overlays? > You probably do not need a GPU. The era when 4-bit visuals were common > was full of dumb memory-mapped framebuffers; modern CPUs are fast > enough that they can probably fake up a 4-bit overlay visual and still > run at least as fast as the hardware your client software was designed > to run against. > > I'm not sure how hard it would be to do. My X server hackery has never > included faking something the hardware doesn't support, so my > experience is rather limited in that direction. But I've done DDX > layers for at least three widely disparate framebuffers, and I feel > reasonably confident what you want could be done. Depending on the cost, hacking the X-Server might be an option. I have presented MS-Windows as a given here, but in fact I could also use Linux. Do any possibilities involving Linux occur to you? cheers, Rob
Re: X11 expertise on ancient HW sought... (4-plane visual (overlay) via X-server on MS-WIndows)
> Mouse, thanks very much for taking the time to comment.
You're welcome! I've received so much help over the years from the
net, paying it back in some small part is the least I can do. Even in
a case like this where I can't really help except in generalities.
>> [...] comparatively small code changes [...]
> [...] does not have the source.
Oh. Yeah, that runs the cost of that option up. Substantially. :-(
>> You probably do not need a GPU. The era when 4-bit visuals were
>> common was full of dumb memory-mapped framebuffers; modern CPUs are
>> fast enough that they can probably fake up a 4-bit overlay visual
>> and still run at least as fast as the hardware your client software
>> was designed to run against.
>> I'm not sure how hard it would be to do. [...]
> Depending on the cost, hacking the X-Server might be an option. I
> have presented MS-Windows as a given here, but in fact I could also
> use Linux. Do any possibilities involving Linux occur to you?
Nothing specific, because I don't use Linux - and, given what I've seen
in my brushes with Linux-land, I am inclined to doubt that anyone has
built out-of-the-box support for such a thing.
But I would hope Linux is capable of presenting a dumb 24bpp
memory-mapped framebuffer to userland, in which case an X server could
be built which presents whatever it wants to clients and then composits
it all into the framebuffer, with the main CPU if necessary. (This
would make colourmap changes comparatively expensive if the underlying
hardware is TrueColor instead of DirectColor, but only _comparatively_
expensive - I'd guess such a change could probably be done within a
single vertical retrace.)
However, as I said, this is not something I've ever personally tried,
so I have only wild guesses at how easy it would be. At least with
Linux on the (X) server host, I become much more confident it's just
("just", hah!) a SMOP.
/~\ The ASCII Mouse
\ / Ribbon Campaign
X Against HTMLmo...@rodents-montreal.org
/ \ Email! 7D C8 61 52 5D E7 2D 39 4E F1 31 3E E8 B3 27 4B
Re: X11 expertise on ancient HW sought... (4-plane visual (overlay) via X-server on MS-WIndows)
Mouse, thanks very much for taking the time to comment. On 06/06/2015 02:35 PM, Mouse wrote: >> Am I correct to assume that the GPU must support 4bpp in order for it >> even to be possible for the X-server to propagate a 4-plane visual to >> a client? > No. However, it is substantially more difficult for an X server to > present a visual that the hardware doesn't support, which is probably > why the server you have doesn't do it. Well-behaved clients must be > prepared to handle whatever capabilities the server presents; the > problem here is that the client you have handles the server you have by > complaining and dying. > > Depending on what other capabilities the client is using, you might be > able to get the overlay effect using colourmap hackery with > comparatively small code changes. What PseudoColor or DirectColor > visuals are available? If you are talking about hacking the x-client, no chance. At least, that would probably entail unacceptable cost for the customer. >> If yes, how can I determine if a GPU supports 4bpp? > Read the documentation, of course, or contact the manufacturer's > support department. Kind of hopeless in the case of manufacturers of mass-market hardware... >> Nvidia is very sparing with the information in their specs for the >> Quadro 400 GPU. > You may be out of luck without repalcing the hardware, then. (That's > one of the prices of buying undocumented hardware) The Quadro 400 is coincidence. The test hardware was randomly chosen before all this 4-plane stuff was known. The solution can be anything reasonable. >> Assuming I can find a GPU that supports/offers 4bpp, does anyone know >> an X-server product/project that can provide 4-plane overlays? > You probably do not need a GPU. The era when 4-bit visuals were common > was full of dumb memory-mapped framebuffers; modern CPUs are fast > enough that they can probably fake up a 4-bit overlay visual and still > run at least as fast as the hardware your client software was designed > to run against. > > I'm not sure how hard it would be to do. My X server hackery has never > included faking something the hardware doesn't support, so my > experience is rather limited in that direction. But I've done DDX > layers for at least three widely disparate framebuffers, and I feel > reasonably confident what you want could be done. > I have presented MS-Windows as a given here, but in fact I could also use LInux. Do any possibilities involving Linux occur to you? cheers, Rob
Re: X11 expertise on ancient HW sought... (4-plane visual (overlay) via X-server on MS-WIndows)
On Thu, Jun 04, 2015 at 08:24:31PM +0200, IMAP List Administration wrote: > [also posted to comp.graphics.x today] > > Hello Folks, > > I'm trying to get an application that currently uses a local display > on an ancient DEC Alpha workstation with a (for the time) > mid-to-high-end graphics controller (ZLX-E2) to instead use an > X-server running under MS-Windows. > > The application is complaining that it cannot find a "4/5-bit > visual". It almost certainly wants to use this visual for an > overlay, as the application displays moving objects superimposed on > a map. > [...] > I have tested VcXsrv, Reflection-X, Exceed (with 3D option), X-Win32 > and even the ancient DEC Pathworks X-server eXcursion with no > success. I'm working on getting an evaluation copy of PTC's MKSTools > X/Server. Of the X-servers I've tested, Exceed seems to offer the > most configuration parameters. > > I'm not even sure the Quadro 400 can handle 4bpp "visuals", or > whatever MS-Windows calls them. In fact, I wonder if any modern > hardware offers 4bpp capability. On my Linux box with a GeForce GT > 430 I don't have any 4-plane visuals, and xprop doesn't mention any > overlays either. Pardon my ignorance, I'm definitely not an expert and after sending this msg I am going to hide and pretend to be dead. However, I cannot see if you tried running your app on Linux-based X server. This might be easy to do and does not even require installing Linux. When I want to test some new hardware, I usually download GRML onto pendrive and boot from it. It's small, really, like half of CD-ROM. And it has relatively new X-server, plus few other things like compiler and Open Office, but in this case it doesn't matter. https://grml.org/ In case you go this way, it may make sense to read some manpages and experiment with options to server - different color depths, screen size (a.k.a. resolution). Perhaps there are some options to the client (DEC app) too, and maybe there is a match. BTW, is the problem related to just one app your customer wants? Can you run xterm from your DEC and display it on new X-server? I'm not sure what GPU has got to do with 4 bits per pixel stuff. I have always thought the problem with bpp is only when graphics chip doesn't have enough memory to handle, say, 1024x768 pixels, each containing 4 bits (16 colors mode), i.e. it has less than 384kB of RAM. And in such case, every app could have its own colormap, which caused display to blink in a psychodelic way. The last time I even bothered was probably when I added 1MB of video memory to empty socket on S3 video card, which allowed me to run X in 800x600 24bpp mode... about 17 years ago. So I think it would be a bit strange to have such problem today. Anyway, I'm definitely not an expert. Consider my questions as controls or noise, I have no idea what they are. Now I will run and duck. -- Regards, Tomasz Rola -- ** A C programmer asked whether computer had Buddha's nature. ** ** As the answer, master did "rm -rif" on the programmer's home** ** directory. And then the C programmer became enlightened... ** ** ** ** Tomasz Rola mailto:tomasz_r...@bigfoot.com **
Re: X11 expertise on ancient HW sought... (4-plane visual (overlay) via X-server on MS-WIndows)
On Sat, 6 Jun 2015, IMAP List Administration wrote: > Depending on the cost, hacking the X-Server might be an option. > I have presented MS-Windows as a given here, but in fact I could also use > Linux. > Do any possibilities involving Linux occur to you? What kind of use do you require here for your app -- is that a one-off installation or a more widespread deployment? For a single use your simplest solution might be using similar hardware. ZLX-E2 is the PMAGD-B or HX+ TURBOchannel option (24-bit DirectColor, no Z-buffer, using a Bt463 RAMDAC), also known as SFB+ (Smart Frame Buffer+) after the graphics ASIC. A 5V PCI variant was manufactured under the name of ZLXp-E2, more commonly known as TGA (I don't know offhand what the acronym expands to). Both are supported under Linux, using the `tgafb' kernel driver and a corresponding X11 DDX module. So you might try this instead, chasing a 24-plane PCI TGA board should be doable. The board has no VGA compatibility AFAIK so there should be no odd issues with legacy address decoders or initialisation on non-Alpha systems -- the board should simply sit there quietly until initialised by the kernel driver and then taken over by the X server. The DDX module for the X server might not be built by default for TGA on non-Alpha systems though, so you might have to do some tweaking. Just a thought, maybe it'll help. Maciej
Re: X11 expertise on ancient HW sought... (4-plane visual (overlay) via X-server on MS-WIndows)
Sorry for undiggind this subject so many years later, but have you found a solution to your problem? I am facing exactly the same issue, and i have tried all possible windows X server options without success. The only possibility I am working with is with a Mac OS X server, which reports a 5-bit visual.
Re: X11 expertise on ancient HW sought... (4-plane visual (overlay) via X-server on MS-WIndows)
On 29 January 2018 at 14:22, Dimitris Theodoropoulos via cctalk wrote: > Sorry for undiggind this subject so many years later, but have you found a > solution to your problem? I am facing exactly the same issue, and i have > tried all possible windows X server options without success. The only > possibility I am working with is with a Mac OS X server, which reports a > 5-bit visual. What bit-depth do you require on the X-server? David
Re: X11 expertise on ancient HW sought... (4-plane visual (overlay) via X-server on MS-WIndows)
I believe that my case is identical to the original message of the list and 24-bit is required. The problematic visual (the one which is not provided by the external X-server) is the following (I cite an extract from xdpyinfo on the original system): * visual:* *visual id:0x36* *class:PseudoColor* *depth:4 planes* *available colormap entries:16* *red, green, blue masks:0x0, 0x0, 0x0* *significant bits in color specification:4 bits* At another part of xdpyinfo, I also get the following info: *screen #0:* * dimensions:1280x1024 pixels (342x274 millimeters)* * resolution:95x95 dots per inch* * depths (4):8, 12, 24, 4* Please forgive my technical ineptitude, in case I did not answer your question, but I am not experienced in this domain. D. Theo 2018-01-30 12:42 GMT+02:00 David Brownlee : > On 29 January 2018 at 14:22, Dimitris Theodoropoulos via cctalk > wrote: > > Sorry for undiggind this subject so many years later, but have you found > a > > solution to your problem? I am facing exactly the same issue, and i have > > tried all possible windows X server options without success. The only > > possibility I am working with is with a Mac OS X server, which reports a > > 5-bit visual. > > What bit-depth do you require on the X-server? > > David >
Re: X11 expertise on ancient HW sought... (4-plane visual (overlay) via X-server on MS-WIndows)
On 31 January 2018 at 08:21, Dimitris Theodoropoulos wrote: > I believe that my case is identical to the original message of the list and > 24-bit is required. > The problematic visual (the one which is not provided by the external > X-server) is the following (I cite an extract from xdpyinfo on the original > system): > > visual: > visual id:0x36 > class:PseudoColor > depth:4 planes > available colormap entries:16 > red, green, blue masks:0x0, 0x0, 0x0 > significant bits in color specification:4 bits > > At another part of xdpyinfo, I also get the following info: > > screen #0: > dimensions:1280x1024 pixels (342x274 millimeters) > resolution:95x95 dots per inch > depths (4):8, 12, 24, 4 > > Please forgive my technical ineptitude, in case I did not answer your > question, but I am not experienced in this domain. OK, so it looks like it might be looking for a PseudoColor display, which I'm not sure is supported by current X11 (*), check if your MacOS server is supporting PseudoColor - I think Apple dropped it around a decade or so ago. You *might* be able to run Xephyr to present a sub XServer which can run 8 bit PseudoColor (its freely available under (*nix), not sure about Windows. (*) Current X11 on x86 looks to have dropped PseudoColor, though there may be some odd VESA or other target drivers. Current X11 on non x86 hardware still supports 8bit, 4bit and other crazy historical bitdepths, which is awesome in its own right, but probably not directly useful in this context. So if it is PseudoColor and your target display is Windows, some increasingly crazy options: Native Windows X servers - Look for a Windows Xephyr port and run under MobaXterm or similar - Find a really ancient (read, probably crashes a lot and may not work on recent Windows) Windows X server which support PseudoColor - Fire up a Linux VM and run Xephyr in there with the display set to MobaXterm Run your actual other OS image on a really fast PC in an emulator which simulates the actual display hardware it needs Using some other OS in a VM full screen as an X server - Check to see if there is some way to get the Linux VM to use a VESA or some other driver which comes in 8 bit - Fire up an ancient Linux (or similar) with X11 PseudoColor support in a VM - Fire up a modern (or even older) *nix for hardware which is native 8 or 4 bpp in qemu (I said increasingly crazy)
