Re: Writing an Operating System in Ada

["Dmitry A. Kazakov" <mailbox@dmitry-kazakov.de>]

On Sat, 16 Jan 2010 13:18:46 -0800 (PST), Maciej Sobczak wrote:

> On 15 Sty, 22:48, "Dmitry A. Kazakov" <mail...@dmitry-kazakov.de>
> wrote:
> 
>>> The blob becomes useful only when it is loaded into memory.
>>
>> How does it?
> 
> By becoming a typed entity that is referenced within the actively
> executed program. This is the moment when blob becomes useful.

So the file containing a movie is useless?

> Several posts ago you have requested to execute actions on objects.
> "Play" was an action on the movie - you wanted that. When I said that
> the blob (a file, essentially) becomes useful when it is loaded into
> memory, I meant that it is the interpretation of the blob that gives
> it some requested capabilities.

Is there any player that loads movies into the memory? That must be a very
curious one. I also assume that data bases are all totally useless if their
contents do not fit into the memory? There still exist text editors capable
to handle files bigger than the virtual memory.

Note that your argument its bogus, because being logically continued, no
object is useful if not in the CPU cache. But, wait, those are useless too,
because not in the registers. Wait, no, they must be on the wire in the CPU
to be useful.

>> In any case,
>> "understanding" is not the purpose of the file system.
> 
> Well, I use it for this purpose.

Others uses microscopes to crack nuts...

>> There are certain functionality a persistent store must provide. Among them
>> enumeration, efficient indexing, naming, notifications, journaling,
>> identification, distribution, authentication, consistency and so on. Modern
>> file systems is a persistent store that fulfills some of these
>> requirements. "Understanding" is not in this list.
> 
> Because you did not put it on that list. I'm not going to discuss it
> this way.
>
> http://en.wikipedia.org/wiki/File_System
> 
> "file system is a method of storing and organizing computer files and
> the data they contain to make it easy to find and access them"

System is a method? (:-)) That sort of definition does not deserve to be
discussed.

>>> And the fact that my USB stick works everywhere shows that this
>>> assumption is realistic. The assumption that the target OS is pure-OO
>>> would not be.
>>
>> Assumption? It is not an assumption, it is a fact, that somebody sat down
>> and wrote the implementation of FAT for the OS X1. Why he or someone else
>> could not do this for an OO X2 persistence layer?
> 
> The only way to implement any persistence layer across different OSs
> is to have it not coupled to any OS.

I cannot comment on that, because I don't know the meaning of the word
"coupled" you are using here.

> If you have pure-OO OS (that is,
> when *everything* in this system is OO) then it's persistence layer
> cannot be implemented on other, non-OO OSs.

That is a logical fallacy:

   X has no A AND X has A => X is not X

X=OS, A=OO. Yes, if the OS is defined as not supporting persistent objects,
then no OS can support them.

> Current file systems (the ones with blobs) serve well as "common
> denominators" and this is exactly their value.

And the unicycle is the common denominator of all vehicles, therefore cars
with four wheels are useless...

>>> Are you sure you are still living in a world where
>>> "commercial" (whatever that means) is equivalent to "leading"?
>>
>> It is an equivalent to "measurable". You can measure involvement in serious
>> projects by investments in.
> 
> That's one approach. Another is to measure involvement by
> participation. In the age of social-this, social-that and social-
> whatever, the accounting is different.

I must disappoint you. That measurement was used first in the age of
slavery, when the time assigned to the activity (then performed by slaves)
was the measure. This measure is quite poor because workers are different
and activities are too. This is why slavery, corv�e, socialized economies
gave way to the free market ones. One can use that "different" accounting
to fool others, but better not oneself.

-- 
Regards,
Dmitry A. Kazakov
http://www.dmitry-kazakov.de