Boston, I verified that it works on Windows, Linux and OS X already.

As long as you carefully follow the steps in the right order (and backups never hurt), it should work fine.

Regards,
Simon.

---

Donate to SETI@Home via PayPal!

Optimized SETI@Home apps + Information

Can one not do like this. Get Seti to download stock app, then renameing the optimized app to exactly the same name and overwritng the stock app. Will this not make the program think that I´m using stock but I´m using the optimized geting the speed from it. I want to keep crunching for Seti but not give up the extra speed now that I know it´s there.

I tried this and it appeared to work. That is, BOINC used the renamed app (Chicken app renamed as stock app) but my result got a validate error. This may just be a coincidence (since there seems to be a more general problem with validation errors). It might also be because this result was a hybrid (started with the real stock app and finished with the renamed app). I have another one "In Progess" that should finish in about an hour. It will have used the renamed optimized app for the whole way. Guess we can wait and see if it valdates.

Well, it appears the renamed application trick is working. So far, I have two successful results this way: resultid=536418005 and resultid=536325442. For the record, I was never able to get work via the renamed app_info file/reset method.

---

Boston, I verified that it works on Windows, Linux and OS X already.

As long as you carefully follow the steps in the right order (and backups never hurt), it should work fine.

Regards,
Simon.

it worked, so far...

but this is not an replacement for the mormal....

Greetings from Germany NRW
Ulli

SETI@home Enhanced 5.15 is now the name for Ben_Joe on my puter and the Loony Whangy puter. And I have the Bogus named app_info file on my puter and Whangy has it deleted.

Hi again.

More observations. What BOINC Manager says in Tasks-pane as the application name is apparently a user friendly name + version number. If you are running with app_info.xml then the "friendly name"="app name"="setiathome_enhanced". If you are running without app_info.xml BOINC takes the friendly name from scheduler's reply. In that case you'll get "SETI@home Enhanced".

As Philadelphia said you can get executable's name from Task Manager or you could take a look at the slots directory. A UNIX equivalent would be 'top', I think.

Sidenote: Those backups I took while playing around with the files are sure coming handy. Makes it easy to compare various settings and what they look like in config files.

-Juha

---

A thought:

When it will be needed the app_info.xml again?

---

With all the people saying the workaround works/doesn't work/almost works/almost doesn't work I say it's time to list what actually works or doesn't work. Spiced up with excessive amount of useless details that'll just stir this mess up until no-one can make any sense out of it. :) Ok? Well, me first.

My machine: AMD Duron 1.2 GHz, 1024 MB ram, Windows 2000 SP4 + not exactly the latest hotfixes

1. I was born[fast forward]...[play]
2. I installed BOINC 5.2.13, downloaded from BOINC's homepage, a year or two ago, give or take a year or two.
3. Last April I upgraded BOINC to 5.8.16, downloaded from BOINC's homepage. Before upgrade I had made sure that the work queue was empty. That is, no project had any work. I upgraded by installing over the old version. Everything went just fine.
4. Shortly after (a couple of days, week max) Step 3 I installed KWSN's 2.2B (specifically the SSE one).
4.1. Downloaded KWSN's 2.2B from lunatics.at.
4.2. I shutdown BOINC.
4.3. Copied KWSN's app to project folder.
4.4. Didn't like the idea of having a version listed in app_info.xml that was never than the stock app, so I removed the 5.17 from it and left everything else as it was.
4.5. Restarted BOINC.
4.6. It works.
5. At some point BOINC decides stock apps are a waste of disk space and deletes them. I can't really tell when this happened, it's not in the logs, although I am pretty sure it didn't happen immediately.
Thumper dies and is resurrected. Everything still works.
6. Got some workunits.
Tuesdays maintenance break and server's software upgrade. Nothing works.
7. I keep reading the news and since even getting connection to Berkeley is almost impossible I keep SETI at No new tasks most of the time.
8. Workunits downloaded after Thumpers resurrection complete and BOINC uploads results. No luck reporting them.
9. I read about restarting BOINC to get rid of errors so I do that. BOINC reports previous WUs. Still can't get work so SETI stays at No new tasks most of the time. I do manage to get two ghosts though.
10. I read about app_info.xml workaround.
10.1. I shutdown BOINC. Note: I have no work for SETI at this point.
10.2. Rename app_info.xml to app_info.xml.bak.
10.3. Encourage (read: force update) BOINC to contact Berkeley to get work.
10.4. Got a couple of workunits and BOINC downloads them.
10.5. I'm waiting until Spinhenge's WU completes so I can shutdown BOINC. (They don't have checkpointing yet.)
10.6. Spinhenge's WU completes and BOINC starts to crunch SETI's WU.
10.7. For some strange reason I take a look at Task Manager and see KWSN's app there. Whatta heck?!
10.8. After thinking about it for a second or two I _dont_ shutdown BOINC and instead let it continue - just to see what happens.
11. BOINC completes SETI's WUs, uploads results, downloads more WU, reports results. No sign of stock app anywhere. Downloads are one WU at a time. Queue doesn't get empty at any time.

At the moment my machine has download about ten WUs. So far none of the completed (and validated) ones have had any errors. I haven't shutdown BOINC after getting it to work.

I tried to get all the details correct but there may be errors and something important may be missing. Feel free to ask.

Hope this helps

-Juha

---

Can one not do like this. Get Seti to download stock app, then renameing the optimized app to exactly the same name and overwritng the stock app. Will this not make the program think that I´m using stock but I´m using the optimized geting the speed from it. I want to keep crunching for Seti but not give up the extra speed now that I know it´s there.

I tried this and it appeared to work. That is, BOINC used the renamed app (Chicken app renamed as stock app) but my result got a validate error. This may just be a coincidence (since there seems to be a more general problem with validation errors). It might also be because this result was a hybrid (started with the real stock app and finished with the renamed app). I have another one "In Progess" that should finish in about an hour. It will have used the renamed optimized app for the whole way. Guess we can wait and see if it valdates.

Well, it appears the renamed application trick is working. So far, I have two successful results this way: resultid=536418005 and resultid=536325442. For the record, I was never able to get work via the renamed app_info file/reset method.

After the above results were reported, my cache of SETI work was drained. For several hours, I was not able to get any new work. Finally, a few minutes ago my host was able to connect and download a new WU. But, when this happened, it also downloaded a new copy of the stock application and it overwrote the renamed Chicken app. In other words, the renaming trick only works temporarily.

---

If having to do all these various fixes is the the only way to get things going again, I see a problem. Some of us might have enough computer knowledge to do all of this, but where does that leave all of the people that don't? How do they get started again?

---

If having to do all these various fixes is the the only way to get things going again, I see a problem. Some of us might have enough computer knowledge to do all of this, but where does that leave all of the people that don't? How do they get started again?

That's Right A/C, Also the workarounds ( not fixes ) will only work for some types of platform. There are many Platforms ( That use the Anonymous Platform Mechanism ) for which there is no known workaround at the moment. Simply put the workaround is just that, and the proper fix will ultimately have to come at the server end.

Basically, just like people without the nerd skills, those people on 'other platforms' are also stuck up the creek without a paddle, and many will simply be unaware their machines are trying for workunits and generating many many 'ghosts'.

Well I haven't looked at the server stats yet, at least the message board seems to be back up :D

[ LOL 1,179,763,353 Results waiting for deletion .... Ghosts ? ]
[ More LOL: Current result creation rate 94.78/sec, now that's better :D ]

---

"Living by the wisdom of computer science doesn't sound so bad after all. And unlike most advice, it's backed up by proofs." -- Algorithms to live by: The computer science of human decisions.

I wish Seti hadn't changed the Server Software, I mean It worked didn't It? So since It wasn't broke why change It? I hope someone comes in on Monday and rids US of this New Bloatware and give US back the Old Server Software. Sure I can see New software on Non SUN Servers. I just wish It worked. I'll bet nobody can upload or download now. Some might leave for good.

How many times have the immortal words "...while we're at it why don't we..." been uttered??!??

---

If having to do all these various fixes is the the only way to get things going again, I see a problem. Some of us might have enough computer knowledge to do all of this, but where does that leave all of the people that don't? How do they get started again?

That's Right A/C, Also the workarounds ( not fixes ) will only work for some types of platform. There are many Platforms ( That use the Anonymous Platform Mechanism ) for which there is no known workaround at the moment. Simply put the workaround is just that, and the proper fix will ultimately have to come at the server end.

Basically, just like people without the nerd skills, those people on 'other platforms' are also stuck up the creek without a paddle, and many will simply be unaware their machines are trying for workunits and generating many many 'ghosts'.

Well I haven't looked at the server stats yet, at least the message board seems to be back up :D

[ LOL 1,179,763,353 Results waiting for deletion .... Ghosts ? ]
[ More LOL: Current result creation rate 94.78/sec, now that's better :D ]

Well it's one big mess right now. Yeah for now the boards are back up. Last night when the boards went down I thought that maybe something got nuked again. :D

---

Nix didn't work, I'm getting a hole new error message, never seen before..



5/21/2007 7:46:32 PM|SETI@home|Message from server: Project encountered internal error: shared memory

This is not related to the problems that started yesterday. If you see something similar to the follwing:


5/21/2007 2:43:02 PM|SETI@home|Scheduler RPC succeeded
5/21/2007 2:43:02 PM|SETI@home|Message from server: Project encountered internal error: shared memory
5/21/2007 2:43:02 PM|SETI@home|Deferring communication for 1 hr 0 min 0 sec
5/21/2007 2:43:02 PM|SETI@home|Reason: project is down

The above issue does not seem to be related to the Optimized App bug. seti@home is encountering a few new snags it seems. I would just sit tight for now.

---

Having WUs without app_info.xml, can anyone give a time span when we can use it. Or do we have to wait until the next optimised apps appears with this new server.

---

Having WUs without app_info.xml, can anyone give a time span when we can use it. Or do we have to wait until the next optimised apps appears with this new server.

The problem doesn’t involve incompatibility between the new server software and the optimized apps themselves: it’s miscommunication between it and the BOINC clients. I’ve successfully crunched a number of WUs with an optimized app, as have many others, using a workaround that’s been posted and discussed here; they’re getting validated just fine.

I have no idea of when to expect it, but no doubt the server code will either get rolled back to the last working version or be updated with a bug-fix. Until Multibeam is released here, probably requiring a setiathome_enhanced v5.18+ app, the existing optimized v5.13 & v5.15 apps will continue to work—whenever BOINC gives them a chance.

---

I think it might be working now! Using app_info.xml and successfully downloading work, that is.

Mon May 21 15:05:55 2007||Starting BOINC client version 5.9.11 for i686-apple-darwin
Mon May 21 15:05:55 2007||log flags: task, file_xfer, sched_ops
Mon May 21 15:05:55 2007||Libraries: libcurl/7.16.1 OpenSSL/0.9.7l zlib/1.2.3
Mon May 21 15:05:55 2007||Data directory: /Library/Application Support/BOINC Data
Mon May 21 15:05:55 2007|SETI@home|Found app_info.xml; using anonymous platform
Mon May 21 15:05:55 2007||Processor: 8 GenuineIntel Intel(R) Xeon(R) CPU X5365 @ 3.00GHz [x86 Family 6 Model 15 Stepping 7]
Mon May 21 15:05:55 2007||Processor features: FPU VME DE PSE TSC MSR PAE MCE CX8 APIC SEP MTRR PGE MCA CMOV PAT PSE36 CLFSH DS ACPI MMX FXSR SSE SSE2 SS HTT TM SSE3 MON DSCPL VMX EST TM2 MNI CX16 TPR
Mon May 21 15:05:55 2007||Memory: 4.00 GB physical, 629.17 GB virtual
Mon May 21 15:05:55 2007||Disk: 698.32 GB total, 628.93 GB free
Mon May 21 15:05:55 2007|SETI@home|URL: http://setiathome.berkeley.edu/; Computer ID: 3360245; location: work; project prefs: default
Mon May 21 15:05:55 2007||General prefs: from http://www.worldcommunitygrid.org/ (last modified 2007-05-20 21:24:06)
Mon May 21 15:05:55 2007||Host location: work
Mon May 21 15:05:55 2007||General prefs: using separate prefs for work
Mon May 21 15:05:55 2007||Preferences limit memory usage when active to 4055.04MB
Mon May 21 15:05:55 2007||Preferences limit memory usage when idle to 4055.04MB
Mon May 21 15:05:55 2007||Preferences limit disk usage to 93.13GB
Mon May 21 15:05:55 2007|SETI@home|Restarting task 18fe05aa.16472.33488.434644.3.38_1 using setiathome_enhanced version 513
Mon May 21 15:05:55 2007|SETI@home|Restarting task 18fe05aa.16472.33488.434644.3.106_0 using setiathome_enhanced version 513
Mon May 21 15:05:55 2007|SETI@home|Restarting task 18fe05aa.16472.33488.434644.3.51_0 using setiathome_enhanced version 513
Mon May 21 15:05:55 2007|SETI@home|Restarting task 04mr05ab.17213.29024.40886.3.21_0 using setiathome_enhanced version 513
Mon May 21 15:05:55 2007|SETI@home|Restarting task 04mr05ab.17213.29024.40886.3.52_1 using setiathome_enhanced version 513
Mon May 21 15:05:55 2007|SETI@home|Restarting task 04mr05ab.17213.29024.40886.3.38_0 using setiathome_enhanced version 513
Mon May 21 15:05:56 2007|SETI@home|Restarting task 04mr05ab.17213.29024.40886.3.37_0 using setiathome_enhanced version 513
Mon May 21 15:05:56 2007|SETI@home|Restarting task 04mr05ab.17213.29024.40886.3.60_2 using setiathome_enhanced version 513
Mon May 21 15:06:31 2007|SETI@home|Sending scheduler request: Requested by user
Mon May 21 15:06:31 2007|SETI@home|Requesting 1065573 seconds of new work
Mon May 21 15:06:41 2007|SETI@home|Scheduler RPC succeeded [server version 509]
Mon May 21 15:06:41 2007|SETI@home|Deferring communication for 11 sec
Mon May 21 15:06:41 2007|SETI@home|Reason: requested by project
Mon May 21 15:06:41 2007|SETI@home|Deferring communication for 1 min 0 sec
Mon May 21 15:06:41 2007|SETI@home|Reason: no work from project ( <--- probably nothing more than overloaded server)
Mon May 21 15:07:41 2007|SETI@home|Sending scheduler request: To fetch work
Mon May 21 15:07:41 2007|SETI@home|Requesting 1067305 seconds of new work
Mon May 21 15:07:51 2007|SETI@home|Scheduler RPC succeeded [server version 509]
Mon May 21 15:07:51 2007|SETI@home|Message from server: No work sent
Mon May 21 15:07:51 2007|SETI@home|Message from server: (reached daily quota of 400 results) ( <--- valid response, as I already hit my limit today)

But the point here, is that I am no longer getting the "internal html error" message.

---

Dublin, California
Team: SETI.USA

I think it might be working now! Using app_info.xml and successfully downloading work, that is.

...snip...

Mon May 21 15:06:41 2007|SETI@home|Reason: no work from project ( <--- probably nothing more than overloaded server)
Mon May 21 15:07:41 2007|SETI@home|Sending scheduler request: To fetch work
Mon May 21 15:07:41 2007|SETI@home|Requesting 1067305 seconds of new work
Mon May 21 15:07:51 2007|SETI@home|Scheduler RPC succeeded [server version 509]
Mon May 21 15:07:51 2007|SETI@home|Message from server: No work sent
Mon May 21 15:07:51 2007|SETI@home|Message from server: (reached daily quota of 400 results) ( <--- valid response, as I already hit my limit today)

But the point here, is that I am no longer getting the "internal html error" message.

Check Message 572237 (this thread, yesterday).

I think you only get internal server error when it actually tries to allocate work. If it doesn't allocate work (congestion, quota) the problem doesn't arise.

I'm afraid my canary isn't looking very chirpy.

Check Message 572237 (this thread, yesterday).

I think you only get internal server error when it actually tries to allocate work. If it doesn't allocate work (congestion, quota) the problem doesn't arise.

I'm afraid my canary isn't looking very chirpy.

Okay. It'll have to wait until tomorrow to corroborate.

---

Dublin, California
Team: SETI.USA

I think it might be working now! Using app_info.xml and successfully downloading work, that is.

<snippety>

But the point here, is that I am no longer getting the "internal html error" message.

You're lucky, I've been getting nothing but "Scheduler request failed: HTTP internal server error" for the last two days solid, every time BOINC tries to connect to the SETI servers (whether or not "app_info.xml" is present).

I can't even update my preferences from the SETI site!


Oh well, at least these errors are no longer generating "ghost" WUs. :-)


TTFN - Pete.

---

Thanks for the info, I can do the WU with the optimised app but cannot download with it. SO will carry on not using it until it is fixed.

---

I have stopped BOINC, renamed the app_info, and restarted BOINC (even upgraded to the 5.9.11 beta), and they all continue to crunch with the optimized application. It downloaded the regular application, but still uses the optimized one, without issue.

me too on my 7 computers (although I did not upgrade to 5.9.11) and they are all working fine.

I made the changes on a friend's 4 computers and they are working fine as well.

Thanks,
Bill

---

Bill
Auckland, NZ