At 12:24 Uhr -0400 28.04.2002, Bill Bumgarner wrote:
>BTW: If you do a 'fink --yes fetch-missing' and hit ctrl-c in the middle,
> it will abort the current download attempt, but continue the
>fetch-missing. If you hit ctrl-z and kill the fink process, it
>appears that fetch-missing keeps running in an orphaned process.
>Likely as a result of a user level shell not having permissions to
>kill a root child, therefore the child is orphaned??
Never tried that. What I do if I want to kill it completly is to
press Ctrl-C twice, quickly. T'is a feature, not a bug, sometimes a
download is super slow and I just want to go on to the next :-)
>(I bet the christian right wouldn't be too happy about the last
>sentence of that paragraph. Especially if I were to mention that
>all of this should be triggered by the cron daemon.)
LOL, yeah, we are so damn evil ]:-)
^
Devil horns
[...]
>Unfortunately, --yes causes Fink to retry forever on files that
>could not be downloaded because the server is down or the file has
>been renamed/moved. Ahh-- but it gives up after 5 tries and moves
>on to the next file. Cool. Anyway to adjust the retries? I
>would like it to retry once for the give-up/retry questions and have
>it cycle through several mirrors before giving up on those.
Would be nice to have, yeah, esp. the cycling (right now Fink tries
mirrors randomly, which is not always optimal). But I don't want to
just make it cycle, cause that would mean the "first" mirrors in the
list would be used a lot, while the others get fewer traffic. One way
to solve this would be: Take the list of mirrors, shuffle it, and
then walk it in order. This way we ensure all mirrors get equal
trafifc (in theory at least), and still we don't try the same mirror
twice.
The "shuffling" could still be partially sorted: if you are in
sam-BR, it would first sort all mirrors in sam-BR, then all in sam-?,
then other country mirrors, and finally the primary mirrors followed
by the secondary mirrors... does that make sense? :-)
>
>>>*sigh* I wish there were 40 hours in the day!
>>
>>Be careful for what you wish :-)
>
>Clones are likely more feasible anyway. I just don't want to deal
>with myself as a two year old.
LOL.
Max
--
-----------------------------------------------
Max Horn
Software Developer
email: <mailto:[EMAIL PROTECTED]>
phone: (+49) 6151-494890
_______________________________________________
Fink-devel mailing list
[EMAIL PROTECTED]
https://lists.sourceforge.net/lists/listinfo/fink-devel
