[linux-elitists] MTA roundup

Andrew L33tsmith akohlsmith-le@benshaw.com
Tue Apr 30 17:34:48 PDT 2002


> And reasonably lucky people may have the same experience.  Or they may
> not.  The point is that public combination testing (let alone regression
> testing) is not feasible.

I agree.  I'm not typically known as being reasonably lucky, but that's 
neither here nor there.  :-)

> Somehow, we're miscommunicating, and I'm not sure exactly how.  You do
> know that modern implementations of UFS are called FFS = Fast
> FileSystem, right?  So, quoting:

> He doesn't approve of async, he particularly doesn't approve of async on
> ext2, and he doesn't even approve of async with softupdates.  (His rants
> specifically against ext2 have been on mailing lists.)

Ah...  I'm not on those mailing lists, which is why I had brought that up.  I 
read the page provided and came to the conclusion that he *prefers* that you 
use sync FFS, not that he dispises anything else, including journalled file 
systems.

> I hope this isn't going to be one of those things of violently arguing
> over interpretations of wording.  What Dan said on the matter is visible
> on the URL-referenced Web page, mostly, and I'll be delighted to leave
> it at that.

Oh hell no.  I'm not trying to start anything; I was just confused where you 
drew some of your conclusions from based on the same text I read.  What it 
turns out is that there are some other texts from DJB that I haven't read.

Regards,
Andrew



More information about the linux-elitists mailing list