Nextor 2.1.0 beta 1 released

Paladin (998)

08-02-2019, 17:42

And now, folks, it's your turn: I really need help to fix the known bugs. I haven't been able to reproduce many of them (I'm especially worried about the freezing Sunrise IDE one). Repro steps, hints, pull requests... anything is welcome.

And in case you haven't noticed it... I've added a link to NextorBuilder in Nextor's README. That's a nifty project that allows to compile Nextor on non-Windows machines, by using Vagrant and a virtual machine. So now everybody can build Nextor at home! :)

By mfeingol

Champion (266)

09-02-2019, 04:25

Konamiman: thanks for what you do! I have a bug report.

I have a MegaFlashROM SCC+SD with a single slot and no additional memory. I'm using it on an ST with stock 256kB. I've been on 2.1.0 alpha for a long time.

When I flash 2.1.0 beta 1, the MSX hangs after printing "SD card slot 1:SDHC". However, 2.0.5 beta 1 works fine.

I'm flashing as follows:
- Make a 1GB FAT partition on an SDHC card. Format the partition using Windows Explorer.
- Copy Nextor-2.1.0-beta1.MegaFlashSDSCC.1-slot.Recovery.ROM to the partition, using Windows Explorer.
- Insert the SDHC card in the MFR SCC+SD
- Boot the MSX pressing cursor Up to enter the recovery menu
- Press F4 to flash the Nextor kernel

By gdx

Prophet (2612)

09-02-2019, 08:54

I tried this new version with the Sunrise driver on the FS-A1ST (1024kB).

- I can boot on FAT16 partition (CF 1GB). So I create a partition of 32MB in FAT12 and add a second in FAT16, CALL FDISK displays bad information when I create the second partition but FAT12 partition works. By cons the FAT16 partition is unuable (same as with one partition: "Error not formated DOS2"). I use COMMAND2.COM v2.44.

- NEXTOR crashes when I insert the line below in AUTOEXEC.BAT.
IF %_MSXVER%==MSX SOFARUN.BAT

Question: How do I copy files from one partition to another on a MSX without floppy disk drive when I have one flash card only? Because now in this case there is only one drive to assign a partition.

konamiman wrote:

I really need help to fix the known bugs. I haven't been able to reproduce many of them (I'm especially worried about the freezing Sunrise IDE one).

I noticed a crash when a CF formatted for the Sunrise interface was insered in slave slot.

By konamiman

Paladin (998)

09-02-2019, 23:59

mfeingol wrote:

When I flash 2.1.0 beta 1, the MSX hangs after printing "SD card slot 1:SDHC". However, 2.0.5 beta 1 works fine.

Damn. I must have messed up something in the last minute. I have updated the kernel files for MFRSD in GitHub, they should work fine (or boot, at least!) now.

By konamiman

Paladin (998)

10-02-2019, 00:19

gdx wrote:

I can boot on FAT16 partition (CF 1GB). So I create a partition of 32MB in FAT12 and add a second in FAT16, CALL FDISK displays bad information when I create the second partition but FAT12 partition works. By cons the FAT16 partition is unuable (same as with one partition: "Error not formated DOS2"). I use COMMAND2.COM v2.44.

I need more details, as I can't see this happening. Pretty please, describe step by step how to reproduce this and tell me what exact "bad information" are you seeing. But preferably as a new issue in GitHub, instead of here (feel free to post here a link to the issue for reference, of course).

By konamiman

Paladin (998)

10-02-2019, 00:08

gdx wrote:

Question: How do I copy files from one partition to another on a MSX without floppy disk drive when I have one flash card only? Because now in this case there is only one drive to assign a partition.

You can assign extra drives for other partitions using MAPDRV.COM (or CALL MAPDRV).

By konamiman

Paladin (998)

10-02-2019, 00:17

gdx wrote:

NEXTOR crashes when I insert the line below in AUTOEXEC.BAT.
IF %_MSXVER%==MSX SOFARUN.BAT

What's your %_MSXVER%? What happens if you run SOFARUN.BAT directly? What about "IF %_MSXVER%==MSX echo foobar"? (or running a different BAT) What if you run that outside AUTOEXEC.BAT? Again, I need more details, please

By Louthrax

Paragon (2012)

10-02-2019, 10:21

Thanks Nestor for that new release.

I was playing with Nextor on my old VG-8235, with only 128KB or RAM yesterday, and was wondering if there's a way to disable the TAB autocompletion ? That should free an extra 16KB segment right ? So having 2 free 16KB segments instead of just 1 on 128KB machines.

That can make a difference for some of my apps here, that could be tweaked a bit to work with 2 segments (but not hope with only 1).

Also, that extra segment could be used for memoy-mapper ROM emulation mode of SofaROM (no need for the TAB completion when running a ROM, and an extra 16KB segment can significantly improve things when emulating all the 8KB ROM pages combinations...).

By gdx

Prophet (2612)

10-02-2019, 10:51

konamiman wrote:

I need more details, as I can't see this happening. Pretty please, describe step by step how to reproduce this and tell me what exact "bad information" are you seeing.

I tested quickly, I do not remember exactly. It was something like "32M partition added".

konamiman wrote:

You can assign extra drives for other partitions using MAPDRV.COM (or CALL MAPDRV).

Thanks. It works.

konamiman wrote:

What's your %_MSXVER%?

It indicate the MSX generation. This condition is to launch Sofarun on my MSX1 otherwise Multi-mente.

konamiman wrote:

What happens if you run SOFARUN.BAT directly?

Sofarun is executed and works.

konamiman wrote:

What if you run that outside AUTOEXEC.BAT?

I define some pathes before the condition to execute MM or sofarun according the MSX generation.

Please help me ! I would like back to the original firmware because finally I noticed that the FAT12 does not work perfectly. The FAT is corrupt no matter how I copy the files, but I can't!

When I use IDEFLLOAD I get the following message: "This ROM is currently used by the system! Erasing aborded. (Try again in DOS1 mode; boot with [ins] pressed)"

However I boot in DOS1 and [ins] has no effect.

By Louthrax

Paragon (2012)

10-02-2019, 11:20

Louthrax wrote:

I was playing with Nextor on my old VG-8235, with only 128KB or RAM yesterday, and was wondering if there's a way to disable the TAB autocompletion ? That should free an extra 16KB segment right ? So having 2 free 16KB segments instead of just 1 on 128KB machines.

Seems more related to the version of COMMAND2.COM (2.40 uses 1 more extra segment compared to 2.31).

By Alexey

Guardian (1985)

10-02-2019, 21:47

Here's my bug report, it affects earlier Nextor (Alpha), also probably applicable to the Beta. On KYBT and KYBT2 systems (Russian versions of YIS503II and YIS503III) the keyboard behaves differently than on other machines - the number keys are shifted to the right. For example if you press '9' at Nextor's bootup screen, it acts as if '0' is pressed on other machines to stop using the previously assigned bootable image. This makes switching disk images and ignoring booting of the assigned DSK image a bit difficult.

Just tried Nextor Beta1 on my Carnivore2. Flashed the "Nextor-2.1.0-beta1.SunriseIDE.ROM" file into the cartridge, copied the "Nextor.sys" and tools into the root folder. As a result the cartridge no longer boots to MSX-DOS2. The system just hangs after Nextor's copyright message.

WARNING! Do not try this Beta1 on Carnivore2 unless you have the second C2 cartridge or something else to boot to MSX-DOS! If your C2 gets unbootable with this Beta1 Nextor BIOS, the only way to restore it would be booting from another IDE-capable device, running the C2MAN utility and flashing the old Nextor IDE BIOS into the C2 cartridge.

@Konamiman, do you have Carnivore2 to test your software? There's a lot of people with C2 these days, so making the next version of Nextor C2-compatible would be really nice. I created reports for both issues on Github.

By edoz

Prophet (2122)

10-02-2019, 19:20

Wow! Cool! I will do some tests on the Sunrise IDE. (But last time i was not able to re-flash it again..) So not sure if it would be possible to test on that device Seems some hardware issue. Nextor is my favorite software! Thank you so so much for your hard work on this piece of software! Without a proper OS we can't use MSX

By mfeingol

Champion (266)

10-02-2019, 21:17

konamiman wrote:

I must have messed up something in the last minute. I have updated the kernel files for MFRSD in GitHub, they should work fine (or boot, at least!) now.

I redownloaded and flashed 2.1.0 Beta 1. The ST now boots fine. Thanks for the fix.

By gdx

Prophet (2612)

11-02-2019, 08:29

I flashed again the "Nextor-2.1.0-beta1.SunriseIDE.ROM" without using floppy disk drive,
I used a Mega-sram. Luckily I had already made a image file containing the necessary to flash the firmware. I had to make a hardware remaniement so that the cartridge did not start.

Good new Nextor works much better! Even the condition with %_MSXVER% works. So forget all that I said above. Everything seems more stable than with previous versions. I will surely keep Nextor from now on.

I noticed only following problems.

- I still can not use my SD to CF adapters.
- The system freezes when I do not insert the second CF card.
- Would not it be better to keep the [INS] key to disable the cartridge as with the Sunrise firmware? I added it myself for next time just in case.

I hope you can secure the flashing, it will avoid problems. Anyway, thanks to Nestor and P. Punk for your efforts.

By Eugeny_Brychkov

Paragon (1063)

11-02-2019, 08:33

CTRL-STOP issue in PING.COM (and other software using TCP/IP UNAPI switching slots) I reported earlier?
What is the last stable release?

By gdx

Prophet (2612)

11-02-2019, 09:32

There are still some FAT corruption issues.

By konamiman

Paladin (998)

11-02-2019, 10:24

Eugeny_Brychkov wrote:

CTRL-STOP issue in PING.COM (and other software using TCP/IP UNAPI switching slots) I reported earlier?

I've been able to reproduce it, but my attempts to get to the root cause of the problem have been unsuccessful so far. It would be great if I could get some more eyes to look at the code.

gdx wrote:

There are still some FAT corruption issues.

Please write a report for this but with excruciating detail. Write down all the steps including how you generate the partitions with FDISK, how you copy files to it, and how you detect the corruption issues. If possible also provide a link to a DSK file containing some of the files you are copying.

Please bear in mind that I don't own any Sunrise IDE (or compatible) device, so all I can do is testing with emulators.

By Eugeny_Brychkov

Paragon (1063)

11-02-2019, 12:27

konamiman wrote:

I've been able to reproduce it, but my attempts to get to the root cause of the problem have been unsuccessful so far. It would be great if I could get some more eyes to look at the code.

Emegrency exit routine (there's an address being jumped to when CTRL-STOP is pressed) must explicitly restore slots in CPU banks 1 and 2.

I do not remember exactly, sound like this one

F302h 2 Pointer to abort handler. (MSX-DOS)

By sd_snatcher

Prophet (2945)

11-02-2019, 12:43

@konamiman

Congrats on the new version!

I really would like to help with the issues I opened on GitHub and with more testing, but right now I'm out of free time even for myself. I hope you understand.

By konamiman

Paladin (998)

11-02-2019, 13:06

sd_snatcher wrote:

I really would like to help with the issues I opened on GitHub and with more testing, but right now I'm out of free time even for myself. I hope you understand.

Don't worry, we are all on the same boat But I'm really concerned about all those bugs I'm not able to squash by myself.

By gdx

Prophet (2612)

11-02-2019, 14:09

konamiman wrote:

gdx wrote:

There are still some FAT corruption issues.

Please write a report for this but with excruciating detail. Write down all the steps including how you generate the partitions with FDISK, how you copy files to it, and how you detect the corruption issues. If possible also provide a link to a DSK file containing some of the files you are copying.

Please bear in mind that I don't own any Sunrise IDE (or compatible) device, so all I can do is testing with emulators.

All I can say is that I copy some of my files and directories (the same ones I use for a long time) to the partitions from the Mac. At first it seems to work but after several modifications and additions, some directories and files disappear. Yet when I insert the CF card into the Mac, everything looks ok. And the more I persist, the worse it is. I try with xopy, multimente or Mac only. Result is about same. I have not tried it yet without Multi-Mente.

By konamiman

Paladin (998)

11-02-2019, 16:12

gdx wrote:

All I can say is that I copy some of my files and directories (the same ones I use for a long time) to the partitions from the Mac. At first it seems to work but after several modifications and additions, some directories and files disappear. Yet when I insert the CF card into the Mac, everything looks ok. And the more I persist, the worse it is. I try with xopy, multimente or Mac only. Result is about same. I have not tried it yet without Multi-Mente.

Could you please convert that CF card into a disk image file and send it to me?

By gdx

Prophet (2612)

12-02-2019, 00:46

OK, I send it right away.

By OeiOeiVogeltje

Paragon (1232)

15-02-2019, 14:57

i just flashed the recoveryversion and now ANY sound is on an extremely low volume
i have to crank up the monitor speak to msx to hear just a little bit

other carts work fine

By KdL

Paragon (1146)

16-02-2019, 20:56

Hi Konamiman.

I've tested these kernels with 1chipMSX:
Nextor-2.1.0-beta1.MegaFlashSDSCC.1-slot.ROM
Nextor-2.1.0-beta1.MegaFlashSDSCC.2-slots.ROM

My report:
- The single slot version do not load COMMAND2.COM, sometimes the LED blinks endlessly, sometimes it hangs the system.
- The double slot version is simply okay and also enabling more of two drives the COMMAND2.COM is loaded.
- The single slot version of driver v1.3 by M. Pazos works fine with the base rom of Nextor Kernel v2.1.0-alpha2. So I think there is an issue inside beta1.

By konamiman

Paladin (998)

17-02-2019, 18:21

Regarding the Sunrise IDE version: Piter Punk just sent me an email noting that I'm distributing these versions with an outdated driver, and including a newer one. So I have added a new file to the release in GitHub, the file name is Nextor-2.1.0-beta1.SunriseIDE.driverv017.ROM. Could you all having a Sunrise IDE try this new version please?

By Aquijacks

Champion (257)

19-02-2019, 07:41

Flashjacks uses adapted sunrise Ide driver. I try it.
Altrought I suspect that it will have the same bug that I have compiled.
It does not work on MSX without a disk drive.
Regards.

By gdx

Prophet (2612)

19-02-2019, 12:43

konamiman wrote:

Regarding the Sunrise IDE version: Piter Punk just sent me an email noting that I'm distributing these versions with an outdated driver, and including a newer one. So I have added a new file to the release in GitHub, the file name is Nextor-2.1.0-beta1.SunriseIDE.driverv017.ROM. Could you all having a Sunrise IDE try this new version please?

Thanks, it works much better! All the bugs I met before seem to have disappeared.
Even my SD card to CF adapters is recognized now and I can create partions, but when insered in master slot only and in this case the slave slot is ignored. This is the only (little) problem that seems to remain.

PS: I'm surprised that even with this driver (as original firmware), the Sunrise CF interface does not like some cards when there are two inserted.

Aquijacks wrote:

It does not work on MSX without a disk drive.

Sunrise CF interface do not have this problem.

By ssfony

Resident (58)

24-02-2019, 16:11

Thanks for your hard work!

My experience:
MFRSCC+SD with 2 slots works fine.
Sunrise IDE with driver17 version works fine.

Question1:
Which rom do I use for one chip msx (SX-1)?
Or how do I compile it?

Question2:

When I try to compile, fdisk is left out. I get ROM files that work, but _fdisks just returns to basic.

When I run the compile batch file in the source\kernel\bank5 folder, I get error:
Compiling, please be patient...
fdisk.c:150: error 91: extern definition for 'putchar' mismatches with declaration.
C:/Program Files/SDCC/bin/../include/stdio.h:86: error 177: previously defined here
fdisk.c:1547: error 98: conflict with previous declaration of 'putchar' for attribute 'type' at C:/Program Files/SDCC/bin/../include/stdio.h:86
from type 'int function ( int fixed) fixed'
to type 'void function ( unsigned-char fixed) fixed'
C:\Nextor\Nextor-2.1\source\kernel\bank5>

Why is that?

By sd_snatcher

Prophet (2945)

28-02-2019, 02:01

konamiman wrote:

Regarding the Sunrise IDE version: Piter Punk just sent me an email noting that I'm distributing these versions with an outdated driver, and including a newer one. So I have added a new file to the release in GitHub, the file name is Nextor-2.1.0-beta1.SunriseIDE.driverv017.ROM. Could you all having a Sunrise IDE try this new version please?

Hey, this driver took a heck lot of work to put together. It deserves a proper announcement by itself, doesn't it? :)

---------------------------------

Sunrise IDE driver v0.1.7, by FRS

Almost complete rewrite.
Huge data transfer performance optimisation:
- Z80/3.58MHz: from 141.8KB/s to 171KB/s (20.6% faster)
- Z80/7.16MHz: from 283.6KB/s to 342.4KB/s (20.7% faster)
- Z80/3.58MHz with system timer present: from 141.8KB/s to 178KB/s (25.5% faster)
- R800 from 335.6KB/s to 564KB/s (68% faster)
Much quicker boot up time
Replaced ESC with CTRL+STOP to abort the devices detection on boot. This is a more standard way to abort processes on the MSX, and doesn't conflict with the ESC key that is used by the MegaFlashROM SCC+ to skip the boot from the games
Added MYSETSCR, a routine that handles the screen initialisation, so only the 1st driver sets the screen mode and clears the screen. The philosophy behind this is:
- To avoid multiple drivers to keep clearing the screen one after another like happened on DOS1/DOS2. This was annoying, hard for the user to check the errors when they were printed, and slowed the boot process. It was also nearly impossible to pause the boot sequence unless your machine had a hardware PAUSE button.
- Since the screen isn't cleared by every driver, there's no need for delays to allow the user to see the messages, or to wait for some form of user input. This speeds up the boot process a lot, specially with multiple interfaces. So the artificial delays that were executed before each IDE cartridge initialisation could be removed.
- Removed the copyright messages that were printed by default. Now the driver only prints the minimum necessary information that the user needs to see on every boot, to avoid visual pollution. The copyright message can still be seen, just check the INICHKSTOP below.
- Added INICHKSTOP, to check, at the end of the driver initialization, if the user has pressed STOP at any point before (even on the bootlogo):
  - If STOP was pressed, it then pauses the boot process to allow the user to read any messages or errors he needs.
  - Once paused, INICHKSTOP also offers the user the option to press "i" once, to show the copyright info for this driver.
Replaced all empty loops with hardware frame based delays. This way the delays will always take the same time regardless of the CPU speed, resulting in better turbo support. Only the VDP vertical refresh rate will affect the delays: they'll be 20% slower on 50Hz video systems, but that's tolerable.
Fixed bugs where flags were being checked without waiting for the device BSY flag first
Added data structures to the code, making it easier to maintain and less prone to bugs when the structures need to be changed
Complete rewrite of the ATA/ATAPI detection routine, following the standard detection procedure recommended by the official documentation. This seems to have solved all detection issues that happened until now. The detection is also a lot quicker now.
The detected device type is now shown after its name (CHS, LBA or ATP)
Added preliminary support for ATAPI devices. They are handled by the detection routines now, but the block read/write routines still need more work and must be considered alpha code. There's also preliminary CD-ROM support
The drives are now configured to work in PIO MODE3, as it enables the IORDY signal generation. This solved a lot of data transfer instabilities.
The interrupts are now disabled for all drives. This improved the stability since some devices internal micro-controller code seem to have been freezing by waiting for an interrupt handling that the MSX driver wouldn't perform, as the Sunrise IDE standard doesn't support it
Fixed the device name fetch routine. It now copies the whole device name to a buffer, and only then process the string. This fixed bogus name strings that happened for devices that expect this data transfer to happen at a minimal and constant transfer rate
Developers: Added DIAGERRPRT to report detailed diagnostic information on detection, for easier troubleshooting

By sdsnatcher73

Champion (307)

28-02-2019, 04:09

Wow, indeed an impressive update! Great job. I personally only have a C2 which I think may benefit from this (it also uses Sunrise driver but 0.15 as for now).

By Grauw

Enlighted (7893)

28-02-2019, 08:50

Awesome! The old driver didn't work on my Sunrise CF-IDE, so I'll give this one a shot for sure! Also, nice transfer rates!

By gdx

Prophet (2612)

28-02-2019, 12:55

This new driver for Sunrise has been significantly improved but before you try it, know that it will be difficult to update it next time or go back to original firmware because CTRL+STOP does not prevent the installation of the cartridge. So IDEFLOAD.COM or IDEFL128.COM will refuse to flash again.

By sd_snatcher

Prophet (2945)

28-02-2019, 13:19

@gdx

That was for DOS2. Remember, on Nextor you have a standardized way to prevent a specific kernel+driver from being installed: on boot, just hold the keys “QWER/ASDF” etc, according to the slot where the interface is connected.

For more detail, please read the Nextor manual.

By gdx

Prophet (2612)

01-03-2019, 00:55

Thanks sd_snatcher, it does not work for DOS2 but it works for Nextor.
It's weird to have used so many keys.

Now, I have only one problem with the Sunrise CF when I put two cards at the same time but the cause is not the driver.

By sd_snatcher

Prophet (2945)

01-03-2019, 22:59

gdx wrote:

Now, I have only one problem with the Sunrise CF when I put two cards at the same time but the cause is not the driver.

This is a hardware problem with some Chinese cards and <insert your preferred card here>-to-CF adapters. They were built under the presumption that they would be only used on cameras and PC USB readers, and those only have one slot.

So, some of them just seem to ignore the card /CSEL pin (used for master/slave selection), and others seem to have the IORDY pin implemented as push-pull outputs, instead of open-collector as they should.

By gdx

Prophet (2612)

02-03-2019, 13:11

It would be nice to make a compatibility list. I never had a problem with CF cards from Nagiwara Sys-Com but cards from Transcend (2GB), Onefavor (32MB) and Sandisk (32/128MB) do not like the slave slot. They work but badly.

By Louthrax

Paragon (2012)

02-03-2019, 13:33

sd_snatcher wrote:

This is a hardware problem with some Chinese cards and <insert your preferred card here>-to-CF adapters. They were built under the presumption that they would be only used on cameras and PC USB readers, and those only have one slot.

Aaaah, lots of PITA finally explained Good to know, thanks Snatcher !

By gdx

Prophet (2612)

03-03-2019, 13:04

Has anyone managed to use two CF cards at the same time with driver v0.17 on an MSX without floppy disk drive? Today I did some tests with my cards, I did not succeed.

By konamiman

Paladin (998)

03-03-2019, 20:59

I'm really puzzled with that bug on diskless machines. I don't own any IDE device and I can't reproduce the bug on emulators, but as fas as I understand, the Carnivore2 is using the same standard Sunrise IDE driver I use on the emulators. Also a bug that appears only when no other disk controllers are present is something that in principle doesn't make any sense.

I really need more information. What's the exact hardware configuration of the Carnivore2? Does it emulate subslots? In this case, what's in these? How does the bug exactly manifest? What happens when there is no card inserted? Is really everything fine when there is other disk controller present? Does changing the order of the slot numbers of the controllers change anything?

By gdx

Prophet (2612)

04-03-2019, 09:54

Sorry I can't pin down the problem precisely. I no longer have a monitor, it broke down. So I test my three MSX (CF-7000, FS-A1 and FS-A1ST) with a down scaler on my PC monitor.

When I insert a single card in the master slot or slave, everything works fine on all MSX. (My micro SD to CF works only in master slot though)

With two CF cards on MSX1/2 (without floppy disk drive) the second card can't mount at all. On the FS-A1ST, some CF cards mount and seem work at first glance but oddities appear more or less quickly depending on the cards used as and when copy and use. The problem of using two cards at the same time seems magnified with the Nextor driver.

I also noticed a little bug that is not related to what I said above. The disk can't mount when NEXTOR.SYS is present on the disk but not the COMMAND2.COM. Normally the Basic should start.

The driver is almost mature. I'm already happy to use it like that.

By konamiman

Paladin (998)

04-03-2019, 10:13

gdx wrote:

When I insert a single card in the master slot or slave, everything works fine on all MSX

Now I'm confused "On all MSX" includes diskless ones?

gdx wrote:

The disk can't mount when NEXTOR.SYS is present on the disk but not the COMMAND2.COM. Normally the Basic should start.

What do you mean by "can't mount"? In this case what you should see is an error message saying that COMMAND2.COM can't be loaded, right?

By gdx

Prophet (2612)

04-03-2019, 10:28

konamiman wrote:

Now I'm confused "On all MSX" includes diskless ones?

Yes all MSX I have.

konamiman wrote:

What do you mean by "can't mount"? In this case what you should see is an error message saying that COMMAND2.COM can't be loaded, right?

The system crashes. No message.

By konamiman

Paladin (998)

04-03-2019, 10:51

Then again, what's exactly what works and what doesn't work on diskless machines, compared to machines with disk?

By gdx

Prophet (2612)

04-03-2019, 11:38

DOS behaves strangely with two CF cards. On diskless machines it's immediately visible since the drive B does not rise with almost all cards, and sometimes even both drives. On my turbo R, many CF cards mount and seem work at first glance but oddities appear more or less quickly.

Share

Related content

Comments (45)