From iaufwg-request Thu Mar 3 12:52:26 1994 Received: by fits.cv.nrao.edu (4.1/DDN-DLB/1.5) id AA04116; Thu, 3 Mar 94 12:52:26 EST Return-Path: Message-Id: <9403031752.AA01648@ns2.hq.eso.org> To: iaufwg@fits.CV.NRAO.EDU Subject: Vote on the three FITS proposals Date: Thu, 03 Mar 94 18:52:07 +0100 From: pgrosbol@eso.org Sender: iaufwg-request@fits.CV.NRAO.EDU Garching, Mar. 3, 1994 Dear IAU-FWG members, I have received no objections concerning the level of interchange tests we have reached. It should be added that I also read the following test files available at the NRAO archive: gbfits2.fits testdata.fits swp06542llg.fits vogtstar_awt.fits with the current release of MIDAS without problems and was able to display the array elements in them. Other reports of exchange are still very welcome! As announced, I would ask you to submit any issue cancerning the wording of the proposals before Mar. 14, 1994. It may be easier if you use the 'iaufwg@fits.CV.NRAO.EDU' mail exploder for this purpose so the points can be discussed directly! With reference to the current discussion on unsigned integers, I would like to hear if you have any special opinion. As stated, I would not support a proposal for unsigned integers. Best regards, Preben Grosbol Chairman, IAU FITS WG From iaufwg-request Thu Mar 10 11:37:02 1994 Received: by fits.cv.nrao.edu (4.1/DDN-DLB/1.5) id AA16250; Thu, 10 Mar 94 11:37:02 EST Return-Path: Message-Id: <9403101636.AA20523@ns2.hq.eso.org> To: iaufwg@fits.CV.NRAO.EDU Subject: On final text of BINTABLE proposal Date: Thu, 10 Mar 94 17:36:39 +0100 From: pgrosbol@eso.org Sender: iaufwg-request@fits.CV.NRAO.EDU Garching, Mar. 10, 1994 Dear IAU-FWG members, I have the following comments on the BINTABLE proposal (ref. Draft May 4, 1993, as available at fits.cv.nrao.edu /fits/documents/proposals/: 145877 May 18 16:29 bintable2.ps): 1) page 3, point 9, par. 2: 'The number of bytes determined .. should equal NAXIS1 but NAXIS1 should be used as the definition of the actual length of the row.' This statement is vague. It only 'suggests' that the sum of TFORM's is equal to NAXIS1 and does not state the exact rule for decoding if they are not. There are two options: A) clearly to state that they must be equal! (It should be possible to write a program which does it correct). B) maintain the option of having them different in which case a sentence must be added to specify the decoding rule e.g. the columns which storage exceed NAXIS1 are regarded as NULL columns. I would prefer option A. 2) page 4, point 2: The sentence 'The default value for EXTVER should be 1' should be added in conformance with the general rules for extensions. 3) page 5, point 12: With reference to the discussion of the usage of the keyword AUTHOR (i.e. only for the human author who compiled the data and not the program which wrote the file), a sentence of clarification should be added (or duplication from the ASCII table definition). 4) page 5, point 1: The action or meaning of other characters that T,F and NUL should be defined. The three main options: A) abort decoding - non-conforming data set! B) treat value as invalid value, C) treat value as false I would prefer option B. I recommend that the above mentioned clarifications are implemented in the final proposal if both authors and IAU-FWG members agree. Best regards, Preben Grosbol Chairman, IAU FITS WG From iaufwg-request Mon Mar 14 05:32:03 1994 Received: by fits.cv.nrao.edu (4.1/DDN-DLB/1.5) id AA08859; Mon, 14 Mar 94 05:32:03 EST Return-Path: Message-Id: <9403141031.AA03523@ns2.hq.eso.org> To: iaufwg@fits.CV.NRAO.EDU Cc: "VILSPA::JDP"@mc0.hq.eso.org Subject: On final text of IMAGE and Blocking proposals Date: Mon, 14 Mar 94 11:31:54 +0100 From: pgrosbol@eso.org Sender: iaufwg-request@fits.CV.NRAO.EDU Garching, Mar. 14, 1994 Dear IAU-FWG members, I have the following comments on the IMAGE and Blocking proposals (ref. NRAO FITS archive /fits/documents/proposals/ files 94980 Jun 4 1992 X_image.ps and 2183 Oct 1 1991 FITS_blocking90.txt): IMAGE extension: In table 1 'Mandatory FITS keywords for ... IMAGE extension' the values of PCOUNT,GCOUNT are explicitly given as 0,1, respectively. On the other hand section 3.1.5 and 3.1.6 state that they 'shall contain a non-negtive integer ..' and that 'A simple IMAGE extension will have PCOUNT=0 and GCOUNT=1' which imply possible other values for PCOUNT/GCOUNT in contradiction with the table. The two options for clarification: A) Only PCOUNT=0/GCOUNT=1 is allowed (as specified in Table 1). In this case section 3.1.5 and 3.1.6 should be reworded to state that these shall be the values. B) Any non-negative values is allowed. This would require the explicit value (0,1) in table 1 to be removed. I would prefer option A. Blocking proposal: The sentence 'Reading an incomplete FITS logical record should be regarded as an end-of-file.' in section 2 is not totally clear. It should rather be worded as 'After detection of an end-of-file on reading, any incomplete logical FITS record should be disregarded.' Please be reminded that today (Mar. 14) is the formal deadline for comments on the explicit wording of the proposals! Best regards, Preben Grosbol Chairman, IAU FITS WG From iaufwg-request Mon Mar 14 08:18:46 1994 Received: by fits.cv.nrao.edu (4.1/DDN-DLB/1.5) id AA08968; Mon, 14 Mar 94 08:18:46 EST Return-Path: Date: Mon, 14 Mar 94 08:18:43 EST From: bcotton@gorilla.CV.NRAO.EDU (Bill Cotton) Message-Id: <9403141318.AA29498@gorilla.cv.nrao.edu> To: iaufwg@fits.CV.NRAO.EDU Subject: Re: On final text of BINTABLE proposal References: <9403101636.AA20523@ns2.hq.eso.org> Sender: iaufwg-request@fits.CV.NRAO.EDU I have the following responses to Preben's comments (10 Mar 94) on the binary table draft: 1) Possible ambiguity between NAXIS1 and the sum of the TFORMs. The reason this was originally left vague was that it allowed a way to pad a row to a "magic" number of bytes in a simple way. Since there is a proper way to do this (given in Section 5, page 5) it's not really necessary to allow NAXIS1 and the sum of the byte counts from the TFORM to differ. Thus, I would agree with Preben's option A and suggest that the last sentence of point 9, page 3 read: "The number of bytes determined from summing the TFORMnnn values must equal NAXIS." 2) The default value of EXTVER should be 1. I agree. 3) Usage of AUTHOR. I agree with Preben and suggest the following sentence be added to point 12, page 5. "This is the human or organization that collected the information given in this table." 4) What do logical values other than "T", "F" or NULL mean? There shouldn't really be any values other than these three and if there are then probably something went wrong in the FITS writer. I think any other values should be disallowed and that the following be added to the end of point 1, page 5: "Any other values are illegal." (I would also agree to Preben's option A, consider other values undefined or equivalent to NULL but I think option C, any other value = false, is a bad idea.) -Bill Cotton From iaufwg-request Tue Mar 15 15:47:16 1994 Received: by fits.cv.nrao.edu (4.1/DDN-DLB/1.5) id AA13336; Tue, 15 Mar 94 15:47:16 EST Return-Path: Date: Tue, 15 Mar 94 15:47:08 EST From: pence@tetra.gsfc.nasa.gov (William Pence) Message-Id: <9403152047.AA11132@tetra.Gsfc.NASA.Gov> To: iaufwg@fits.CV.NRAO.EDU Subject: Re: On final text of BINTABLE proposal Cc: pence@tetra.gsfc.nasa.gov Sender: iaufwg-request@fits.CV.NRAO.EDU I agree with Preben's comments on the binary table draft. In particular, 1. The sum of the TFORMS must be equal to NAXIS1. 2. The default value for EXTVER should be 1 3) Author is the person or group creating the data, not the name of the program that wrote the FITS file. 4) characters other than T,F and NUL are illegal, but if this happens, the FITS reader should treat it the same as a NUL -Bill Pence From iaufwg-request Tue Mar 15 16:06:21 1994 Received: by fits.cv.nrao.edu (4.1/DDN-DLB/1.5) id AA13412; Tue, 15 Mar 94 16:06:21 EST Return-Path: Date: Tue, 15 Mar 94 16:06:06 EST From: pence@tetra.gsfc.nasa.gov (William Pence) Message-Id: <9403152106.AA11144@tetra.Gsfc.NASA.Gov> To: iaufwg@fits.CV.NRAO.EDU Subject: Re: On final text of IMAGE and Blocking proposals Sender: iaufwg-request@fits.CV.NRAO.EDU Regarding the value of PCOUNT and GCOUNT in IMAGE extensions, I have no strong preference. My FITSIO package already supports any non-negative values for these keywords, but it would be simple to restrict this to PCOUNT=0/GCOUNT=1 if that is what the community wants. -Bill Pence From iaufwg-request Tue Mar 15 16:59:00 1994 Received: by fits.cv.nrao.edu (4.1/DDN-DLB/1.5) id AA13489; Tue, 15 Mar 94 16:59:00 EST Return-Path: Date: Tue, 15 Mar 94 14:58:49 MST From: tody@noao.edu (Doug Tody) Message-Id: <9403152158.AA04736@lepus.tuc.noao.edu> To: iaufwg@fits.CV.NRAO.EDU Subject: Re: On final text of IMAGE and Blocking proposals Sender: iaufwg-request@fits.CV.NRAO.EDU Preben, I have reviewed your points (10Mar94) and the current bintable draft, as well as the comments already posted. I agree with the comments already made, so it is unanimous. None of these points represent any changes to the proposal, rather they address ambiguities in the draft, and they are all worthwhile. - Doug For reference, the four points were: 1) Require that NAXIS1 be the sum of the TFORMs (no implicit padding, explicit padding remains possible). 2) Extension version (EXTVER) should be 1. 3) Clarify use of AUTHOR keyword (this is not restricted to BINTABLE). 4) Values other than `T' `F' or NUL for a logical are illegal.