Gearing up for Multibeam

[Simon]

Hi folks,

Multibeam apps will move to the main SETI@Home project shortly. We have been working on a new app release for a while, and decided to wait until MB was rolled out so we wouldn't have two releases in quick succession.

In any case, whenever the project staff are ready to move Multibeam to main, our new release will follow it in quick succession (or be simultaneous, depending on any last-minute issues).

So keep an eye out on the main S@H boards as well as here, new apps should shortly be incoming.

Regards,
Simon.

[Devaster]

read : http://setiathome.berkeley.edu/sah_sci_newsletters.php?frag=news-24.inc 

[Haselgrove]

New information in the SETI technical blog:

Happy Monday, one and all. Not much really exciting to report except that I just code signed the Windows version of the new client, which Eric and I fully plan to release to the public tomorrow. We'll start splitting multibeam workunits shortly after that (the new client can and will process classic data until the new workunits appear). Expect a press release shortly after that (we hope).

(http://setiathome.berkeley.edu/forum_thread.php?id=41393&nowrap=true#615939)

Tomorrow never happened, of course, because the server broke, but does this give you the green light for a final build ready for release?

[Simon]

Pretty much,

though we're still sorting out some last-minute changes regarding the credit multiplier (3.81 at beta, 2.8 for the new apps, 3.35 currently).

When we have reliable info on what the new chosen default multiplier will be, that's when stuff will go live.

Regards,
Simon.

[Gecko_R7]

Just FYI.

I got 5.27 for -doze on main a little bit ago.

[Simon]

Yeah,

the tech news board has a post from Matt saying the Windows version got released and that Linux and Mac versions were to follow shortly.

In any case, I posted a question asking about the default multiplier in that thread a few minutes ago - shucks, would be interesting to know, wouldn't it

If I get a quick answer, I'll start populating the downloads section with 2.4 apps (else I'll go to sleep first).

Now for a front page item about MB apps and switching to them...

Regards,
Simon.

[Josef W. Segur]

Yeah,

the tech news board has a post from Matt saying the Windows version got released and that Linux and Mac versions were to follow shortly.

In any case, I posted a question asking about the default multiplier in that thread a few minutes ago - shucks, would be interesting to know, wouldn't it

If I get a quick answer, I'll start populating the downloads section with 2.4 apps (else I'll go to sleep first).

Now for a front page item about MB apps and switching to them...

Regards,
Simon.

The Windows 5.27 on Main is binary identical to the Windows 5.27 on Beta, hence has the 2.85 multiplier. I'm sure that applies to the Linux 5.27 also, Eric would bump the version number if he made a change.
                                                                              Joe

[Josef W. Segur]

There are a couple of Multibeam splitter processes running on SETI Main now. One interesting change is that the chirp limits are 30 and 100 rather than the 20 and 50 which has been used for Line feed and for Multibeam in SETI Beta. We'll need to add some of these even longer ones to our testing, of course, but it may take awhile to get a useful variety.
                                                                        Joe

[Haselgrove]

The difference between credit claim (2.2B) and credit award (5.27) cross-checks pretty well with a 2.85 multiplier, too.

[Crunch3r]

OK i just received am email from Eric that confirms the 2.85 multi.

Yes, 2.85.  If the <credit_rate> tag is present in the workunit
analysis_config, the value present there will be used.    That will
allow us to vary the multiplier without releasing a new application.

Eric

On 8/10/07, Crunch3r  wrote:
>
>
>
>
> Hello Eric,
>
>
>
> could you please confirm that the new multiplier used for MB apps is 2.85 ?
>
>
>
> We (lunatics.at) need to know for sure before we can release our
> optimized apps.
>
>
>
> Thanks
>
>

[Josef W. Segur]

OK i just received am email from Eriic that confirms the 2.85 multi.

Yes, 2.85.  If the <credit_rate> tag is present in the workunit
analysis_config, the value present there will be used.    That will
allow us to vary the multiplier without releasing a new application.

Eric

On 8/10/07, Crunch3r  wrote:
>
>
>
>
> Hello Eric,
>
>
>
> could you please confirm that the new multiplier used for MB apps is 2.85 ?
>
>
>
> We (lunatics.at) need to know for sure before we can release our
> optimized apps.
>
>
>
> Thanks
>
>

Crunch3r, I have been unable to find anything in the code which transfers the <credit_rate> value to LOAD_STORE_ADJUSTMENT. Would you send him a reply asking about that, he's about to go on vacation and a reply is more likely to catch his attention than a new query from me. I think we have a situation where he intended to add the code but didn't.
                                                                          Joe

[Crunch3r]

I'll send him your reply to my post as well as a link to here. Hopefully i can catch him before he leaves to vacation.

[EDIT]
eMail send. Cross your fingers that we catch him befor he's off.

-----------------------------------------------------------------------------------------


Here's the email that i send to him:

Hi Eric,

I've got a reply from Joe Segur over at lunatics.at.
This is a direct quote of what he wrote over at lunatics.at

http://lunatics.at/discussion-forum/gearing-up-for-multibeam.msg4293.html#msg4293

"Crunch3r, I have been unable to find anything in the code which transfers the <credit_rate> value to LOAD_STORE_ADJUSTMENT. Would you send him a reply asking about that, he's about to go on vacation and a reply is more likely to catch his attention than a new query from me. I think we have a situation where he intended to add the code but didn't.
                                                                          Joe "

Hopefully this mail reaches you before you're headding of to vacation.

Thanks again.





-----Ursprüngliche Nachricht-----
Von: Crunch3r [mailto:.....@......]
Gesendet: Freitag, 10. August 2007 23:05
An: 'Eric J Korpela'
Betreff: AW: MultiBeam apps.

Thanks Eric!

Now we can release the opt apps as well.


-----Ursprüngliche Nachricht-----
Von: k....a@...... [mailto:.....@.....] Im Auftrag von Eric J Korpela
Gesendet: Freitag, 10. August 2007 21:15
An: Crunch3r
Betreff: Re: MultiBeam apps.

Yes, 2.85.  If the <credit_rate> tag is present in the workunit
analysis_config, the value present there will be used.    That will
allow us to vary the multiplier without releasing a new application.

Eric

On 8/10/07, Crunch3r <.....> wrote:
>
>
>
>
> Hello Eric,
>
>
>
> could you please confirm that the new multiplier used for MB apps is 2.85 ?
>
>
>
> We (lunatics.at) need to know for sure before we can release our
> optimized apps.
>
>
>
> Thanks
>
>

[Josef W. Segur]

For the record, the email to Eric got through and he replied:

Damn, I remember writing the code in one of the constructors in seti_header.cpp, but it's definitely not there now.  That's the problem with developing on multiple machines simultaneously, sometimes a change can get lost.  It looks like the 5.27 application is constant at 2.85, and we'll need to do one more release to change it.

If you want yours to do it right and not need to do a later release, the function seti_analyze() in analyzeFuncs.cpp is probably a good place to add an "if (swi.analysis_cfg.credit_rate != 0) LOAD_STORE_ADJUSTMENT=swi.analysis_cfg.credit_rate;"  Like around line 189.

Whatever you do, I'll do the same.

Eric

I went ahead and replied to him:

Eric,

We've already released with the fixed 2.85, so will defer to you
rather than going ahead with a change. Simon and Crunch3r have
been kept pretty busy doing many builds and such, unless you
anticipate using the feature very soon it seems best to leave
the situation as it stands while things settle down.

I apologize for having taken your "(potentially)" in the SETI
Beta post where you mentioned the feature as indicating completion
of the feature was intentionally deferred.

I suppose you also will want to adjust SETUP_FLOPS ?

So no action is needed for now, but we'll probably need to do another set of builds after enough MB work has been done at SETI Main to establish whether further adjustment is needed.
                                                                                     Joe

[comfortablynumb]

We will crunch citizen,just give us optimised apps. We will crunch faster,better,farther than any app b4 us.

There is a theory which states that if ever anybody discovers exactly what the Universe is for and why it is here, it will instantly disappear and be replaced by something even more bizarre and inexplicable.
There is another theory which states that this has already happened.

- Douglas Adams