
	SDAST Chat meeting #64: Thursday 5th September 2002
	
Participants: Carol Anne Oxborrow, Stefan Larsson, Sami Maisala

Subjects discussed: Detailed component descriptions
		Changes in deadtime calculation
		Setting up the Stockholm repository
		Scientific validation
		
Action Items: none		


**** Logging Started : Thu Sep 05 13:28:58 MEST 2002
> Hi Sami! I see this time suits you better than 10.30 - I think we should stick
> with it. END
*** larsson (~larsson@tenma.dsri.dk) has joined channel #jemxadr
<larsson> Hi CA and Sami! END
> Hi Stefan! How're things going with the data repository in Stockholm? END 
<smaisala> Hi all. Yes, this time suits perfectly. END
<larsson> Hi CA. Things will not be physicly set up untill the end of this
<larsson> month. We will probably use one of our existing machines
<larsson> initially, and then move over to a RAID system later this year. END
> You'll be able to manage that switch over with the data pouring in? END
> NJW isn't coming to chat today - he's sick. Does anyone know if PK's in
> the office today, or was he off somewhere to give that talk? END
<larsson> Yes. That part is not a problem (with a little help from my friends).
<larsson> One or two of our grad students will be responsible for the
<larsson> maintanence. END
> That's good news that there's some person power already allocated for these
> things. I'll just see if I can get hold of Peter on the phone. END
<larsson> I dont know about PK. However since I am sick too I am able to sit
<larsson> home and do Jem-X work instead of teaching today! END
> Okay guys, no reply at the Kretschmar homestead - I suggest a very short
> meeting in that case. I expected Peter to comment on the detailed descriptions
> we all sent him and masha. You have done that haven't you? END
<larsson> Sort of... END
> I'll admit I'm still bogged down in the description of dead-time-calc because
> we
> 've got to change the way it works totally, and I'm still  trying to work out
> if Niels Lund's very nice and simple equation is correct. I'm thinking of
<smaisala> I sent something to Peter but I forgot Masha ...
> sending Masha what I have so far and leaving that last component till the 
> SCREW's accepted and the component re-coded. END
> Masha's happy to get anything that's user friendly (remember the theoretical
> user in Timbuktu), so if you send her what you sent Peter she can at least
> add it to the User Manual. END
<smaisala> Ok. I think I like to little bit improve my text before I send it to the Masha. END
> I'd also expected Peter to tell us about his presentation, but we'll have
> to hear about that next week. Stefan do you have anything to say about
> Scientific Validation? END
<larsson> Sorry no. I am still struggling with my delivery of j_src_spectra
<larsson> I have finally got the problems with flux normalization sorted
<larsson> out. At least my tests give nice results. I am just doing
<larsson> the corresponding update of the error estimates.
<larsson> will try (again) to get it delivered by this evening so NJ
<larsson> can (I hope ) test it too. END
> The way NL wants the deadtime done now will involve calculating a factor
> by which to multiply all measureed events to get the actual number of events
> entering the detector - I'll then convert this into effective deadtime to
> stick with the data structure format we've agreed for JMXi-DEAD-SCP. However,
> NL's equation also takes events lost due to grey filter into account, so it
> would be wrong to correct for both deadtime and grey filter further down the
> pipeline. How much does this affect you further-down-the-pipeline guys? END
<larsson> It is easy for me to adapt to that.
<larsson> From your deadtimes and greyfilters I actually calculate precisly
<larsson> that type of correction faktor which Niels eq. gives. END.
> Would it be a whole lot better then if I simply provided that factor in
> DEAD-SCP? END
<larsson> I think it would me cleaner to have that as a column too.
<larsson> I dont mind to change to take the factor directly but maybe both
> yes, an extra column in the data structure is probably the way to go. END
<larsson> columns should be there to make sure all software is consistent. END
> Okay, this is a couple of weeks away yet, so I'll solicit some other 
> end-of-the-pipeline opinions before bombarding Marc with new templates. END
> Isabelle has asked me to provide a montly report update next Friday
> (the 13th as it happens) - could you all be sure to send me a list of what
> you've done in the last few weeks, especially on deliveries, documentation
> and setting up the Stockholm repository. END
<larsson> I'll add it to my "todo"-list. END
> Thanks. What about you Sami? Any comments so far? How does a change in
> the composition of deadtime affect you? END
<smaisala> It has no effect at all. I do not use that information. END
> That's very good news. END
> Okay point 5: Background handling - I really wanted to ask for the opinions
> of the astronomers on this one: are they happy with what's there etc. etc.
> but the astronomers are in the minority today......any comments Stefan? END
<larsson> I must admitt that I am so deep into other tasks right now that
<larsson> I have had not spare time to think about this. END
> That's fine with me! Has anyone done an AIs? SL here's one just for you:
> AI020502_1 SL	Aug 2002 Spend a day performing analysis of SV results made by
OPEN			Peter, and make a report of the results for the SVR


> How about it? Sami seems to be AI-free except for AIs on all of us. eND
<larsson> Well actually I did spend a few days on thoose simulations in june.
> Does that mean the AI's closed? (Hint: please say `yes')END
<larsson> I also reached some conclusions but since the analysis was run with
<larsson> by now obsolite version of j_src_spectra I have simply moved on.
<larsson> NJs newer simulations have been usefull in sorting out the flux
<larsson> problem which emerged from the previous simulations. END
<larsson> YES. yes, yes END
> I'll assume that particular AI's closed: if we need more analyses we'll make
> a new AI. END
<larsson> NJ and I have had an email exchange on some of the issues (including Peter
<larsson> in the loop of discussion). END
> That just leaves AOB: anyone for AOB? END
<larsson> No OBs
<smaisala> Question for you CAO. How your v3 coding is going on?
> I'd really like to get back to my documentation.
> all my V3 stuff is on hold: there's NO code yet, but at least one more of the
> ical tools is unneeded since there won't be any deadtime model to update.
> Perhaps if I wait long enough all my V3 will just float away. END
<smaisala> Ok. I'm struggling with ROOT. It's not so easy... :(
> I don't have time for ROOT now - I had big plans at the beginning of the
> year, but it's just so time-consuming make those nice  GUIs - our instrument
> guys certainly don't need anything that fancy, and my suspicion is that they'll
> use all their own tools for making updates of the instrument model tables:
> so no GUIs for the guys. END
> We'll if that's all, I have to run, my husband'd coming back from London today
> AND it's his birthday, so that means a fair deal of shopping has to be done for
> supper. Chat with you both next week. END
*** peter (~pkretsch@isdcul14.unige.ch) has joined channel #jemxadr
<smaisala> Ok. Have a nice day. END
