[Info-vax] HP wins Oracle Itanium case

[Johnny Billquist]

On 2012-08-23 18:22, ChrisQ wrote:
> On 08/23/12 01:15, Johnny Billquist wrote:
>> Except you will get a block, and not a stream of bytes, when you read
>> from a disk. If your "stream of bytes" supposedly were just 10 bytes,
>> you would still get one block of bytes from the disk. The other 502
>> bytes will also be something. You will then have to write a layer on top
>> of the block nature of a disk to try and hide this truth.
>>
>
> Sigh, your still missing the point. If I do a file open and read from
> unix, I get the whole file as a stream of bytes loaded into memory. The
> underlying disk format is irrelevant. All that stuff is handled by the
> files system and the underlying block size may be 512 or any other
> convenient
> value. If the file size is not an integral block count in length, then the
> last block is truncated before being returned to me. Thus, from a system
> programming pov, the written or read data is a lowest common denominator
> byte stream onto which you can impose a structured format.
>
> No one thinks about disk blocks other than driver writers and yes, I do
> that
> sort of work from time to time as well...

So basically, what you are saying is that when using the file system 
interfaces, a file (on disk or anywhere else) is a stream of bytes (in 
Unix).
I have no problem with that. It's just not the "lowest level of a disk", 
nor can all I/O be done through the file system.

	Johnny