is this a bug, or not? 

> inc
Incorporating new mail into inbox...

2233+ 12/07 sssandy@mit.edu    subdirector<<want to be a really cool powerful s
2234  12/07 greg@duke.cs.unlv  Zephyr 2.0.2 checksum byte order wierdness<<Ther
2235  12/07 rivest@theory.lcs  Friday's meeting // abstract from Goldreich // Y
2236  12/07 ken@analytix.com   Charts for December HIS Newsletter<<--==========
2237  12/07 foner@media.mit.e  (Boston-area) Anyone tape DS9 this weekend?<<I w
inc: unable to write 2238: No space left on device
> inc
Incorporating new mail into inbox...

2238+ 12/07 sssandy@mit.edu    subdirector<<want to be a really cool powerful s
2239  12/07 greg@duke.cs.unlv  Zephyr 2.0.2 checksum byte order wierdness<<Ther
2240  12/07 rivest@theory.lcs  Friday's meeting // abstract from Goldreich // Y
2241  12/07 ken@analytix.com   Charts for December HIS Newsletter<<--==========
afs: failed to store file (over quota or partition full)
inc: write error on 2241: No space left on device
> scan last:10
2227  12/07 rei@mit.edu        [John_Wang@mktplace.com: IAP job opportunities f
2232  12/07 jhbrown@ai.mit.ed  [sajit@ai.mit.edu: forwarded message from jeff@a
2233  12/07 sssandy@mit.edu    subdirector<<want to be a really cool powerful s
2234  12/07 greg@duke.cs.unlv  Zephyr 2.0.2 checksum byte order wierdness<<Ther
2235  12/07 rivest@theory.lcs  Friday's meeting // abstract from Goldreich // Y
2236  12/07 ken@analytix.com   Charts for December HIS Newsletter<<--==========
2237+ 12/07 foner@media.mit.e  (Boston-area) Anyone tape DS9 this weekend?<<I w
2238  12/07 sssandy@mit.edu    subdirector<<want to be a really cool powerful s
2239  12/07 greg@duke.cs.unlv  Zephyr 2.0.2 checksum byte order wierdness<<Ther
2240  12/07 rivest@theory.lcs  Friday's meeting // abstract from Goldreich // Y

> mkdir /var/tmp/mail
> inc +/var/tmp/mail 
Incorporating new mail into /var/tmp/mail...

   1+ 12/07 ken@analytix.com   Charts for December HIS Newsletter<<--==========
   2  12/07 foner@media.mit.e  (Boston-area) Anyone tape DS9 this weekend?<<I w
   3  12/07 jweiss@mit.edu     banana's mouse<<Banana-sr's mouse seems to have 
   4  12/07 jweiss@mit.edu     banana's mouse found<<So as soon as I send the m

the first time, I run over quota.  I delete some stuff, inc again, run out of
quota again.  so, I decide to inc into temp space.  this time, it doesn't
inc from the begining, it inc's from where it left off last time.  What
would cause this inconsistent behavior?