// DBOINCP-300: added node comment count condition in order to get Preview working ?>
Jos van Wolput
Joined: 11 Feb 05
Posts: 47
Credit: 800840
RAC: 0
25 Jan 2007 3:36:20 UTC
Topic 192356
(moderation:
)
When an unfinished EAH task is suspended with boinc_5.8.4_i686-pc-linux it jumps to 100%, marks the task as complete and uploads it resulting in a validation error.
For the sake of not confusing those casual and/or inexperienced readers of individual project message boards, please realise that discussion of the problems of various alpha test versions of BOINC is probably best done on the appropriate BOINC message board rather than a project message board. Both the people who need to know and the people able to help are more likely to be there rather than here, even if the problem was noticed while a particular project's science app was running.
Obviously, the development of BOINC needs a group of dedicated alpha testers and a decision to be one of those is highly commendable. However it's not a job for everybody and particularly not for someone who has recently taken up a BOINC based project for the first time. The recommended 5.4.11 version is what such a person should be using. Don't be tempted to go for something like 5.8.5 because it looks "brand new" and therefore must be "better". A particularly bad time for an inexperienced person to trial a "new" version like this is when the minor version number is changing pretty rapidly. That's a sure sign that the "bleeding edge" is bleeding rather badly and that extra band-aids are being applied rather rapidly :). Please note that we have seen a very rapid progression from 5.8.0 -> 5.8.5 with no sign yet that things are stabilising.
For the sake of not confusing those casual and/or inexperienced readers of individual project message boards, please realise that discussion of the problems of various alpha test versions of BOINC is probably best done on the appropriate BOINC message board rather than a project message board. Both the people who need to know and the people able to help are more likely to be there rather than here, even if the problem was noticed while a particular project's science app was running.
Actually, it's even better to report problems like this directly to the BOINC Alpha email list, as that's where the developers read first.
Suspend problem with boinc_5.8.4 (testing)
)
I should have written
...suspended by clicking the pause button...
5.8.5 Beta release is out.
)
5.8.5 Beta release is out. Maybe a known bug in .4?
For the sake of not confusing
)
For the sake of not confusing those casual and/or inexperienced readers of individual project message boards, please realise that discussion of the problems of various alpha test versions of BOINC is probably best done on the appropriate BOINC message board rather than a project message board. Both the people who need to know and the people able to help are more likely to be there rather than here, even if the problem was noticed while a particular project's science app was running.
Obviously, the development of BOINC needs a group of dedicated alpha testers and a decision to be one of those is highly commendable. However it's not a job for everybody and particularly not for someone who has recently taken up a BOINC based project for the first time. The recommended 5.4.11 version is what such a person should be using. Don't be tempted to go for something like 5.8.5 because it looks "brand new" and therefore must be "better". A particularly bad time for an inexperienced person to trial a "new" version like this is when the minor version number is changing pretty rapidly. That's a sure sign that the "bleeding edge" is bleeding rather badly and that extra band-aids are being applied rather rapidly :). Please note that we have seen a very rapid progression from 5.8.0 -> 5.8.5 with no sign yet that things are stabilising.
Cheers,
Gary.
RE: For the sake of not
)
Actually, it's even better to report problems like this directly to the BOINC Alpha email list, as that's where the developers read first.