Flash disks and FFS layout heuristics

Matthew Dillon dillon at apollo.backplane.com
Tue Apr 1 17:47:58 PDT 2008


:My, my. Mr Dillon likes to be rude to people and tell them they are =
:'stupid' and 'silly', but when he makes na=EFve comments about systems =
:he doesn't understand and gets called on it, suddenly it's "complete =
:bullshit."
:
:I can see why PHK broke off trying to educate you.

    I really have no love for people who are so disrespectful to their
    peers.  A few of the people unfortunately associated with a project I
    had an interest in fit that category, some more then others. 
    Not too many, only two (well, three if you count yourself).  On the
    bright side my list is very limited.   

    I do not believe that you are any more qualified then you think I am.
    Clearly it is an issue for you and just as clearly you are unwilling
    to engage in any sort of technical conversation about the matter.  I
    really have no idea why.  If you decide you want to have a technical
    conversation, where you actually post meaningful information useful not
    only to me but to everyone reading this thread, instead of vague, broad,
    uninteresting references, then please go ahead and do so.  If you
    think those vague bits of information you post, condescending and
    secretive as if they were something so secret and special nobody needs
    to know the details... if you think those actually contribute to the
    conversation, then you are deluded.

    If it is important to you, then perhaps you should consider that the
    characteristics of NAND flash are only a small part of the equation.
    The characteristics are not this mystical scary beast that nobody
    understands, they are very well defined and fairly limited in scope,
    and thus can be discussed, theorized, implemented, and tested.  None
    of these processes are absolute.  Hell, filesystem design is just as
    important and I dare say that the only person on this list with more
    experience then I have on filesystem design is Kirk.

    I'm a technical theorist, a dreamer, and an implementer.  Theory
    always comes before function, always.  I don't know what your problem
    is and I really don't care, but it absolutely does not and never has
    required direct experience to have a technical conversation.  If that
    were true nobody would ever invent anything, try anything, or make
    any progress.

    So, yes, there is a great deal of value to having a technical
    conversation that mixes theory and actual direct experience.  Very
    few people have the breadth of direct experience required to be able
    to comment definitively on something.  Not a single person on this
    list, not myself, not you, not Poul... nobody has anywhere near the
    level of experience required to come to any sort of conclusion with
    regards to the material we are discussing.  All we can do is experiment,
    theorize, and have a technical conversation about the merits of one
    thing or another.

    So, again, if you have something to contribute to our technical
    conversation, perhaps some direct experience you've had trying to
    actually implement one of these 'failed' schemes???, I'm all ears.
    If not, then I recommend you stop posting.

						-Matt



More information about the freebsd-arch mailing list