Show Posts

This section allows you to view all posts made by this member. Note that you can only see posts made in areas you currently have access to.

Messages - Olafur Gunnlaugsson

Pages: [1] 2 3 ... 15
Setup & Installation / Re: OS2 Warp 4.52 "vdi2hardware" result
« on: June 06, 2021, 03:07:41 pm »
I have done this for zonks, I believe there was a small article on the eComStation Facebook page that I wrote on the subject ca 2011, that was removed by Facebook when they changed the rules on article and file permissions when they split groups into groups and pages. I never bothered restoring it elsewhere since no-one ever showed any interest in the technique anyway, just like there was no interest in running OS/2 in a Hyper-V container.

Although in general it is very helpful, there are a few issues doing this, while VB etc. are highly compatible with real hardware there are edge cases of incompatibility, you can end up with an image that boots on VB but not on real hardware and vise versa. It can sometimes by tricky to place an image onto a drive correctly especially if you have multiple partitions and boot drives, for those cases investing in DFSee becomes an absolute must, alignment issues are frequent with any sort of solid state media for instance. The weird behavior of the VirtualBox image manager in recent versions can make maintaining a large  number of VBox images and copies a real pain (unlike VirtualPC, VBox registers (and re-registers) each image into a database). Each version of VBox introduces a new collection of bugs which can throw a spanner into the works, these get fixed eventually but OS/2 is not high upon their priority list (and they are downright hostile to ArcaOS), not a huge issue with images but this makes install in-place frequently impossible on modern VBox which was relatively easy on older VBox versions. In-Place installs were extremely convenient since you could remount the drive on VBox in you had any problems on real hardware. On older versions of OS/2 getting data onto the virtual machine can in some cases be much more time consuming than on real hardware, weirdly enough. etc etc etc

I only really use this technique these days for when installing "legacy" versions of the system, that means I do not have to bother with floppies and I can get the system into a fixpack state that will actually run on the hardware (memory & disk size & driver issues etc). Apart from that and for the fact that I already have a large collection of ready made VBox images, I am not altogether sure I would use the technique except for installing onto systems that run on consumer solid state media eg. embedded systems with SDCards, thin clients with Compact Flash drives or old IDE PC's with CF adapters. In those cases it is very quick, just plonk a converted VBox image onto start of the SD drive and then create other partitions as needed from the install.

I am currently setting up a couple of old VIA C3 systems for compatibility testing, basically each system has 2 or 3 "legacy" Windows installs on primary partitions and a host of various "common" OS/2 versions in an extended partitions. I had intended to convert and copy already made basic installs from VBox onto the extended partition but in the end it actually made more sense time wise to buy a floppy emulator off eBay, load it with a large collection of OS/2 installer floppy images and install from there.

Using PCE rather than VirtualBox also has some advantages since Windows can mount the PCE images directly, you can load software onto the PCE image that contains a windows readable partition type directly from the file manager or folder. It also has a better hardware compatibility in some cases.

And remember you can also go the other way round, DFSee can easily convert your existing OS/2 installs into VBox images.

Hardware / Re: IBM Thinkpad with floppy drive
« on: May 26, 2021, 08:42:47 am »
As far as an UltraBay Floppy that can be swapped in place of the CD drive, I think the T540p series is the last one.

I thought the ultrabay 2000 was the last ultrabay that was hooked up to the legacy controller, was there an internal floppy drive available for the 540?

Hardware / Re: IBM Thinkpad with floppy drive
« on: May 25, 2021, 12:12:00 pm »
What is the latest model of thinkpad with integrated floppy drive ?
Thank you.
IBM stopped supplying floppie drives as standard back in the 90's, they replaced them with "Ultrabay" a hot swappable slot that can take CD-ROM's, floppies or batteries etc., depending on need. I think the last Thinkpad series that came standard with a floppy were some of the 300 series (486 era), AFAIK all the newer models had the floppy drives as Ultrabay options, the last models that could take an Ultrabay with an internal Floppy (Ultrabay 2000) were the T2x series and the A3x series.

Programming / Re: Installing Prominaire Designer
« on: April 30, 2021, 11:44:53 am »
You can try downloading the demo version, from the publishers old web page on the Internet Archive if it is not on one of the FTP sites, install it and then installing your archive over that.

Hardware / Re: Hardware Wiki report
« on: March 28, 2021, 08:26:08 pm »
Some USB audio devices were also tested, all with the same hardwareID: 8086:0808
Fairly certain this is the default HiD for the C-Media CM108 USB codec

Multimedia / Re: Updated Uniaud32 build
« on: March 21, 2021, 11:59:01 am »

  Thanks for the head's-up... it is working well here. One thing I notice - with your build on my machine, the driver uses IRQ16, but David A's version uses IRQ55. Does he add some secret sauce for that? Is there an advantage to a higher IRQ?

Yes, in that AT compatibles had 15 interrupts available while early APIC based system had 24 interrupts, with the extra 9 interrupt lines being handled dynamically in a way that could cause problems with certain drivers and hardware, in some cases drivers either did not see any interrupt above 15 or saw any interrupt above 15 a IRQ16 even though the hardware itself supported higher IRQ's. Modern MSI based APIC systems have 224 interrupt lines available, BTW.

A number of old OS/2 drivers will only work correctly with interrupts <16 or <25 and a number of them showing problems with IRQ's higher than 15. While sharing an interrupt is possible both from a hardware and software perspective it can have side effects both on the software side and with the hardware itself such as the interrupt switching modulating the analogue gain stages on a soundcard or a video card with analogue (VGA) outputs. And there are some other issues, drivers like the old Martox ones being quite aggressive in claiming interrupts, clashing with other dirvers and sometimes with the OS itself.

By moving as much of the modern drivers/hardware to interrupts higher than 24 you are simply leaving more space free to legacy drivers and hardware to play with, so to speak, minimising the possibility of software/hardware clashes.

Yeah, good catch

Storage / Re: FAT32 - Netlabs or AN?
« on: March 17, 2021, 10:11:01 am »
It is not legal to use FAT32, without paying Microsoft for the right to use it (even if it is only a dollar).

Yes it is, they ran out years ago or were invalidated with the exception of the long filename patent that was invalidated everywhere except for the USA, and is easily circumvented anyway.

exFAT is another issue.

I don't get this about Win10 support. Are you saying that they won't work on Win10 either?
All the USB audio 1 devices I have do work under Win10. But I have not tested headphones as I don't have a USB headphone.

Windows 10 recently dropped class1 (1998) support on Win10 devices, none of my early USB DAC's work on Win10 since late fall last year, some others are 1.1 compliant and seem to work, some are supposedly 1.1 but do not work or work only partially (Phillips CODEC's only work if volume is 100%), some are class one compliant and are no longer seen as class compliant devices but have drivers supplied on the windows distribution (CMedia et al) and work, but others like Roland/EDIROL/Boss units no longer work in class compliant mode but work in "driver mode", if you manage to install older drivers that is. Others like Focusrite Saffire 6 USB 1.1 only work with Windows using the ASIO drivers which is a world of pain for those that do not know what ASIO is and how to route it to the system.

Note that there were 2 versions of the Saffire 6, a USB1.1 and USB 2 version, the latter still works, but both are only about a decade old, Ploytec only started shipping their asynchronous USB 2 development package in 2011, it was all USB 1 up until then. CMedia also has the bad habit of supplying different chips with a very similar or sometimes even the same name, some early CMedia USB audio chips no longer work with either the Win class drivers or CMedia drivers, even though the CMedia drivers supposedly work with that part number, turns out that they recycled the part number for a later and different USB CODEC.

I am not necessarily unhappy about this development, I have been buying weird old USB audio products for peanuts off eBay in the last couple of months. USB powered digital amplifiers and old hi-fi DAC's with programmable DAC chips and isolation transformers on the audio output, so on, so forth.

Hardware / Re: Motherboard audio in Warp 4.52
« on: February 24, 2021, 10:24:06 pm »

There are some old Ensoniq driver on hobbes, but I would try Uniaud first.


The Ensoniq driver on hobbes et al, are for Ensoniq ISA cards not the Audio PCI card. Ensoniq was bought out by Creative and the Audio PCI was re-released as Sound Blaster PCI 128, PCI 512 and so on, using the same Ensoniq control chip but using a inferior AC97 chip as a DAC, rather than the reasonably good DAC on the original. In general you should search for SB PCI 128 drivers rather than Ensoniq.

Since it does install correctly, announces itself correctly and the same results happen with a SB Live with Sander's completely different set of drivers my guess is either a simple misconfiguration in config.sys or corrupted MMINSTALL config, corrupted or partially missing MM subsystem, or other OS to MM subsystem misconfiguration. It is worth a try uninstalling the MM subsystem and reinstalling it from the original CD, or at the least manually taking a look t the configuration files in the MM subfolder.

It's strange that Linux will not accept audio class 1 devices (at least not as a default). Everything that is simple and still about 90% of all USB audio devices offered are audio class 1 devices.

The Linux kernel is monophilic, every device driver lives in the kernel. They can be dynamically loaded as well. It really depends on the choices the person/group compiling the kernel chooses. Lots of dists are based on Ubuntu, which seems to compile most everything in. Ubuntu is based on Debian, so perhaps Debian compiles everything in. Examples are HPFS and JFS support. Ubuntu derivatives support them out of the box. Other Linuxes, not so much.
So really it depends on the distribution. The kernel supports both.

Not 100% sure on its accuracy since I deleted my last Linux partition a decade ago after using it since v0.1, but I have had numerous complaints about usba1 audio devices not working in Debian that work fine in Ubuntu (I ran a Linux audio site in the 90's, and still get emails 20 years later). I am personally more annoyed about Windows 10 support sine it means millions of headsets are no longer working.

DSD enabled devices mostly use an XMOS front end (Although ESS has recently released their own frontend for testing), while they mostly default to usba1, they should switch to usba2 if they get AudioClassRequests_2 according to my understanding of the XMOS docs[design-guide](6.16.1alpha1).pdf

Also come to think of it, anything with class compliance and support for frequency rates above 96k is bound to be class 2 compatible.

Can anyone point me to a "audio standard 2.0" device ? The very cheap devices usually implement "audio standard 1.0" but I need "audio standard 2.0" in order to test. Hopefully, I can find something that is not overly expensive.

Here is how you can find out about what you have:
1) get lsusb.exe. This tool allows you to list the complete descriptor contents of a USB device
2) run "lsusb.exe". It will list all attached devices with their vendorid:productid. You will have to guess what is what by attaching and detaching USB devices.
3) run "lsusb.exe -v -d vendorid:productid where "vendorid" and "productid" are the numbers from step 2.Pipe the tool output into a file.
4) in that file, search for something like this:
      AudioControl Interface Descriptor:
        bLength                 9
        bDescriptorType        36
        bDescriptorSubtype      1 (HEADER)
        bcdADC               1.00
        wTotalLength           38
        bInCollection           1
        baInterfaceNr( 0)       1

If "bcdADC" says 1.00 then you have an audio standard 1.0 device. If it says 2.00, then you have an audio standard 2.0 device.

The problem is that if I search in Amazon or elsewhere, they never give that level of detail in their stupid product descriptions.

Anything that advertises Linux compatibility? The Linux kernel has not had usba1 compiled in as standard for at the least 6 years although some distributions like Ubuntu do supply kernels with usba1 enabled. As I am still on Firewire mostly I cannot help you as the only USB devices that I am certain of have usba2 are Micromega MyDAC and Arcam rPAC, both have switchable class compatibility but are relatively expensive, although the rPAC is frequently going cheap on eBay these days because it switches back to usba1 on every restart, which means incompatibility with modern Linux and Windows if you do not, or are not aware of how to switch it back.

Applications / Re: PMMail Where to buy
« on: February 10, 2021, 02:00:24 am »
Hi there once again,

which is the best email address to contact voice concerning my payment and the license for pmmail. THe mails I sent are all returned?


Try the head honcho himself:

Pages: [1] 2 3 ... 15