Re: Linux community software-update-anarchy polemic

From: Anonymous Coward (acoward_at_mail.ru)
Date: 01/30/04 

Date: 30 Jan 2004 12:23:27 -0800

"Shmuel (Seymour J.) Metz" <spamtrap@library.lspace.org.invalid> wrote in message news:<400f1e92$2$fuzhry+tra$mr2ice@news.patriot.net>...
> In <ab356123.0401210428.779aef7e@posting.google.com>, on 01/21/2004
> at 04:28 AM, acoward@mail.ru (Anonymous Coward) said:
>
> >the (again, practical) solution is simply to put a battery in the
> >computer and obviate the need for journaling and syncing (at least in
> >the case that the disk is local, with the RAM<->disk connection being
> >axiomatically reliable).
>
> No, that is not the solution
Why?

> and, no, it is not axiomatically
> reliable.
The statement "the RAM<->disk connection is axiomatically reliable"
isn't an assertion of a law of the universe. It's the assertion of an
axiom. In any case, I didn't even make that statement; I said "...in
the case that the disk is local, with the RAM<->disk connection being
axiomatically reliable." Axioms are statements presumed for the sake
of argument to be true. In this case, I was making clear my definition
of "local" as meaning that the connection is presumed reliable,
because if the connection is not presumed reliable, then merely
providing a battery for the RAM no longer obviates the need for
journaling and syncing (unless the persistence of the battery-backed
RAM is considered an acceptable permanent substitute for the
persistence of the disk, but I'm not considering that case here, only
the case that it's considered an acceptable temporary substitute with
just sufficient permanence to enable dumping the RAM to disk in the
event of a primary power loss). The connection here refers to the data
path all the way to the disk's surface. Other local things in computer
systems are commonly presumed reliable (that is, their reliability is
axiomatic): for example, the cpu<->localmemory connection; in
contemporary systems, the possibility of the severing of that
connection is not a consideration in software design. You can assert
different axioms if you want, but asserting that the RAM<->disk
connection is reliable isn't unreasonable, especially in systems that
already assert the reliability of the cpu<->RAM connection, and even
moreso in systems that assert the reliability of the RAM itself
without any ECC or even parity checking.

> Power failures are not the only source of outages, and a
> battery will not obviate the need for recovery techniques.
Please expand on this so that I can reply precisely.

> >Ok ok I'm just a monkey,
>
> Then why should anybody trust your advice on technical matters?
You should not trust me or my advice on technical matters or any other
matters, but the reason that you should not trust me is not that I'm a
monkey, but that I'm posting anonymously. But I never said that
anybody should trust me or my advice; indeed, an anonymous person
saying "trust me!" is absurd. I posted information to be accepted or
rejected on the basis of its own merits, not to be accepted or
rejected because of any identity or qualifications of the source. See
my comments about random number generators in my thread with Peter
Breuer.

Relevant Pages

  • Re: Question Magic Jack
    ... same connection. ... We have tried VOIP (Voice over Internet Protocol) telephone systems. ... And found that voice quality and reliability is highly variable even ...
    (alt.home.repair)
  • RE: confused
    ... reliability issue by having a backup mail server and adding a second MX ... because you have a concern about internet connection reliability too. ...
    (microsoft.public.windows.server.sbs)
  • Re: Cox Communications Usenet service entering 10-3-39
    ... alternatives for extra money. ... But the reliability gets worse each time another wave of new people ... I'm having connection problems almost daily, ... As an example they have a Mini account that's $12 for 6-months or $19.95 for ...
    (rec.sport.football.college)
  • Re: SNR margin & router questions
    ... connection was more reliable at the cost of sync speed with the latter. ... increase in speed along with rock solid reliability. ... that I expect my BRAS rate to imminently fall from my current 2.5 Mbps ... lower speed when the noise increases at the start of opening hours. ...
    (uk.telecom.broadband)
  • Re: SNR margin & router questions
    ... connection was more reliable at the cost of sync speed with the latter. ... series of Openreach engineers found various minor faults but no ... increase in speed along with rock solid reliability. ... that I expect my BRAS rate to imminently fall from my current 2.5 Mbps ...
    (uk.telecom.broadband)