[Coco] OS9Boot bugs

[Gene Heskett]

Robert Gault wrote:
> Chester Patterson wrote:
>> Hi all,
>>
>> OS9 ver2 standard. Coco3 512, MPI. All worked great 12-14 years ago. But
>> now... While making boot disk for my ST225 (B&B interface) I have run 
>> into
>> some curious behaviour that I haven't seen before. First a standard 
>> bootup,
>> load bbfhdisk and /H0. Was able to format and dircopy to /H0. And 
>> chd/chx as
>> well. However when I went to use any command that access' /h0/cmds, error
>> 215. OK, I thought, boot order bug. So ezgen a proper boot disk with
>> bbfhdisk and H0 right after RBF and before cc3disk. Boot up. Am able to
>> chd/chx and access /h0 just fine but none of the /w descriptors work! 
>> (error
>> 215) All I have is the 32 column term. So I turn everything off, 
>> disconnect
>> the power to the ST225, but not the cables, boot up, and presto, all the
>> windows work fine. How do I get this to work? Thanks.
>>
>> /Chester
>> Costa Rica
>>
>>
>>
> 
> Please clear up several points of information. How long have you had the 
>  B&B interface and hard disk? Has the hard disk system ever worked? With 
> the standard boot, was bbfhdisk and h0 in a single merged file? If not, 
> you usually can't load them separately and get things to work correctly. 
> What happens if you power the hard drive but remove the cable from the 
> B&B interface? Does everything other than the hard drive work correctly?
> 
> 
>  First, the "boot order bug" should not be an issue here. It really 
> applies to the floppy controller not hard drive systems.

I would take exception to this as a blanket, floppy only problem 
generalized statement.

The blob as it came to be called, was actually a gime init problem, and 
as such it may have had something to do with the penchant my B&B XT-RTC 
had to return an empty, all zeroed out, 256 byte sector to rbf.mn when a 
search of the fat was being done.  As you can imagine this lead to my 
being intimately familiar with the os9 file system as I was always 
rebuilding what was left with ded.  I eventually switched to a scsi 
drive and interface using a disto 4n1 and never had that problem again. 
  The blob was fixed not too long after that, maybe a year max.  But 
I'll always wonder if that was how it would manifest itself if it took a 
notion to hit someplace besides the floppy.

The blob fixes are in most of the later clock modules.


>  At first glance, you seem to have a power supply or memory problem. 
> However, it would help to see an exact copy of the Ident results on your 
> two os9boot files. There may be something not quite right with them. 
> Make sure that everything in the os9boot has good CRC values.
>  MPI units can have power supply problems. My own had one that was fixed 
> by reseating the lock washer/bolts on the main power transistor. The 
> contacts between the MPI and Coco or the controller/interface and MPI 
> are known to cause problems; clean them.
> 
>  It actually reads as if you have two problems. Since turning on/off 
> power to the hard drive effects the success of the boot, that suggest a 
> power supply problem. Rearranging modules in the os9boot file and 
> subsequent problems, seems to suggest a bad os9boot component.
>  If you repeat your tests several times making them exactly the same 
> test, do you get the same results or are things intermittent?
> 


-- 
Cheers, Gene