[svn.haxx.se] · SVN Dev · SVN Users · SVN Org · TSVN Dev · TSVN Users · Subclipse Dev · Subclipse Users · this month's index

Re: Implement major FSFS performance related changes in the experimental FSX format

From: Stefan Sperling <stsp_at_elego.de>
Date: Tue, 30 Sep 2014 19:30:06 +0200

On Tue, Sep 30, 2014 at 05:45:48PM +0100, Julian Foad wrote:
>   ***
>   Who has reviewed this or intends to do so? And for what aspects -- overall
>   design, compatibility/upgrade issues, threading issues, style, ...?
>   ***
 
I am not planning to review the FSFSv7 changes because the changes
are too large and complex for me to digest in a reasonable amount
of time. From my point of view they have accumulated to a giant code
dump developed by a single developer (much like the original FSFS code,
as Greg once told me in Berlin). They are the result of much of Stefan's
time and effort, which is awesome, but which he also developed mostly
alone, which is bad for collaboration if it happens over such a long
period of time. I wish Ivan and Stefan had developed these changes
together, in lock step, from the beginning. I hope we can learn from
this in the future and adapt our way of working accordingly, somehow.

I hope to have made a small but valuable contribution towards FSFSv7's
quality by making sure there is no 1980's atoi()-style code in it that
never heard of integer overflows. But beyond that I see no way of
catching up. I came to this conclusion after glancing over the files
that Ivan pointed out earlier we should read.

I'd rather invest my time elsewhere and hope the community will settle
on something that is good enough. Given that this discussion is still
on-going and obviously considered important, I still have hope that
this will be the case.

And I hope I won't be left alone when having to fix FSFSv7 corruption
should it ever occur in the wild ;-)
Received on 2014-09-30 19:30:39 CEST

This is an archived mail posted to the Subversion Dev mailing list.