The Os 3.1.4 Thread

[cehofer]

Thomas,

I am sorry I started a fight here.  First off, I want to say I for one greatly appreciate your help and knowledge of the Amiga OS in general.  Thank you for that!  I was hoping you were going to say to just add some "tooltypes" to the printer prefs and I would be good to go.  I have been busy so I haven't had a chance to try restoring 3.9 printer prefs.  It doesn't sound too difficult.

Kolla is correct that netprinter is a network printer redirection program on aminet.  I will have to restore 3.9 printer prefs and see if I keep getting that crash when I try to use or save netprinter prefs. 

I am currently using netpar, also on Aminet, and it works nice with my new HP CLJ Pro MFPM281 printer.  I have printed from Pagestream4 and Finalwriter97 in color.  What seems to work best in Finalwriter is using the Postscript driver.  It is nice that HP still makes printers on PCL5, 5e and 5c along with Postscript even though that is emulated but it still works.  The only beef I have with NetPar is that it has a banner page and wastes a page of paper every print.  I cannot figure out how to shut it off in the docs.  It is annoying!  So if somebody knows hot to stop the Banner print page, I would highly appreciate it!

[guest11527]

I am sorry I started a fight here. 

No worries, you haven't. Just be aware that I am not the Hyperion support. If you have issues, I need all your help if you want me to hunt them down.

Concerning your particular problem, as stated, increase the stack size and all will be fine. Or use the 3.9 prefs if that suits you better.

The only beef I have with NetPar is that it has a banner page and wastes a page of paper every print.  I cannot figure out how to shut it off in the docs.  It is annoying!  So if somebody knows hot to stop the Banner print page, I would highly appreciate it!

Not one of my programs, nor part of 3.1.4, so I'm out here.

[guest11527]

Unfortunately, it is up to me to report a bug myself and inform the users. As it seems, a bug present in the 3.9 version of the scsi.device made it into the 3.1.4 release without being identified in the code review.

The problem is that the scsi.device erraneously reports a problem if a block transfer of more than 255 blocks crosses a 4GB boundary. The transfer would have worked, data integrity is not harmed, but the command is aborted.

One can work around the problem by setting (*sigh* again) MaxTransfer to 1FE00, limiting block transfers to 255 blocks, as we had it before. So, we have - unfortunately yet again - revert to the old MaxTransfer workaround to cure glitches in the drivers, but this is the only practicle way how to get rid of the issue.

Thanks to Toni Willen for identifying and reporting the problem.

[kolla]

Well, duh. I noticed that I needed max transfer settings pretty much right away half a year ago, but you insisted on calling me a liar and conspiracy theorist, and whatnot when I mentioned the problem.

[Gulliver]

It is all about me! 

[chris]

I am currently using netpar, also on Aminet, and it works nice with my new HP CLJ Pro MFPM281 printer.  I have printed from Pagestream4 and Finalwriter97 in color.  What seems to work best in Finalwriter is using the Postscript driver.  It is nice that HP still makes printers on PCL5, 5e and 5c along with Postscript even though that is emulated but it still works.  The only beef I have with NetPar is that it has a banner page and wastes a page of paper every print.  I cannot figure out how to shut it off in the docs.  It is annoying!  So if somebody knows hot to stop the Banner print page, I would highly appreciate it!

Maybe try Olaf's lpr.device instead.  I'd hope that is fully compatible given the author's involvement in OS3.1.4 (also I'm pretty sure it doesn't try to do anything fancy like banner pages, although it has been some years since I last used it)

[cehofer]

I installed the 3.9 printer prefs, printer gfx and printer PS along with the env-archive/sys printer.prefs.  When I tried to set up netprinter.prefs, it still crashed on use or save.

Chris, I will have to try lpr.device.

thanks

[cehofer]

Thomas,

Would this SCSI.device bug cause my harddrive light to come fully on and computer to crash.  I have had this issue with my 2 A4000Ts but not my 4000D with oktagon scsi.  What will happen without rhyme or reason, I have tried different size drives and different size partitions but on boot sometimes or sometimes drive access the HD light will go on full bright and computer will crash.   Since I have re-formatted my partitions without "international mode" it has occurred much less.  This has happened on 3.9 and 3.1.4  I am using the current 46.13 FFS.  On 3.9 I was using 45.13 the "unofficial ffs".  I have installed the new 46.13ffs on my 3.9 now.

I have the SD2SCSI V6 and  I am able to capture scsi traffic.   I contacted Michael McMaster, developer of SD2SCSI.  He analyzed the logs and told me that it was occurring during sync transfers.

"Please use scsi2sd-util to set the "speed" limit option to Async 5MB/s. I can see that the logs all show synchronous data transfers, and the hang occurred during a sync transfer. It would be good to know if the problems go away in async mode."

I could get this to occur on a regular basis using Samba swat web page on the A4000T.  I just go to the browse to the swat page and browse around on help, config, etc and the drive light would go fully on and A4000T would crash.   I did as Michael suggest and it did stop it for awhile but would still occur from time to time.  What has made my A4000T the most stable is formatting the drives without "international mode".  I have since put the SD2SCSI to full speed again and  I really don't see any difference.

My SCSI is configuration is I have a physical SCSI II 4gig HD partitioned :

1gig System
1.4 gig Work
1.6 gig Data.

I have several drives on the SD2SCSI 32Gig SD card divided into 2 SCSI IDs 4 and ID5 16G each:
900MB
1.8G
3.9
3.8
5.6

3.5
3.8
3.8
1.8

The controller is configured for Sync transfers and SCSI II.  The Termination is the SD2SCSI active termination built in it.

[guest11527]

, I have tried different size drives and different size partitions but on boot sometimes or sometimes drive access the HD light will go on full bright and computer will crash. 

No. This has absolutely nothing to do with the issue. The issue, to repeat this again, only affects transfers over 255 sectors that *also* cross a 4GB boundary. Hence, they are pretty rare. Even then, the system would not react with a deadlock, but with an error condition. What you see here is likely a hardware issue. I would also suggest to check the power supply.

I have the SD2SCSI V6 and  I am able to capture scsi traffic.   I contacted Michael McMaster, developer of SD2SCSI.  He analyzed the logs and told me that it was occurring during sync transfers.

The issue is also *only* in the IDE version of the (not-so) scsi.device. "Sync" transfers - and the SCSI2SD - are SCSI devices, not IDE devices.  Termination is one issue, if the host cannot catch up with the transfer speed of the SCSI2SD, then this is another issue.

[kirk_m]

Has anyone encountered any issues with SCSI drives no longer working after installing 3.1.4 on an A2000 that has a Blizzard 2060 accelerator card?  I swapped out my 3.1 ROMs for 3.1.4 yesterday, and, now the SCSI drives do not show up in the early startup, or, anywhere else for that matter.  The IDE drives are still visible, as I was able to install 3.1.4 onto a drive connected to the Buddha IDE controller.

If I reinsert the 3.1 ROMs, my SCSI devices (SCSI2SD and CDROM) are again visible and usable (indeed, the SCSI2SD is my 3.9 boot drive on this computer).

[guest11527]

Has anyone encountered any issues with SCSI drives no longer working after installing 3.1.4 on an A2000 that has a Blizzard 2060 accelerator card?

Nope. Given that I own exactly this combo, and I develop (amongst others) on this combo - no, that should be (and provably, actually is) working quite fine.

I swapped out my 3.1 ROMs for 3.1.4 yesterday, and, now the SCSI drives do not show up in the early startup, or, anywhere else for that matter.  The IDE drives are still visible, as I was able to install 3.1.4 onto a drive connected to the Buddha IDE controller.

That sounds more like a hardware issue to me. Oh, yes, a SCSI2SD here, too.

[giZmo350]

Has anyone encountered any issues with SCSI drives no longer working after installing 3.1.4 on an A2000 that has a Blizzard 2060 accelerator card?

I have the same setup, including a SCSI2SD V5 (SCSI2SD not in my sig A2000 though) and would try it out but alas, I do not yet own OS3.1.4.

Kirk, I thought you owned a SCSI card reader, i.e., a PCD-50B? If you do, you should use it with your Blizz 2060. It's ideal! 

[kirk_m]

Thomas,

The HD activity light stays illuminated on the 2000, as well as the LED on the SCSI2SD when the 3.1.4 ROM is installed.  Also, 2060scsi.device shows up nowhere in the list of devices that are currently active on the system.  If I plug the 3.1 ROM back, all of this comes back, and, the LEDs behave as they should.   Also, reinstalling the 3.1 ROM makes everything work again -- all drives appear, I can boot from the SCSI drive, etc.  As soon as I put the 3.1.4 back, it disappears.

Does this steadily illuminated LED indicate anything?

Also what version firmware is on your 2060's ROM?  Sysinfo reports v8.2 on mine.  It seems that 8.5 is the last version made (at least according to the archive on A1K).  Does this make any difference?


Gizmo:

No, I don't have one of them there fancy card-readers

[Rotzloeffel]

Has anyone encountered any issues with SCSI drives no longer working after installing 3.1.4 on an A2000 that has a Blizzard 2060 accelerator card?

Same setup here! No problems and I am using a scsi2SD V5.....

[kirk_m]

Has anyone encountered any issues with SCSI drives no longer working after installing 3.1.4 on an A2000 that has a Blizzard 2060 accelerator card?

Same setup here! No problems and I am using a scsi2SD V5.....

Funny you should reply... I got my 2060 from you

[Rotzloeffel]

Funny you should reply... I got my 2060 from you

haha.... OK Never sell things, you not have twice

did you set the Jumper which copies the ROM into the FastRAM? Maybe there is a Problem? I did not set this Jumper!

I use MUFastROM instead....