Happy Hackin’ days…
Check out what my BeOS mail_daemon does!
I also applied a ‘fix’ (haha more like a kludge) for people who, like me, formatted their BFS drives with a block size of 1024. In case you hadn’t noticed, R5 (no idea if this persists in dano / Zeta) will occasionally loose updates to the queries, resulting in incorrect information being displayed in the deskbar. It gets -really- annoying. Most common is the case where the icon says I’ve got mail, but I really don’t. The only cure was to manually restart the mail_daemon. Until now.
Mine will automatically rebuild the query (full re-issue) every 20 seconds if there’s a BFS volume with a block_size of 1024 attached to the system. So even if the count does get off by a bit, in twenty seconds, it’s accurate again!
Now to finish up the log handling stuff I was working on…
December 30th, 2005 at 11:17 pm
Bad Bryan ! Bad Kludge ! Go sit in the corner !
December 31st, 2005 at 1:38 pm
Controversial and NEW http://www.quickfox.illhostit.com/bq2/quake3_beos.html
January 1st, 2006 at 4:44 pm
AFAICT it still happens with BFS volumes with a block size of 2048 bytes, although it happens less often. I have been told that this bug of BFS has been fixed in Zeta, but I didn’t check it :-)
January 1st, 2006 at 8:54 pm
Hi,
I posted this on your previous post, but for some reason it did not show up.
Since you are working on MDR, would it be possible to make BeMail use the same encoding as the original message when you reply or forrward a message?
Thanks!
Koki
January 2nd, 2006 at 12:34 am
Was this fixed in OpenBeFS?
January 2nd, 2006 at 8:03 am
> Was this fixed in OpenBeFS?
Sure, Haiku’s BFS won’t have any of these problems. Why should we try to create the same bugs?? :-)
January 2nd, 2006 at 10:14 am
Axeld Aahhhhhh NathanW told me it didn’t exist on 2048 partitions a while back. I’m glad to hear that it does, which means I’ve not been missing anything by not reformatting my disc.
I’ll make sure I remove the 1024 check before I submit a patch, so it’ll work on all the BFS discs.
Which brings me to another question. Is there a way to tell in the fs_stat structure which driver (bfs or openbfs) is controlling a volume? If so, that would be nice to have. :-)