09:25:08	 From Andrew Nelson : I feel seen right now.
09:52:41	 From Murphy, Dr. rer. nat. Bridget : Thanks Stuart for a very thought provoking talk and for clearly outlining best practice.
09:52:43	 From Andrew McCluskey (he/him)  To  Stuart Prescott(privately) : Lovely talk (and cheers to the shout out). I am curious about the comment that Notebooks probably aren’t the best way to share reproducible workflows (for the reasons I totally agree with). What do you think it the best option instead? Raw py?
09:55:00	 From Stephen Holt : Interesting presentation thanks Stuart.  It was however very much from the perspective of data collected at a large facility. We have recently been looking at new labortory x-ray instruments. These all include their own closed data reduction and analysis packages.  While we may not use these ourselfes I suspect that there are many users of laboratory instruments out there who do use the supplied closed tools.  There is therefore potentially a large barrier to providing the information related to reduction and analysis.  Something to think about for workshop discussion.  While we can encourage/agitate for changes at large facilities what about laboratory sources ?
09:57:29	 From Adrian Rennie : Stephen - I agree.  We need to address the needs of the large community working with laboratory X-rays.
10:07:35	 From Stuart Prescott : Stephen & Adrian: the culture of the facility is definitely much more open (formats, tools, collaboration) than the average benchtop instrument. That said, other communities have had success in defining formats and interfaces that are "industry standard" and then the instrument manufacturer falls into line.
10:09:35	 From Andrew McCluskey (he/him) : Furthermore, perhaps we should be engaging facility-like instruments. I am thinking of Brian Pauw’s Maus SAXS instrument, which is a Xenocs lab SAXS but he has worked hard to enable the CanSAS standards aspects.
10:12:54	 From Kinane, Christy (STFC,RAL,ISIS) : Rigaku have in the past asked us what we are developing so they might not be as closed a shop as we often think. Especially for new stuff. its mainly when it come to anything proprietary that we have hit obvious barriers. 
10:13:14	 From Andrew Nelson : This is also an issue for ellipsometry, where there is little open-source analysis, and little manufacturer to manufacturer interoperability.
10:19:28	 From THOMAS HASE : I am having the same fight with TXRF manufactures where they are not very willing to release the information required to perform first-principal data analysis with users often using the proprietary software to perform the analysis which returns a "quantifiable result" irrespective of whether the sample meets the requirements of the technique in the first place.  There is tension between "ease of use" and "non-expert user" with the drive for more open and transparent processing which necessarily requires the various tools to not be black box, turn-key operations...
11:07:08	 From Murphy, Dr. rer. nat. Bridget : For Tom. Do you thing  machine learning can advance the technique?
11:37:07	 From Murphy, Dr. rer. nat. Bridget : in response to Adrian: only the talks and written comments are recorded  as i understand it.
11:38:00	 From Andrew McCluskey (he/him) : @Bridget this is true
11:38:52	 From Adrian Rennie : I see the recording symbol on Zoom - that may inhibit some remarks.
11:39:50	 From Thomas Arnold : @Adrian: The working group sessions will not necessarily be recorded (depends on the session)… the intention is for free discussion and we won’t publish anything sensitive. This session we are recording the talks but stop for the discussion.
11:41:43	 From THOMAS HASE : In terms of the standard data files; it makes sense from a central facility perspective but x-ray lab based systems usually perform XRR as just one of a range of techniques so it will be difficult. You might well have issues of getting the relevant information from the manufacturers (detector windows, processing algorithms etc.).  The FAIR data principal is taking time to get into University labs....
11:44:42	 From Andrew Nelson : Straight XRR reduction from lab sources is (usually) not too difficult, at least our Panalytical is ok (omega/twotheta/intensity is provided). Resolution information is harder.
11:46:29	 From Adrian Rennie : For Andrew: often people do not know what footprint corrections are applied or if the alignment was such that the footprint correction was valid.  Sadly this is rarely propagated in data files from laboratory sources.
11:47:08	 From Andrew Nelson : True
11:48:04	 From Thomas Gutberlet : An interesting aspect here might be to go for virtual twins of the Instrument and of experiments. There all parameters have to be known properly.
11:48:11	 From THOMAS HASE : @Andrew  -  agreed, but it is difficult to have a "reflectivity" data file for anything (beamline or lab source) that is not a reflectometer beamline.  I like the idea of trying to combine the data, parameters and fitting into a single repository.
11:53:19	 From Tatsuro Oda : NIST staff develops a graphical hdf5 reader on Web browser https://github.com/usnistgov/jsfive
12:00:16	 From Andrew McCluskey (he/him) : Just to say that there will be some “required reading” for the tutorial meeting posted on Wednesday!
12:00:42	 From Andrew McCluskey (he/him) : Posted on the reflectometry.org webpages for this workshop

