Re: Systems lie about persistence and consistency promises. A lot.

Show/hide message thread

scsh-0.6.7 -> scsh-0.7 has substantial changes, and introduces grievous error hga@xxxxxx (10 Jul 2019 18:30 UTC)

Re: scsh-0.6.7 -> scsh-0.7 has substantial changes, and introduces grievous error Lassi Kortela (10 Jul 2019 18:51 UTC)

Re: scsh-0.6.7 -> scsh-0.7 has substantial changes, and introduces grievous error hga@xxxxxx (10 Jul 2019 19:43 UTC)

Re: scsh-0.6.7 -> scsh-0.7 has substantial changes, and introduces grievous error John Cowan (10 Jul 2019 19:56 UTC)

Re: scsh-0.6.7 -> scsh-0.7 has substantial changes, and introduces grievous error hga@xxxxxx (10 Jul 2019 20:05 UTC)

Re: scsh-0.6.7 -> scsh-0.7 has substantial changes, and introduces grievous error Lassi Kortela (10 Jul 2019 20:32 UTC)

Systems lie about persistence and consistency promises. A lot. hga@xxxxxx (10 Jul 2019 22:10 UTC)

Re: Systems lie about persistence and consistency promises. A lot. John Cowan (10 Jul 2019 22:20 UTC)

SQLite only SRFI? hga@xxxxxx (11 Jul 2019 01:49 UTC)

Re: SQLite only SRFI? John Cowan (11 Jul 2019 03:23 UTC)

Re: SQLite only SRFI? hga@xxxxxx (11 Jul 2019 08:19 UTC)

Re: SQLite only SRFI? John Cowan (11 Jul 2019 13:32 UTC)

Re: Systems lie about persistence and consistency promises. A lot. Lassi Kortela (10 Jul 2019 22:26 UTC)

Re: Systems lie about persistence and consistency promises. A lot. Lassi Kortela (10 Jul 2019 22:43 UTC)

Re: Systems lie about persistence and consistency promises. A lot. John Cowan (10 Jul 2019 22:50 UTC)

Re: Systems lie about persistence and consistency promises. A lot. hga@xxxxxx (10 Jul 2019 23:02 UTC)

Re: Systems lie about persistence and consistency promises. A lot. Arthur A. Gleckler (10 Jul 2019 22:51 UTC)

Re: Systems lie about persistence and consistency promises. A lot. Lassi Kortela (10 Jul 2019 22:59 UTC)

Re: Systems lie about persistence and consistency promises. A lot. hga@xxxxxx (10 Jul 2019 23:19 UTC)

Schemepersist, anyone? hga@xxxxxx (10 Jul 2019 23:40 UTC)

Re: Systems lie about persistence and consistency promises. A lot. Lassi Kortela 10 Jul 2019 22:59 UTC

> It's not clear to me that the 30-second auto-sync is reliable, either.
> I can't find it right now, but I read an ACM paper a few years ago that
> compared several filesystems and showed that they had widely varying
> behavior with respect to unplanned shutdown.  Varying the hardware makes
> reliable behavior even less likely.  I don't think waiting thirty
> seconds will fix that.

It's most likely not reliable. I find your argument easy to believe,
even without any sources :)

I wonder how many programs dutifully use sync() but forget to flush
their own output buffers.