Robert, keep in mind that BOINC has two main directories now. One for the BOINC Manager, and one for the "data directory" where the projects and slots reside (and they could be the same).

I believe the files they are looking for are located in the data directory. The one just OVER the projects and slots subdirectories. And they start with "std".

You could EMail me the files and I could try to forward for you if the EMail address still isn't working for you.

My current problem is that I'm having trouble finding the BOINC data directory, since for this combination of BOINC version and Windows version the frequently written files are in the data directory tree instead of the program-oriented BOINC directory. The Vista SP2 update seems to interfere with using the search function to find a directory if you know the lowest level name in the directory path, but not much more about it. Since the "std" files aren't directories, I'll try searching for them instead and see if that gets around this problem.

In case I need to search for a complete filename instead, what is the current full name of the files minirosetta uses as lockfiles?

[quote]Under that BOINC, where is the slots subdirectory? A files search was unable to find it.]

In my case the slots directory is at:

D:Documents and SettingsAll UsersApplication DataBOINC

---

:) Robert, it's much easier then that. If your messages tab doesn't show clear back to when you last started BOINC Manager, restart the Manager. Within the first 20 messages of starting it will show the data directory path in a message.

---

Rosetta Moderator: Mod.Sense

:) Robert, it's much easier then that. If your messages tab doesn't show clear back to when you last started BOINC Manager, restart the Manager. Within the first 20 messages of starting it will show the data directory path in a message.

While waiting for an answer, I found that it's C:ProgramDataBOINC on my machine - a directory that my search window treats as hidden and not easily searchable. The messages tab agrees.

Now, I'm still looking for the proper commands to zip up all the *.txt and *.old files from that directory and its subdirectories. The help files I can find don't seem to offer that information anywhere.

lb_thread_all_multi_hb_t328__IGNORE_THE_REST_12734_577_1 ran over the my time limit and errored out. It almost looks like it ran twice.


Outcome Client error
Client state Compute error
Exit status 0 (0x0)
BOINC:: CPU time: 29151.9s, 14400s + 14400s[2009- 6-12 13:21: 5:]

then there is this: <file_xfer_error>
<file_name>lb_thread_all_multi_hb_t328__IGNORE_THE_REST_12734_577_1_0</file_name>
<error_code>-161</error_code>
</file_xfer_error>

2009-06-13 00:42:33 rosetta@home Starting tails_homnative_relaxed_9ctf_m00_p00_SAVE_ALL_OUT_12749_9578_0
2009-06-13 00:42:34 rosetta@home Starting task tails_homnative_relaxed_9ctf_m00_p00_SAVE_ALL_OUT_12749_9578_0 using minirosetta version 171

2009-06-13 11:17:16 rosetta@home Computation for task tails_homnative_relaxed_9ctf_m00_p00_SAVE_ALL_OUT_12749_9578_0 finished
2009-06-13 11:17:16 rosetta@home Output file tails_homnative_relaxed_9ctf_m00_p00_SAVE_ALL_OUT_12749_9578_0_0 for task tails_homnative_relaxed_9ctf_m00_p00_SAVE_ALL_OUT_12749_9578_0 absent

And the screen of the window with the error informations:

---

Sorry for the late post - I had some errors while I was away:

lb_thread_all_multi_hb_t374__IGNORE_THE_REST_12748_968_1

CPU Time: 17863.3 (default 14400)
stderr out <core_client_version>6.6.20</core_client_version>
<![CDATA[
<message>
Incorrect function. (0x1) - exit code 1 (0x1)
</message>
<stderr_txt>
<stderr_txt>
value legal range
sin_cos_range ERROR: 1.#QNAN00 is outside of [-1,+1] sin and cos value legal range
(Repeat many times)
dof_atom1 atomno= 3 rsd= 1
atom1 atomno= 1 rsd= 1
atom2 atomno= 2 rsd= 1
atom3 atomno= 5 rsd= 1
atom4 atomno= 6 rsd= 1
THETA1 1.#QNAN00
THETA3 1.#QNAN00
PHI2 1.#QNAN00

ERROR: AtomTree::torsion_angle_dof_id: angle range error
ERROR:: Exit from: ....srccorekinematicsAtomTree.cc line: 754
BOINC:: Error reading and gzipping output datafile: default.out
called boinc_finish

</stderr_txt>

Credit was awarded, but looks nasty and not an error message I've seen reported before.

---