I do not pay much attention to BOINC and just let it run.
However I recently noticed that I was getting LHC jobs but no credit.
There appear to be two problems -
1. Some tasks ran till 100% complete but did not stop and continued to run and eventually (after several days) I aborted them.
2. Later and unrelated to the above when examining the Event Log I found the following errors.

23/02/2026 00:24:42 | LHC@home | Output file Theory_2922-4876778-624_0_r188348125_result for task Theory_2922-4876778-624_0 absent
23/02/2026 00:24:55 | LHC@home | Output file Theory_2922-4800650-624_0_r2096107958_result for task Theory_2922-4800650-624_0 absent
23/02/2026 00:25:29 | LHC@home | Output file Theory_2922-4766022-622_1_r333088561_result for task Theory_2922-4766022-622_1 absent
23/02/2026 00:28:42 | LHC@home | Output file Theory_2922-4869178-624_0_r416457011_result for task Theory_2922-4869178-624_0 absent

HELP!

Bill McLachlan

Which computer? 10850788 was working but now it says:

boinc-buda-runner is missing Podman
wsl_init(): no usable WSL distro

So it seems like its not setup correctly anymore.

The others didn't do anything?

I am seeing that on several systems the task shows 100% but is continuing to run.

Is the remaining (est) guestimate not calibrating to the system running it?

---

Let's crunch for our future.

This delightful feature is the result of the way the duration estimate does not take notice of a major part of the calculations - the "final integration" step, which can take hundreds of steps is not used (or at least appears not to be used) in the initial duration calculation. This may be because the actual number of integration steps is not known at the outset, but that's only an excuse, it would be perfectly possible to take a sensible estimate (say 1000 of these steps) at the outset then use the time for successive steps to recalibrate the estimated duration.

Another real gotcha is that if for any reason the task is not stopped correctly the whole calculation is restarted from the very first part (this may only apply to tasks running in a vb under Windows). As a result you may have a task that really nearing completion having taken several days to get there starts again, rattles through the first few stages (they normally take a few minutes), only to apparently stall at 100% completion, but still have many hundred integration steps to complete, so taking another few days.....

Thanks, rob that makes some sense.

Still a bit disconcerting. I seem to be getting a bunch of them on my desktop.

I've paused new work for now until they complete or time out.