Eric's biannual post #6: You can tuna fish, but you can't tune a TCP

Message boards : SETI@home Staff Blog : Eric's biannual post #6: You can tuna fish, but you can't tune a TCP
Message board moderation

To post messages, you must log in.

Previous · 1 . . . 3 · 4 · 5 · 6 · 7 · 8 · Next

AuthorMessage
Richard Haselgrove Project Donor
Volunteer tester

Send message
Joined: 4 Jul 99
Posts: 14644
Credit: 200,643,578
RAC: 874
United Kingdom
Message 571340 - Posted: 19 May 2007, 16:26:22 UTC - in response to Message 571317.  

Well since It's a server problem, Will Seti go back to the Previous software or Are they going to stick anyone who uses any non stock app in the "No Work Zone"? Cause until It's fixed one way or the other, It's like only the Stock app is supported by Seti and It may as well be closed source, Cause Seti is sure acting like It unless One tricks Seti with a work around just to get downloads Which is Tedious on more than 1 PC and I've got 5 PCs now. So Somebody fix this, Please? :(

I think SETI has two choices. Go forwards, or go back. Standing still isn't (or shouldn't be) an option.

I would guess that SETI would prefer to go forwards, rather than backwards. It should be possible to write server code which includes the new functions they're trying to roll out, but at the same time maintains compatibility with the older recommended clients. The question is, how long will that take?

I've done everything I can to alert the BOINC programmers that there's a problem which needs solving urgently, but so far I've had no feedback at all to indicate whether they're working on it. That may be a good sign (too busy to blog), or a bad sign (they're all on a weekend hike in the Rockies).

Also I have no idea at all whether either the BOINC team or the SETI team are going to come into their offices this Saturday. We'll wait and see - their call.

What I do hope is that they have a review meeting on Monday morning, decide which direction to go in - and then let us know what they've decided.
ID: 571340 · Report as offensive
Profile zoom3+1=4
Volunteer tester
Avatar

Send message
Joined: 30 Nov 03
Posts: 65690
Credit: 55,293,173
RAC: 49
United States
Message 571373 - Posted: 19 May 2007, 16:58:24 UTC - in response to Message 571340.  
Last modified: 19 May 2007, 17:03:29 UTC

Well since It's a server problem, Will Seti go back to the Previous software or Are they going to stick anyone who uses any non stock app in the "No Work Zone"? Cause until It's fixed one way or the other, It's like only the Stock app is supported by Seti and It may as well be closed source, Cause Seti is sure acting like It unless One tricks Seti with a work around just to get downloads Which is Tedious on more than 1 PC and I've got 5 PCs now. So Somebody fix this, Please? :(

I think SETI has two choices. Go forwards, or go back. Standing still isn't (or shouldn't be) an option.

I would guess that SETI would prefer to go forwards, rather than backwards. It should be possible to write server code which includes the new functions they're trying to roll out, but at the same time maintains compatibility with the older recommended clients. The question is, how long will that take?

I've done everything I can to alert the BOINC programmers that there's a problem which needs solving urgently, but so far I've had no feedback at all to indicate whether they're working on it. That may be a good sign (too busy to blog), or a bad sign (they're all on a weekend hike in the Rockies).

Also I have no idea at all whether either the BOINC team or the SETI team are going to come into their offices this Saturday. We'll wait and see - their call.

What I do hope is that they have a review meeting on Monday morning, decide which direction to go in - and then let us know what they've decided.

Well except for Seti, I think most or all the other Boinc projects prefer No 3rd party apps, Or am I wrong on this? I hope that the New Boinc code takes into account the 3rd party apps though.

And Yes I agree, It's go forward or go back, But Seti may not want to go back and so I hope they fix this soon.
The T1 Trust, PRR T1 Class 4-4-4-4 #5550, 1 of America's First HST's
ID: 571373 · Report as offensive
Profile Demiurg
Volunteer tester
Avatar

Send message
Joined: 2 Jul 02
Posts: 883
Credit: 28,286
RAC: 0
Sweden
Message 571395 - Posted: 19 May 2007, 17:18:59 UTC

I have posted a warning about the app_info bugfix at beta number crunching. To lazy to repost. To cranky. Stay the *poot* away from it.

Carl
It is SEXY to DONATE!
Skype = demiurg2
ID: 571395 · Report as offensive
picantecomputing
Volunteer tester

Send message
Joined: 1 Jan 07
Posts: 4
Credit: 104,652
RAC: 0
United States
Message 571405 - Posted: 19 May 2007, 17:27:56 UTC
Last modified: 19 May 2007, 17:29:09 UTC

I'm not quite sure why, but I'm running optimized apps just fine. When I first did the app_info fix, it worked (once). Then when I reverted to the original file name, it stopped working. So I did what seemed to be the next logical thing and renamed the app_info file again and left it that way. And I've been uploading and downloading with zero problems (except for occasionally not being able to update, which eventually works itself out) for a full day now. I have no idea why the renamed file isn't fouling things up, but at least on my two crunchers (one PC, one Mac), it's running smooth as butta. Hope that helps.
ID: 571405 · Report as offensive
Profile Blurf
Volunteer tester

Send message
Joined: 2 Sep 06
Posts: 8962
Credit: 12,678,685
RAC: 0
United States
Message 571483 - Posted: 19 May 2007, 18:21:59 UTC

I removed the app_info.xml file (put it on the desktop) and restarted BOINC and got me a full tub of work on all machines. I've put it back in and restored it several times with no difficulty. Just shut down BOINC before doing it


ID: 571483 · Report as offensive
zombie67 [MM]
Volunteer tester
Avatar

Send message
Joined: 22 Apr 04
Posts: 758
Credit: 27,771,894
RAC: 0
United States
Message 571602 - Posted: 19 May 2007, 20:03:11 UTC - in response to Message 571395.  

I have posted a warning about the app_info bugfix at beta number crunching. To lazy to repost. To cranky. Stay the *poot* away from it.

Why? It works perfectly...and it's easy to do. The only problem is that it's not automated.
Dublin, California
Team: SETI.USA
ID: 571602 · Report as offensive
Richard Haselgrove Project Donor
Volunteer tester

Send message
Joined: 4 Jul 99
Posts: 14644
Credit: 200,643,578
RAC: 874
United Kingdom
Message 571617 - Posted: 19 May 2007, 20:20:38 UTC - in response to Message 571602.  

I have posted a warning about the app_info bugfix at beta number crunching. To lazy to repost. To cranky. Stay the *poot* away from it.

Why? It works perfectly...and it's easy to do. The only problem is that it's not automated.

Have a look at Jason Gee's profile
ID: 571617 · Report as offensive
zombie67 [MM]
Volunteer tester
Avatar

Send message
Joined: 22 Apr 04
Posts: 758
Credit: 27,771,894
RAC: 0
United States
Message 571917 - Posted: 20 May 2007, 1:21:04 UTC - in response to Message 571617.  

I have posted a warning about the app_info bugfix at beta number crunching. To lazy to repost. To cranky. Stay the *poot* away from it.

Why? It works perfectly...and it's easy to do. The only problem is that it's not automated.

Have a look at Jason Gee's profile

I don't think that applies to my Macs.... =;^)
Dublin, California
Team: SETI.USA
ID: 571917 · Report as offensive
Brian Silvers

Send message
Joined: 11 Jun 99
Posts: 1681
Credit: 492,052
RAC: 0
United States
Message 571959 - Posted: 20 May 2007, 2:06:02 UTC - in response to Message 571917.  

I have posted a warning about the app_info bugfix at beta number crunching. To lazy to repost. To cranky. Stay the *poot* away from it.

Why? It works perfectly...and it's easy to do. The only problem is that it's not automated.

Have a look at Jason Gee's profile

I don't think that applies to my Macs.... =;^)


...however you could take some time and modify it to apply to your Macs and then post it in your profile and help others... :)

At any rate, I asked some that had the expertise on various platforms to make an automated job for the common platforms, so alternatively you could wait on someone to automate it for Mac platforms...
ID: 571959 · Report as offensive
zombie67 [MM]
Volunteer tester
Avatar

Send message
Joined: 22 Apr 04
Posts: 758
Credit: 27,771,894
RAC: 0
United States
Message 572201 - Posted: 20 May 2007, 13:15:32 UTC - in response to Message 571959.  

Why? It works perfectly...and it's easy to do. The only problem is that it's not automated.

Have a look at Jason Gee's profile

I don't think that applies to my Macs.... =;^)


...however you could take some time and modify it to apply to your Macs and then post it in your profile and help others... :)

At any rate, I asked some that had the expertise on various platforms to make an automated job for the common platforms, so alternatively you could wait on someone to automate it for Mac platforms...


I just don't have the l33t sklz to do something like that. My point earlier in this thread was to show that the work-around...er...worked. No problems at all. Someone was saying that it didn't work.
Dublin, California
Team: SETI.USA
ID: 572201 · Report as offensive
Profile Demiurg
Volunteer tester
Avatar

Send message
Joined: 2 Jul 02
Posts: 883
Credit: 28,286
RAC: 0
Sweden
Message 572248 - Posted: 20 May 2007, 15:50:05 UTC - in response to Message 571602.  

I have posted a warning about the app_info bugfix at beta number crunching. To lazy to repost. To cranky. Stay the *poot* away from it.

Why? It works perfectly...and it's easy to do. The only problem is that it's not automated.


It wrecked boinc so badly that I had 63 beta and main WUs get computational errors in a row. It whacked the ben_joe app beyond repair. So I had to reinstall everything and ditch the 63 WUs:-(
And yes I tested everything before doing it.

Carl
It is SEXY to DONATE!
Skype = demiurg2
ID: 572248 · Report as offensive
Profile Keith Myers Special Project $250 donor
Volunteer tester
Avatar

Send message
Joined: 29 Apr 01
Posts: 13161
Credit: 1,160,866,277
RAC: 1,873
United States
Message 572448 - Posted: 20 May 2007, 20:32:19 UTC - in response to Message 570723.  

Hmmm. OS/2 Warp, if I'm not mistaken. Looking at the SETI applications page, your operating system is not listed, so I guess there's no official, Berkeley-supported and Berkeley-hosted application for you.

So it must be a home-build or team-build version, right? And you must use the app_info.xml mechanism to get it to run?

That is, in fact, the primary purpose of the "anonymous platform" technique. And it's broken. If you use app_info, you get no work. And if you don't use app_info, you get no work. B****r.

All I can suggest is that you put pressure on David Anderson to get the BOINC Server source code fixed, perhaps adding to my bug report (trac ticket 194), or you put pressure on Eric and Matt to down-grade the BOINC servers to the earlier version of the BOINC server software. Neither will be easy, or welcome, but it's the best I can suggest. Sorry.

Edit - third possibility - could your team get the latest BOINC v5.9.11 to compile on OS/2? If so, there's a report in the Bugs...but where? thread - unconfirmed as yet - that the app_info will work again, and your existing application should work too.

Real edit - sorry, hit 'reply' to my own post instead of 'edit' - past midnight here, and I've been working on this one for 14 hours: time for bed very soon. But you get my drift....


Hi Richard, thanks for the skinny. I have been having no luck understanding all the tricks for the Windows platform and how to translate them to OS/2. We had one anonymous benefactor who was able to compile the BOINC and SETI source code for SETI Enhanced after the demise of Classic. But that has been the only application we have ever received. No updates to newer clients or newer BOINC platform software.. It was barely possible to compile the original source code from what I have been told. So unless SETI fixes the problem on their server side, it seems that all of us running OS/2 will have to say goodbye to SETI and all distributed computing. A real shame.

Thanks for the explanation, Keith

Seti@Home classic workunits:20,676 CPU time:74,226 hours

A proud member of the OFA (Old Farts Association)
ID: 572448 · Report as offensive
Brian Silvers

Send message
Joined: 11 Jun 99
Posts: 1681
Credit: 492,052
RAC: 0
United States
Message 572458 - Posted: 20 May 2007, 20:41:57 UTC - in response to Message 572448.  
Last modified: 20 May 2007, 20:42:43 UTC

So unless SETI fixes the problem on their server side, it seems that all of us running OS/2 will have to say goodbye to SETI and all distributed computing. A real shame.


Keith,

Am I remembering correctly that an unofficial addon somewhere enabled Win32 (ala Windows 95) support in a Win-OS/2 emulation mode? If so, couldn't you run in that mode instead of OS/2 native? If it is still stuck at Windows 3.1 emulation (like it was when I left OS/2 at version 2.1), then nevermind...

Brian
ID: 572458 · Report as offensive
Profile Keith Myers Special Project $250 donor
Volunteer tester
Avatar

Send message
Joined: 29 Apr 01
Posts: 13161
Credit: 1,160,866,277
RAC: 1,873
United States
Message 572965 - Posted: 21 May 2007, 17:45:06 UTC - in response to Message 572458.  

So unless SETI fixes the problem on their server side, it seems that all of us running OS/2 will have to say goodbye to SETI and all distributed computing. A real shame.


Keith,

Am I remembering correctly that an unofficial addon somewhere enabled Win32 (ala Windows 95) support in a Win-OS/2 emulation mode? If so, couldn't you run in that mode instead of OS/2 native? If it is still stuck at Windows 3.1 emulation (like it was when I left OS/2 at version 2.1), then nevermind...

Brian


Hi Brian, no it never worked. We have ODIN that can run some very simple Windows 32 bit programs and there is some directly ported stuff from Innotek that uses some Windows 32 bit libraries but they are rare. I tried to run the Windows app under ODIN and it failed miserably and wouldn't even unpack or install. Thanks for the help anyway.

Keith

Seti@Home classic workunits:20,676 CPU time:74,226 hours

A proud member of the OFA (Old Farts Association)
ID: 572965 · Report as offensive
Richard Haselgrove Project Donor
Volunteer tester

Send message
Joined: 4 Jul 99
Posts: 14644
Credit: 200,643,578
RAC: 874
United Kingdom
Message 572971 - Posted: 21 May 2007, 17:48:54 UTC - in response to Message 572965.  

So unless SETI fixes the problem on their server side, it seems that all of us running OS/2 will have to say goodbye to SETI and all distributed computing. A real shame.


Keith,

Am I remembering correctly that an unofficial addon somewhere enabled Win32 (ala Windows 95) support in a Win-OS/2 emulation mode? If so, couldn't you run in that mode instead of OS/2 native? If it is still stuck at Windows 3.1 emulation (like it was when I left OS/2 at version 2.1), then nevermind...

Brian


Hi Brian, no it never worked. We have ODIN that can run some very simple Windows 32 bit programs and there is some directly ported stuff from Innotek that uses some Windows 32 bit libraries but they are rare. I tried to run the Windows app under ODIN and it failed miserably and wouldn't even unpack or install. Thanks for the help anyway.

Keith

Keith,

Have you seen the changes over at trac ticket 194? You may want to add another comment about your (lack of) ability to get new versions of BOINC.
ID: 572971 · Report as offensive
Brian Silvers

Send message
Joined: 11 Jun 99
Posts: 1681
Credit: 492,052
RAC: 0
United States
Message 573000 - Posted: 21 May 2007, 18:09:43 UTC - in response to Message 572965.  
Last modified: 21 May 2007, 18:10:25 UTC

So unless SETI fixes the problem on their server side, it seems that all of us running OS/2 will have to say goodbye to SETI and all distributed computing. A real shame.


Keith,

Am I remembering correctly that an unofficial addon somewhere enabled Win32 (ala Windows 95) support in a Win-OS/2 emulation mode? If so, couldn't you run in that mode instead of OS/2 native? If it is still stuck at Windows 3.1 emulation (like it was when I left OS/2 at version 2.1), then nevermind...

Brian


Hi Brian, no it never worked. We have ODIN that can run some very simple Windows 32 bit programs and there is some directly ported stuff from Innotek that uses some Windows 32 bit libraries but they are rare. I tried to run the Windows app under ODIN and it failed miserably and wouldn't even unpack or install. Thanks for the help anyway.

Keith


OK. I last ran OS/2 in like 1996, so I am not up on all of what happened since about Warp 3. Shame that IBM abandoned it though...

Brian
ID: 573000 · Report as offensive
Profile Keith Myers Special Project $250 donor
Volunteer tester
Avatar

Send message
Joined: 29 Apr 01
Posts: 13161
Credit: 1,160,866,277
RAC: 1,873
United States
Message 573020 - Posted: 21 May 2007, 18:28:28 UTC - in response to Message 572971.  

So unless SETI fixes the problem on their server side, it seems that all of us running OS/2 will have to say goodbye to SETI and all distributed computing. A real shame.


Keith,

Am I remembering correctly that an unofficial addon somewhere enabled Win32 (ala Windows 95) support in a Win-OS/2 emulation mode? If so, couldn't you run in that mode instead of OS/2 native? If it is still stuck at Windows 3.1 emulation (like it was when I left OS/2 at version 2.1), then nevermind...

Brian


Hi Brian, no it never worked. We have ODIN that can run some very simple Windows 32 bit programs and there is some directly ported stuff from Innotek that uses some Windows 32 bit libraries but they are rare. I tried to run the Windows app under ODIN and it failed miserably and wouldn't even unpack or install. Thanks for the help anyway.

Keith

Keith,

Have you seen the changes over at trac ticket 194? You may want to add another comment about your (lack of) ability to get new versions of BOINC.



Richard, I see they have closed the ticket. What I'm not sure of is whether they are implying that the issue or ticket is closed because a new client solves the problem. As you understand well is that we do not have access to a client app programmer. The "Yuri" guy is a phantom of some sorts. It was over a month before our 5.13 and 5.49 client apps showed up after shutting down of Classic. The link to the temporary server that hosted the apps for a short time is long gone. Nobody has a good email address to contact the programmer. We are just hoping that he is monitoring this mess and will once again come to our rescue.

Do you think that SETI's message about the problem from yesterday being solved means that our older anonymous platform apps will work again?

Keith

Seti@Home classic workunits:20,676 CPU time:74,226 hours

A proud member of the OFA (Old Farts Association)
ID: 573020 · Report as offensive
Profile Keith Myers Special Project $250 donor
Volunteer tester
Avatar

Send message
Joined: 29 Apr 01
Posts: 13161
Credit: 1,160,866,277
RAC: 1,873
United States
Message 573024 - Posted: 21 May 2007, 18:34:59 UTC - in response to Message 573000.  

So unless SETI fixes the problem on their server side, it seems that all of us running OS/2 will have to say goodbye to SETI and all distributed computing. A real shame.


Keith,

Am I remembering correctly that an unofficial addon somewhere enabled Win32 (ala Windows 95) support in a Win-OS/2 emulation mode? If so, couldn't you run in that mode instead of OS/2 native? If it is still stuck at Windows 3.1 emulation (like it was when I left OS/2 at version 2.1), then nevermind...

Brian


Hi Brian, no it never worked. We have ODIN that can run some very simple Windows 32 bit programs and there is some directly ported stuff from Innotek that uses some Windows 32 bit libraries but they are rare. I tried to run the Windows app under ODIN and it failed miserably and wouldn't even unpack or install. Thanks for the help anyway.

Keith


OK. I last ran OS/2 in like 1996, so I am not up on all of what happened since about Warp 3. Shame that IBM abandoned it though...

Brian



Brian, actually, Serenity Systems International has been doing a quite credible job of enhancing and promoting the current invocation of OS/2 Warp. They are the defacto handlers of all current OS/2 development along with their parters Innotek Systems GMbH. Yes, shame that Microsoft killed the development of OS/2.

Keith

Seti@Home classic workunits:20,676 CPU time:74,226 hours

A proud member of the OFA (Old Farts Association)
ID: 573024 · Report as offensive
Richard Haselgrove Project Donor
Volunteer tester

Send message
Joined: 4 Jul 99
Posts: 14644
Credit: 200,643,578
RAC: 874
United Kingdom
Message 573093 - Posted: 21 May 2007, 19:14:08 UTC - in response to Message 573020.  
Last modified: 21 May 2007, 19:15:19 UTC


Do you think that SETI's message about the problem from yesterday being solved means that our older anonymous platform apps will work again?

Keith

No, I'm afraid I don't: I think that message just relates to the complete server breakage yesterday.

But I do think that they've got a fresh team in the office today, and they are fully focussed on getting it working: we had a developer call 'Janus' pop up in Number Crunching an hour ago, asking for tests to see if the 'shared memory' error had been fixed. (It has, but not the app_info). We need to watch, test, watch, test and watch again.

I do hope, however, that we'll get a progress blog late afternoon (PDT), then we can start to enumerate what has been fixed and what is still broken. If anonymous platform is still on the second list, we start asking why.
ID: 573093 · Report as offensive
Profile Blurf
Volunteer tester

Send message
Joined: 2 Sep 06
Posts: 8962
Credit: 12,678,685
RAC: 0
United States
Message 573121 - Posted: 21 May 2007, 19:50:09 UTC

Running chicken apps again apparently with no problem...this was after a personal discussion with Eric


ID: 573121 · Report as offensive
Previous · 1 . . . 3 · 4 · 5 · 6 · 7 · 8 · Next

Message boards : SETI@home Staff Blog : Eric's biannual post #6: You can tuna fish, but you can't tune a TCP


 
©2024 University of California
 
SETI@home and Astropulse are funded by grants from the National Science Foundation, NASA, and donations from SETI@home volunteers. AstroPulse is funded in part by the NSF through grant AST-0307956.