From fitsbits-request Sun Dec 3 10:46:15 1995 X-VM-Message-Order: (1 24 3 26 2 21 11 12 16 27 4 25 17 20 18 8 22 9 10 15 14 13 19 30 29 5 31 23 28 6 7 33 32 34 35 36 37 38) X-VM-Summary-Format: "%n %*%a %-17.17F %-3.3m %2d %4l/%-5c %I\"%s\"\n" X-VM-Labels: nil X-VM-VHeader: ("Resent-" "From:" "Sender:" "To:" "Apparently-To:" "Cc:" "Subject:" "Date:") nil X-VM-Bookmark: 38 X-VM-v5-Data: ([nil nil nil nil nil nil nil nil nil] ["1329" "" "26" "November" "1995" "00:07:04" "GMT" "Vaclav Pribik" "emorse at csn.net" "<498b38$q1c at news-2.csn.net>" "36" "Re: Data Compression" "^From:" nil nil "11" "1995112600:07:04" "Data Compression" (number " " mark " Vaclav Pribik Nov 26 36/1329 " thread-indent "\"Re: Data Compression\"\n") "<48g69o$plk at pa.mother.com>"] nil) Received: by fits.cv.nrao.edu (5.x/S2.3/NRAO-CV/2.4) id AA13777; Sun, 3 Dec 1995 10:46:15 -0500 Return-Path: Message-Id: <498b38$q1c at news-2.csn.net> Organization: HAF Path: solitaire.cv.nrao.edu!hearst.acc.Virginia.EDU!portal.gmu.edu!europa.chnt.gtegsc.com!news.kreonet.re.kr!overload.lbl.gov!lll-winken.llnl.gov!simtel!chi-news.cic.net!newsfeed.internetmci.com!in1.uu.net!csn!nntp-xfer-2.csn.net!news-2.csn.net!usenet References: <48g69o$plk at pa.mother.com> Newsgroups: sci.astro.fits content-length: 1327 From: Vaclav Pribik Sender: fitsbits-request at fits.cv.nrao.edu To: fitsbits at fits.cv.nrao.edu Subject: Re: Data Compression Date: 26 Nov 1995 00:07:04 GMT Data compression is useful a mos on Internet (smaller down-uploading times). There are (I think...) two different ways: 1. Use uncompressed FITS and compress it (the best should be comrpess it bu Compress (.Z), UNIX is still standart). Then, after downloading, uncompressed again and easy-to-work FITS is avaible. Problem: Software depended 2. Use directly some compression method: RLE (like IFF, Tiff) or LZW (like GIF89a ) - no quality lost or Frequence chart (like JPEG) - some quality lost, but high-compressive. The result-size is not (= only a little) depending on size. (I tried it - image 1280x1280 pix was 1.5 bigger than the some image, only 320x320 pix.). +: No SW depend Problem: Harder to process. The first method is easier (everyone has, or can have, Compress), but it is not so compressive than method 2. Maybe it will be useful to make new extension using method 2, but it should be used only for large images (ex. 16777216x8388608 pixels :-) ), that is suitable for JPEG. Only the optimisation should be low = most quality. And the format should be also modified - JPEG is for COLOR images and don't like gray-scaled. Vaclav Pribik email now: emorse at csn.net normal email, use from 7/96: pribik at leo1.zlin.vutbr.cz astro home page: http://www.csn.net/~emorse/haf/haf.html sci.astro.fits From fitsbits-request Mon Dec 4 20:01:47 1995 X-VM-v5-Data: ([nil nil nil nil nil nil nil nil nil] ["1760" "" "4" "December" "1995" "13:20:15" "-0000" "Clive Page" "cgp at le.ac.uk" "<49usif$2ab at hawk.le.ac.uk>" "38" "Re: Formal fits grammar (was: Re: Are HEX integers allowed in header?)" "^From:" nil nil "12" "1995120413:20:15" "Formal fits grammar (was: Re: Are HEX integers allowed in header?)" (number " " mark " Clive Page Dec 4 38/1760 " thread-indent "\"Re: Formal fits grammar (was: Re: Are HEX integers allowed in header?)\"\n") ""] nil) Received: by fits.cv.nrao.edu (5.x/S2.3/NRAO-CV/2.4) id AA16622; Mon, 4 Dec 1995 20:01:47 -0500 Return-Path: Message-Id: <49usif$2ab at hawk.le.ac.uk> Organization: University of Leicester, UK Path: solitaire.cv.nrao.edu!hearst.acc.Virginia.EDU!caen!usenet.cis.ufl.edu!usenet.eel.ufl.edu!newsfeed.internetmci.com!in1.uu.net!EU.net!peer-news.britain.eu.net!warwick!leicester!leicester!not-for-mail References: Newsgroups: sci.astro.fits content-length: 1758 From: Clive Page Sender: fitsbits-request at fits.cv.nrao.edu To: fitsbits at fits.cv.nrao.edu Subject: Re: Formal fits grammar (was: Re: Are HEX integers allowed in header?) Date: 4 Dec 1995 13:20:15 -0000 In article , John E. Davis wrote: > etc.... > >I really believe that the fits definition should go into such detail. For >me, there appear to be too many ambiguities in the fits definition. Such a >grammar would resolve these issues. I agree with you. On this particular point of hex constants: it must surely be an unintentional lack of rigour that allows this, as surely only decimal numbers were only intended here. The problem has been that the desire to say "once a FITS file, always a FITS file" (which I've heard Don Wells declaim more than once). This means that no rule should be introduced which invalidates past data files. Because FITS has long been used for archiving data this is in general a sensible rule, but perhaps it can be taken too far. I came upon a similar problem not long ago when making a parser for a table browser. I hoped it would always be possible to distinguish column-names from constants so the user could select rows using an ordinary arithmetic expression with column-names, constants, and operators (similar to FSELECT in FTOOLS). But the FITS standard allows column-names to consist of all digits (and start with a minus sign come to that). Now hardly anyone uses names like that, as it just seems natural to a programmer to start a name with a letter. But you can if you want give your column a name like "42" (and it will break my browser). It would be nice if the FITS Standard disallowed such things, but it may be too late now. -- ------------------------------------------------------------------------- Clive Page, Internet: cgp at star.le.ac.uk Dept of Physics & Astronomy, University of Leicester. From fitsbits-request Tue Dec 5 01:41:10 1995 X-VM-v5-Data: ([nil nil nil nil nil nil nil nil nil] ["1190" "" "30" "November" "1995" "20:59:36" "GMT" "John E. Davis" "davis at space.mit.edu" "" "32" "fits ambiguities" "^From:" nil nil "11" "1995113020:59:36" "fits ambiguities" (number " " mark " John E. Davis Nov 30 32/1190 " thread-indent "\"fits ambiguities\"\n") nil] nil) Received: by fits.cv.nrao.edu (5.x/S2.3/NRAO-CV/2.4) id AA16747; Tue, 5 Dec 1995 01:41:10 -0500 Return-Path: Message-Id: Organization: Center for Space Research Path: solitaire.cv.nrao.edu!hearst.acc.Virginia.EDU!maui.cc.odu.edu!news.larc.nasa.gov!news.msfc.nasa.gov!newsfeed.internetmci.com!bloom-beacon.mit.edu!senator-bedfellow.mit.edu!davis Reply-To: davis at space.mit.edu Newsgroups: sci.astro.fits content-length: 1188 From: davis at space.mit.edu (John E. Davis) Sender: fitsbits-request at fits.cv.nrao.edu To: fitsbits at fits.cv.nrao.edu Subject: fits ambiguities Date: 30 Nov 1995 20:59:36 GMT Hi, In a previous article I mentioned that there are some ambiguities in the fits definition and indicated a need for a formal grammar for fits keywords. I believe that such problems need to be addressed before programs that convert fits to other data formats can be constructed. Here is an example of one ambiguity: 1 2 3 4 5 6 7 1234567890123456789012345678901234567890123456789012345678901234567890 AMBIG1 = 10 is more common AMBIG2 = 1 0 is more common AMBIG3 = 1 10 is less common As far as I can tell, these lines are permitted by the fits standard although they are strongly discouraged. Nevertheless, it is not clear how they should be interpreted. AMBIG1 is a keyword whose value is 1 with comment `0 is more common'. There is no ambiguity here. AMBIG2 is either equivalent to AMBIG1 or represents a complex integer. AMBIG3 is equally problematic. How should one resolve these ambiguities? Again, issues like these have to be dealt with before software will be able to interpret such values. Thanks, --John From fitsbits-request Tue Dec 5 01:52:25 1995 X-VM-v5-Data: ([nil nil nil nil nil nil nil nil nil] ["572" "" "4" "December" "1995" "09:28" "EDT" "Barry M. Schlesinger" "bschlesinger at nssdca.gsfc.nasa.gov" "<4DEC199509284270 at nssdca.gsfc.nasa.gov>" "21" "Re: Simple Header Clarification" "^From:" nil nil "12" "1995120413:28:00" "Simple Header Clarification" (number " " mark " Barry M. Schlesin Dec 4 21/572 " thread-indent "\"Re: Simple Header Clarification\"\n") ""] nil) Received: by fits.cv.nrao.edu (5.x/S2.3/NRAO-CV/2.4) id AA16764; Tue, 5 Dec 1995 01:52:25 -0500 Return-Path: Message-Id: <4DEC199509284270 at nssdca.gsfc.nasa.gov> Organization: NASA Goddard Space Flight Center -- Greenbelt, Maryland USA Path: solitaire.cv.nrao.edu!hearst.acc.Virginia.EDU!caen!uwm.edu!lll-winken.llnl.gov!ames!newsfeed.gsfc.nasa.gov!nssdca.gsfc.nasa.gov!bschlesinger References: Newsgroups: sci.astro.fits content-length: 570 From: bschlesinger at nssdca.gsfc.nasa.gov (Barry M. Schlesinger) Sender: fitsbits-request at fits.cv.nrao.edu To: fitsbits at fits.cv.nrao.edu Subject: Re: Simple Header Clarification Date: 4 Dec 1995 09:28 EDT In article , kkuehl at ih4gp743.ih.att.com (Kevin Kuehl) writes... > ... Can headers (either primary or extension) be >composed of multiple 2880 byte records or only one? In >some places it seems to say they can and in some places >they can't. > In section 4.2 # The primary HDU and every extension HDU shall consist of an integral # number of header records.... which allows more than one record in a header. I am not aware of any text that limits headers to one record. Barry Schlesinger FITS Support Office GSFC/ADF From fitsbits-request Tue Dec 5 02:56:22 1995 X-VM-v5-Data: ([nil nil nil nil nil nil nil nil nil] ["2194" "" "30" "November" "1995" "16:55:22" "GMT" "John E. Davis" "davis at space.mit.edu" "" "46" "Formal fits grammar (was: Re: Are HEX integers allowed in header?)" "^From:" nil nil "11" "1995113016:55:22" "Formal fits grammar (was: Re: Are HEX integers allowed in header?)" (number " " mark " John E. Davis Nov 30 46/2194 " thread-indent "\"Formal fits grammar (was: Re: Are HEX integers allowed in header?)\"\n") nil] nil) Received: by fits.cv.nrao.edu (5.x/S2.3/NRAO-CV/2.4) id AA16793; Tue, 5 Dec 1995 02:56:22 -0500 Return-Path: Message-Id: Organization: Center for Space Research Path: solitaire.cv.nrao.edu!hearst.acc.Virginia.EDU!maui.cc.odu.edu!news.larc.nasa.gov!lerc.nasa.gov!magnus.acs.ohio-state.edu!math.ohio-state.edu!howland.reston.ans.net!swrinde!newsfeed.internetmci.com!bloom-beacon.mit.edu!senator-bedfellow.mit.edu!davis Reply-To: davis at space.mit.edu Newsgroups: sci.astro.fits content-length: 2192 From: davis at space.mit.edu (John E. Davis) Sender: fitsbits-request at fits.cv.nrao.edu To: fitsbits at fits.cv.nrao.edu Subject: Formal fits grammar (was: Re: Are HEX integers allowed in header?) Date: 30 Nov 1995 16:55:22 GMT On 30 Nov 1995 09:09 EDT, Barry M. Schlesinger wrote: : explicit and differentiated from integers. This history supports the : interpretation of integer keyword values as decimal integers. I think that the NOST fits definition paper should explicitly point out that the integers should be expressed in decimal format. As I recall, it only says something like ``ASCII representation''. `7F' (hex) is certainly a valid ASCII representation of `127' (decimal). Perhaps many will think that I am going overboard by saying that I would like the NOST fits definition paper to be quite formal on these issues. For example, look at any RFC, e.g., RFC822 which presents the format for email headers. In it you will find: D. ALPHABETICAL LISTING OF SYNTAX RULES address = mailbox ; one addressee / group ; named list addr-spec = local-part " at " domain ; global address ALPHA = ; (101-132, 65.- 90.) ; (141-172, 97.-122.) atom = 1* authentic = "From" ":" mailbox ; Single author / ( "Sender" ":" mailbox ; Actual submittor "From" ":" 1#mailbox) ; Multiple authors ; or not sender CHAR = ; ( 0-177, 0.-127.) comment = "(" *(ctext / quoted-pair / comment) ")" CR = ; ( 15, 13.) CRLF = CR LF ctext = may be folded ")", "\" & CR, & including linear-white-space> CTL = " "42" "Re: fits ambiguities" "^From:" nil nil "12" "1995120414:59:30" "fits ambiguities" (number " " mark " William Thompson Dec 4 42/1855 " thread-indent "\"Re: fits ambiguities\"\n") ""] nil) Received: by fits.cv.nrao.edu (5.x/S2.3/NRAO-CV/2.4) id AA16825; Tue, 5 Dec 1995 04:49:04 -0500 Return-Path: Message-Id: <49v2ci$jss at post.gsfc.nasa.gov> Organization: NASA Goddard Space Flight Center -- Greenbelt, Maryland USA Path: solitaire.cv.nrao.edu!hearst.acc.Virginia.EDU!caen!uwm.edu!lll-winken.llnl.gov!enews.sgi.com!ames!newsfeed.gsfc.nasa.gov!orpheus.nascom.nasa.gov!thompson References: <49nosb$t5u at post.gsfc.nasa.gov> Newsgroups: sci.astro.fits content-length: 1853 From: thompson at orpheus.nascom.nasa.gov (William Thompson) Sender: fitsbits-request at fits.cv.nrao.edu To: fitsbits at fits.cv.nrao.edu Subject: Re: fits ambiguities Date: 4 Dec 1995 14:59:30 GMT davis at space.mit.edu (John E. Davis) writes: >On 1 Dec 1995 20:34:19 GMT, William Thompson >wrote: > : davis at space.mit.edu (John E. Davis) writes: > : > Here is an example of one ambiguity: > : > > : > 1 2 3 4 5 6 7 > : >1234567890123456789012345678901234567890123456789012345678901234567890 > : >AMBIG1 = 10 is more common > : >AMBIG2 = 1 0 is more common > : >AMBIG3 = 1 10 is less common > : > : >As far as I can tell, these lines are permitted by the fits standard > : >although they are strongly discouraged. Nevertheless, it is not clear how > : >they should be interpreted. > : > : >AMBIG1 is a keyword whose value is 1 with comment `0 is more common'. There > : >is no ambiguity here. > : ^^ > : > : I disagree. This is also ambiguous. It could be interpreted as a character > : string with the quotes left off--many poorly designed FITS writers do this. > : The only completely non-ambiguous form would be > : > : 1234567890123456789012345678901234567890123456789012345678901234567890 > : AMBIG0 = 1 /0 is more common >I can use your argument to argue that this is also ambiguous. There >should be no excuse for fits writers to omit the quotes. If a ``fits'' file >does not follow the syntax that the fits standard requires, then it is not a >fits file. Such a file should have set the boolean flag `SIMPLE' to `F'. Yes, you could argue that if one is allowed to have strings without quotes, then everything is ambiguous. I can't really argue with you there. Mainly, I was complaining about leaving the / off to mark the comment. I consider the / delimiter to be as essential as the quote signs. Bill Thompson From fitsbits-request Tue Dec 5 04:52:04 1995 X-VM-v5-Data: ([nil nil nil nil nil nil nil nil nil] ["1079" "" "4" "December" "1995" "18:22:21" "GMT" "Jesse B. Doggett" "doggett at stsci.edu" "<49ve8t$akq at marvel.stsci.edu>" "38" "Re: fits ambiguities" "^From:" nil nil "12" "1995120418:22:21" "fits ambiguities" (number " " mark " Jesse B. Doggett Dec 4 38/1079 " thread-indent "\"Re: fits ambiguities\"\n") "<49v3k0$edm at news.ccit.arizona.edu>"] nil) Received: by fits.cv.nrao.edu (5.x/S2.3/NRAO-CV/2.4) id AA16835; Tue, 5 Dec 1995 04:52:04 -0500 Return-Path: Message-Id: <49ve8t$akq at marvel.stsci.edu> Organization: Space Telescope Science Institute Path: solitaire.cv.nrao.edu!hearst.acc.Virginia.EDU!maui.cc.odu.edu!beacon.regent.edu!imci3!imci2!newsfeed.internetmci.com!swrinde!cs.utexas.edu!news.eas.asu.edu!noao!stsci!doggett References: <49nosb$t5u at post.gsfc.nasa.gov> <49v3k0$edm at news.ccit.arizona.edu> Newsgroups: sci.astro.fits content-length: 1077 From: doggett at stsci.edu (Jesse B. Doggett) Sender: fitsbits-request at fits.cv.nrao.edu To: fitsbits at fits.cv.nrao.edu Subject: Re: fits ambiguities Date: 4 Dec 1995 18:22:21 GMT > This brings up one of my "favorite" problems in the FITS definitions > which I've read. Suppose one wants the equivalent of: > > Name: O'Malley > > As I've read the standards there are two possibilities: > > NAME = 'O''Malley ' > NAME = 'O'Malley ' > > (Pardon if I misscount spaces, my point is the arrangment of "'".) > For me, the first is ambigious, since it could be: Name: O''Malley > The second is not ambigious, since the ending quote must be at column > 20 or beyond. Of course, I don't know how many FITS readers would > agree with me. The first is not actually ambiguous. Name: O''Malley would be in the header as: NAME = 'O''''Malley ' The second would fail because the string might include "'" beyond column 20, eg, NAME: 1234567890123456789012345678910'ABCD would then presumably be put in the header as: NAME = '123456789012345678901234567890'ABCD' But, this would parse as the string '123456789012345678901234567890' with comment ABCD'. It should be: NAME = '123456789012345678901234567890''ABCD' From fitsbits-request Wed Dec 6 04:17:48 1995 X-VM-v5-Data: ([nil nil nil nil nil nil nil nil nil] ["1114" "" "4" "December" "1995" "23:45:30" "GMT" "Anthony J. Ferro" "tferro at merlin.la.asu.edu" "<4a016q$pe6 at news.ccit.arizona.edu>" "33" "Re: fits ambiguities" "^From:" nil nil "12" "1995120423:45:30" "fits ambiguities" (number " " mark " Anthony J. Ferro Dec 4 33/1114 " thread-indent "\"Re: fits ambiguities\"\n") "<49ve8t$akq at marvel.stsci.edu>"] nil) Received: by fits.cv.nrao.edu (5.x/S2.3/NRAO-CV/2.4) id AA18484; Wed, 6 Dec 1995 04:17:48 -0500 Return-Path: Message-Id: <4a016q$pe6 at news.ccit.arizona.edu> Organization: The University of Arizona Path: solitaire.cv.nrao.edu!hearst.acc.Virginia.EDU!caen!news.umass.edu!umassd.edu!ulowell.uml.edu!wang!news.kei.com!newsfeed.internetmci.com!vixen.cso.uiuc.edu!uchinews!ncar!noao!math.arizona.edu!news.Arizona.EDU!merlin!tferro References: <49nosb$t5u at post.gsfc.nasa.gov> <49v3k0$edm at news.ccit.arizona.edu> <49ve8t$akq at marvel.stsci.edu> Newsgroups: sci.astro.fits content-length: 1112 From: tferro at merlin.la.asu.edu (Anthony J. Ferro) Sender: fitsbits-request at fits.cv.nrao.edu To: fitsbits at fits.cv.nrao.edu Subject: Re: fits ambiguities Date: 4 Dec 1995 23:45:30 GMT Jesse B. Doggett (doggett at stsci.edu) wrote: : The second would fail because the string might include "'" beyond : column 20, eg, : NAME: 1234567890123456789012345678910'ABCD : would then presumably be put in the header as: I agree that after column 20, the double-single quote, "''", is required as is the Fortran standard. Between columns 11 and 20, however, it is not clear what is meant. Any quotes in this range cannot be the end-of-string marker, so why should they be doubled? The standards say, as I recall, that one should follow the Fortran standard. To me this breaks down in the case of short strings. All Fortrans I know of allow short strings but FITS doesn't. (Some versions do require that they use a full word of space, however.) I haven't reviewed the notes, but does "''" count as two characters or one? This might be important in a case such as NAME: 1'A - Tony --------- Anthony J. Ferro aferro at as.arizona.edu Steward Observatory, NICMOS Project Ph: (520) 621-8683 University of Arizona FAX: (520) 621-1891 Tucson, AZ 85721 --------- From fitsbits-request Wed Dec 6 05:14:02 1995 X-VM-v5-Data: ([nil nil nil nil nil nil nil nil nil] ["919" "Tue" "5" "December" "1995" "13:08:55" "GMT" "Kevin Kuehl" "kkuehl at ih4gp743.ih.att.com" "" "24" "Re: fits ambiguities" "^From:" nil nil "12" "1995120513:08:55" "fits ambiguities" (number " " mark " Kevin Kuehl Dec 5 24/919 " thread-indent "\"Re: fits ambiguities\"\n") "<49vmr9$ptb at post.gsfc.nasa.gov>"] nil) Received: by fits.cv.nrao.edu (5.x/S2.3/NRAO-CV/2.4) id AA19248; Wed, 6 Dec 1995 05:14:02 -0500 Return-Path: Message-Id: Organization: AT&T Path: solitaire.cv.nrao.edu!hearst.acc.Virginia.EDU!caen!newsxfer.itd.umich.edu!tank.news.pipex.net!pipex!news.mathworks.com!newsfeed.internetmci.com!chi-news.cic.net!news.midplains.net!gw2.att.com!nntpa!ssbunews!not-for-mail References: <49nosb$t5u at post.gsfc.nasa.gov> <49vmr9$ptb at post.gsfc.nasa.gov> Newsgroups: sci.astro.fits content-length: 917 From: kkuehl at ih4gp743.ih.att.com (Kevin Kuehl) Sender: fitsbits-request at fits.cv.nrao.edu To: fitsbits at fits.cv.nrao.edu Subject: Re: fits ambiguities Date: Tue, 5 Dec 1995 13:08:55 GMT In article <49vmr9$ptb at post.gsfc.nasa.gov>, William Thompson wrote: >Yes, but in section 5.3.1 it states > > The fixed format is required for values of mandatory keywords and > recommended for values of all others. > ^^^^^^^^^^^ > >The "recommended" is an example of the ambiguity that the FITS standard as >written allows, and which the original author was (rightly) complaining about. >I certainly don't condone or defend such practise--note that I said "poorly >dessign"--but I've seen it done. I guess I skipped over that part because I couldn't believe that it meant what you are saying. I am not saying you are wrong, I just couldn't believe that the standard said that a FITS writer did not always have to adhere to the value definitions. How are readers supposed to read FITS files if writers don't have to adhere to the standard? Yuck! -- Kevin kevin.r.kuehl at att.com From fitsbits-request Wed Dec 6 10:08:57 1995 X-VM-v5-Data: ([nil nil nil nil nil nil nil nil nil] ["1257" "Wed" "6" "December" "1995" "10:08:51" "-0500" "Eric Greisen" "egreisen at nrao.edu" "<9512061508.AA25155 at primate.cv.nrao.edu>" "29" "FITS \"ambiguities\"" "^From:" nil nil "12" "1995120615:08:51" "FITS \"ambiguities\"" (number " " mark " Eric Greisen Dec 6 29/1257 " thread-indent "\"FITS \"ambiguities\"\"\n") nil] nil) Received: by fits.cv.nrao.edu (5.x/S2.3/NRAO-CV/2.4) id AA00513; Wed, 6 Dec 1995 10:08:57 -0500 Return-Path: Message-Id: <9512061508.AA25155 at primate.cv.nrao.edu> content-length: 1255 From: egreisen at NRAO.EDU (Eric Greisen) Sender: fitsbits-request at fits.cv.nrao.edu To: fitsbits at primate.cv.nrao.edu Subject: FITS "ambiguities" Date: Wed, 6 Dec 1995 10:08:51 -0500 Reply to: kkuehl at ih4gp743.ih.att.com Are you suggesting in your recent posting that all FITS keywords should be required to have some sort of fixed format no matter what they actually mean and require? We thought long and hard about requiring fixed format for ANY keywords and decided that the very small number of required keywords - all logicals and modest integers - would have a fixed format to help out stupid reading programs. All others are expected to follow Fortran free-format rules which are not ambiguous at all. There is no requirement that string keywords be of length >= 8, there is a requirement that they follow Fortran rules. Thus an embedded ' is represented as '' and the string data must be surrounded by ' marks. Earlier examples such as KEYWORD = 0 1 a sample are just erroneous. If a comment is desired, it must be preceded by a / . KEYWORD can be either a 1-valued or a 2-valued numeric parameter, but it cannot be both a numeric and string array (violates Fortran rules and was not listed as a possibility by us) and, besides, the a sample above is not a Fortran string because it is not surrounded by quotes. Why are we trying to invent ambiguities where there are none? Eric Greisen, Scientist, NRAO From fitsbits-request Wed Dec 6 16:08:35 1995 X-VM-v5-Data: ([nil nil nil nil nil nil nil nil nil] ["5046" "Wed" "6" "December" "1995" "16:06:42" "-0500" "William Pence" "pence at tetra.gsfc.nasa.gov" "<199512062106.QAA18781 at tetra.gsfc.nasa.gov>" "104" "Re: fits ambiguities" "^From:" nil nil "12" "1995120621:06:42" "fits ambiguities" (number " " mark " William Pence Dec 6 104/5046 " thread-indent "\"Re: fits ambiguities\"\n") nil] nil) Received: by fits.cv.nrao.edu (5.x/S2.3/NRAO-CV/2.4) id AA03431; Wed, 6 Dec 1995 16:08:35 -0500 Return-Path: Message-Id: <199512062106.QAA18781 at tetra.gsfc.nasa.gov> content-length: 5044 From: William Pence Sender: fitsbits-request at fits.cv.nrao.edu To: fitsbits at fits.cv.nrao.edu Cc: pence at tetra.gsfc.nasa.gov Subject: Re: fits ambiguities Date: Wed, 6 Dec 1995 16:06:42 -0500 On the subject of FITS keyword 'ambiguities' it might be of interest to give a short explaination of how the FITSIO software package (http://heasarc.gsfc.nasa.gov/docs/software /fitsio/fitsio.html) interpretes header keywords. I believe this interpretation is consistent with all the rules defined in the FITS standard. While the FITS standard could perhaps be made clearer, I don't see any significant ambiguities in how to interpret FITS keywords (except perhaps for complex keywords as illustrated below). Except for the few required FITS keywords (SIMPLE, BITPIX, NAXIS, etc) which have a fixed format integer or logical value, all other keyword values are interpreted following the Fortran free-format rules. First, the keyword value field is defined to be the first 'token' in the FITS 80-byte keyword record following the '= ' characters in column 9 and 10. (If there is no '= ' in columns 9 and 10, then by definition the keyword has no value field). The value token is defined to be either the first consecutive string of non-blank ASCII text characters, or if the first character is a single quote, then as a string of ASCII text characters (including any spaces) enclosed within the first pair of single quotes. The data type and value of this token is then determined from the following rules: 1. If the token begins with a single quote character then the value is a character string. The string extends to the next single quote character in the value token, except that 2 consecutive single quote characters are interpreted as a literal single quote character within the string. The rules for a fixed format character string require that the leading quote character be in column 11 and the closing quote character be no earlier than column 20, but a free format string value is not constrained by this. Leading spaces in the quoted string are significant, but trailing spaces are not. 2. If the token consists of a single upper case T or F character then the value is interpreted as a true or false logical value. 3. If the token contains only digits (0 - 9) then the value is interpreted as a decimal integer. 4. If none of the above then the token must be a floating point numerical value, either in fixed decimal format, or exponential format. If the token does not conform to one of these 4 rules, then the keyword value is invalid and it does not conform to the FITS standard. Any remaining characters following the value token are generally interpreted as a comment field. If the first non-blank character following the value token is a slash (`/`) character then this is interpreted as a separator between the value and comment fields. Alternatively, instead of a comment field, there may be another numerical value field following the first numerical value token. Keywords like this with 2 numeric tokens can be interpreted as a complex number with the first token representing the real part and the second token representing the imaginary part. As I understand the FITS standard, this interpretation is not required, so it is up to the reader to decide whether the second number should be interpreted as the imaginary part of the complex number, or as a comment to the real floating point number. The last example below illustrates this ambiguity. In practice I have never seen a FITS file that uses the complex keyword format. Finally, a few examples. The first example in each group follows the recommended rules for a fixed format keyword. These are all legal FITS string keywords with identical value and comment fields: 1234567890123456789012345678901234567890 STRING1 = 'DOG ' / It's an animal STRING2 = 'DOG ' / It's an animal STRING3 = 'DOG ' It's an animal STRING4 = 'DOG'It's an animal Notice in the STRING4 example, even when the / and space characters are omitted, (which is obviously NOT recommended) there is still no ambiguity between the value and comment fields. These are all legal FITS integer keywords with identical value and comment fields: 1234567890123456789012345678901234567890 INTEGER1= -273 / absolute zero INTEGER2= -273 / absolute zero INTEGER3= -273 absolute zero These are all legal FITS logical keywords with identical value and comment fields: 1234567890123456789012345678901234567890 LOGICAL1= T / a true value LOGICAL2= T /a true value LOGICAL3= T a true value These are legal FITS floating point keywords with identical values and comments: 1234567890123456789012345678901234567890 FLOAT1 = 12345.67890123 / 10 times greater than expected FLOAT2 = 1.234567890123D+04 / 10 times greater than expected FLOAT3 = 1234.567890123 10 times greater than expected This last example is actually ambiguous; it can either be interpreted as a real floating point number with the comment '10 times greater than expected' or as a complex number with the value (1234.567890123, 10.) and a comment 'times greater than expected'. Bill Pence HEASARC/GSFC/NASA From fitsbits-request Thu Dec 7 10:38:11 1995 X-VM-v5-Data: ([nil nil nil nil nil nil nil nil nil] ["289" "Sat" "2" "December" "1995" "14:59:18" "GMT" "Kevin Kuehl" "kkuehl at ih4gp743.ih.att.com" "" "12" "Simple Header Clarification" "^From:" nil nil "12" "1995120214:59:18" "Simple Header Clarification" (number " " mark " Kevin Kuehl Dec 2 12/289 " thread-indent "\"Simple Header Clarification\"\n") nil] nil) Received: by fits.cv.nrao.edu (5.x/S2.3/NRAO-CV/2.4) id AA13775; Thu, 7 Dec 1995 10:38:11 -0500 Return-Path: Message-Id: Organization: AT&T Path: solitaire.cv.nrao.edu!hearst.acc.Virginia.EDU!portal.gmu.edu!newsfeed.internetmci.com!vixen.cso.uiuc.edu!howland.reston.ans.net!nntp.crl.com!pacbell.com!gw2.att.com!nntpa!ssbunews!not-for-mail Newsgroups: sci.astro.fits content-length: 287 From: kkuehl at ih4gp743.ih.att.com (Kevin Kuehl) Sender: fitsbits-request at fits.cv.nrao.edu To: fitsbits at fits.cv.nrao.edu Subject: Simple Header Clarification Date: Sat, 2 Dec 1995 14:59:18 GMT In reading the FITs documentation from NOST I am a little confused. Can headers (either primary or extension) be composed of multiple 2880 byte records or only one? In some places it seems to say they can and in some places they can't. Thanks a bunch, -- Kevin kevin.r.kuehl at att.com From fitsbits-request Thu Dec 7 15:32:54 1995 X-VM-v5-Data: ([nil nil nil nil nil nil nil nil nil] ["2534" "" "5" "December" "1995" "20:50:32" "GMT" "Rob Seaman" "seaman at noao.edu" "<4a2bao$pk9 at noao.edu>" "55" "Re: fits ambiguities" "^From:" nil nil "12" "1995120520:50:32" "fits ambiguities" (number " " mark " Rob Seaman Dec 5 55/2534 " thread-indent "\"Re: fits ambiguities\"\n") ""] nil) Received: by fits.cv.nrao.edu (5.x/S2.3/NRAO-CV/2.4) id AA17476; Thu, 7 Dec 1995 15:32:54 -0500 Return-Path: Message-Id: <4a2bao$pk9 at noao.edu> Organization: National Optical Astronomy Observatories, Tucson, AZ, USA Path: solitaire.cv.nrao.edu!hearst.acc.Virginia.EDU!caen!night.primate.wisc.edu!nntp.msstate.edu!gatech2!news.eas.asu.edu!noao!seaman References: <49nosb$t5u at post.gsfc.nasa.gov> Newsgroups: sci.astro.fits content-length: 2532 From: seaman at noao.edu (Rob Seaman) Sender: fitsbits-request at fits.cv.nrao.edu To: fitsbits at fits.cv.nrao.edu Subject: Re: fits ambiguities Date: 5 Dec 1995 20:50:32 GMT > Yes, but in section 5.3.1 it states > > The fixed format is required for values of mandatory keywords and > recommended for values of all others. The two sentences before this one are: "Unless otherwise specified, the value field must be written in a notation consistent with the list-directed read operations in ANSI FORTRAN-77. The structure shall be determined by the type of the variable." The only explicit exceptions "otherwise specified" in section 5.3 are complex integer and complex floating point valued keywords, which omit the parentheses and separating comma that are required by FORTRAN-77. Complex numbers are indeed ambiguously represented in FITS keywords - assuming the user and the software aren't expecting to read a complex number, and are accepting of nonsensical keyword comments. On the other hand, FORTRAN-77 list-directed reads require that strings be single quoted and that embedded quotes (apostrophes) be represented by two consecutive quotes. Therefore, non-quoted FITS keyword values cannot legally be interpreted as strings. To the question of whether FITS integers are required to be decimal, I don't see any ambiguity at all. Barry's original reply: > In the FITS papers, reference to integer format, when unmodified, is > generally understood to mean decimal integers. In the original FITS > paper, integer format is discussed in the context of FORTRAN, whose > standard integers are decimal. [...] just isn't strong enough. FITS keyword usage isn't only "discussed" in the context of FORTRAN - it is defined in that context. List directed FORTRAN-77 input requires that the form of the input be acceptable for a constant of the given type (ANSI X3.9-1978, section 13.6.1). A FORTRAN-77 integer constant is "interpreted as a decimal number" (ANSI X3.9-1978, section 4.3.1). It wouldn't hurt to emphasize this in the FITS documents, though. FITS is certainly informally specified in computer science terms. It does, however, rely on the most standard scientific computing reference that was available at the time, namely ANSI X3.9-1978 FORTRAN-77. It would be interesting, though, to see someone reconstruct a real grammar for FITS - surely there's FITS software relying on a formal parser that could be unraveled for the job. (Anybody? It would make a good ADASS paper for next year.) Rob Seaman -- seaman at noao.edu, http://iraf.noao.edu/~seaman NOAO, 950 N Cherry Ave, Tucson AZ 85719, 520-318-8248 PGP: 98 8D 8B 49 74 9A 41 88 3A 43 87 54 51 BF 30 4B From fitsbits-request Fri Dec 8 06:05:55 1995 X-VM-v5-Data: ([nil nil nil nil nil nil nil nil nil] ["1708" "" "3" "December" "1995" "17:32:32" "GMT" "John E. Davis" "davis at space.mit.edu" "" "39" "Re: fits ambiguities" "^From:" nil nil "12" "1995120317:32:32" "fits ambiguities" (number " " mark " John E. Davis Dec 3 39/1708 " thread-indent "\"Re: fits ambiguities\"\n") "<49nosb$t5u at post.gsfc.nasa.gov>"] nil) Received: by fits.cv.nrao.edu (5.x/S2.3/NRAO-CV/2.4) id AA25471; Fri, 8 Dec 1995 06:05:55 -0500 Return-Path: Message-Id: Organization: Center for Space Research Path: solitaire.cv.nrao.edu!hearst.acc.Virginia.EDU!portal.gmu.edu!newsfeed.internetmci.com!bloom-beacon.mit.edu!senator-bedfellow.mit.edu!davis References: <49nosb$t5u at post.gsfc.nasa.gov> Reply-To: davis at space.mit.edu Newsgroups: sci.astro.fits content-length: 1706 From: davis at space.mit.edu (John E. Davis) Sender: fitsbits-request at fits.cv.nrao.edu To: fitsbits at fits.cv.nrao.edu Subject: Re: fits ambiguities Date: 3 Dec 1995 17:32:32 GMT On 1 Dec 1995 20:34:19 GMT, William Thompson wrote: : davis at space.mit.edu (John E. Davis) writes: : > Here is an example of one ambiguity: : > : > 1 2 3 4 5 6 7 : >1234567890123456789012345678901234567890123456789012345678901234567890 : >AMBIG1 = 10 is more common : >AMBIG2 = 1 0 is more common : >AMBIG3 = 1 10 is less common : : >As far as I can tell, these lines are permitted by the fits standard : >although they are strongly discouraged. Nevertheless, it is not clear how : >they should be interpreted. : : >AMBIG1 is a keyword whose value is 1 with comment `0 is more common'. There : >is no ambiguity here. : ^^ : : I disagree. This is also ambiguous. It could be interpreted as a character : string with the quotes left off--many poorly designed FITS writers do this. : The only completely non-ambiguous form would be : : 1234567890123456789012345678901234567890123456789012345678901234567890 : AMBIG0 = 1 /0 is more common I can use your argument to argue that this is also ambiguous. There should be no excuse for fits writers to omit the quotes. If a ``fits'' file does not follow the syntax that the fits standard requires, then it is not a fits file. Such a file should have set the boolean flag `SIMPLE' to `F'. I am more concerned about the informal nature of the NOST fits definition. The descriptions should supplemented by something more formal and the ambiguities that I described above should be resolved by the standard. --John From fitsbits-request Fri Dec 8 07:41:13 1995 X-VM-v5-Data: ([nil nil nil nil nil nil nil nil nil] ["437" "" "7" "December" "1995" "02:46:27" "GMT" "Immanuel Freedman" "immanuel at mother.com" "<4a5ki3$mag at pa.mother.com>" "14" "Re: fits ambiguities" "^From:" nil nil "12" "1995120702:46:27" "fits ambiguities" (number " " mark " Immanuel Freedman Dec 7 14/437 " thread-indent "\"Re: fits ambiguities\"\n") "<4a2bao$pk9 at noao.edu>"] nil) Received: by fits.cv.nrao.edu (5.x/S2.3/NRAO-CV/2.4) id AA26296; Fri, 8 Dec 1995 07:41:13 -0500 Return-Path: Message-Id: <4a5ki3$mag at pa.mother.com> Organization: Mother.COM Path: solitaire.cv.nrao.edu!hearst.acc.Virginia.EDU!caen!uwm.edu!homer.alpha.net!news.jersey.net!ns2.mainstreet.net!news.PBI.net!pa.mother.com!news References: <49nosb$t5u at post.gsfc.nasa.gov> <4a2bao$pk9 at noao.edu> Newsgroups: sci.astro.fits content-length: 435 From: "Dr. Immanuel Freedman" Sender: fitsbits-request at fits.cv.nrao.edu To: fitsbits at fits.cv.nrao.edu Subject: Re: fits ambiguities Date: 7 Dec 1995 02:46:27 GMT Rob The COBE project built an "awk"-based parser/converter to convert VAX DDL & COBE RDL record definitions to FITS. The 2-step process involved creating a FITS header from the Record Definition Compiler output & using FITSIO to perform the conversion based on that header. Almost 1000 record structures were converted. If, interested, please contact Don West or John Mather at NASA Code 680, Dr. Immanuel Freedman (619) 658 0236 From fitsbits-request Fri Dec 8 09:24:25 1995 X-VM-v5-Data: ([nil nil nil nil nil nil nil nil nil] ["1332" "" "1" "December" "1995" "20:34:19" "GMT" "William Thompson" "thompson at orpheus.nascom.nasa.gov" "<49nosb$t5u at post.gsfc.nasa.gov>" "35" "Re: fits ambiguities" "^From:" nil nil "12" "1995120120:34:19" "fits ambiguities" (number " " mark " William Thompson Dec 1 35/1332 " thread-indent "\"Re: fits ambiguities\"\n") ""] nil) Received: by fits.cv.nrao.edu (5.x/S2.3/NRAO-CV/2.4) id AA27023; Fri, 8 Dec 1995 09:24:25 -0500 Return-Path: Message-Id: <49nosb$t5u at post.gsfc.nasa.gov> Organization: NASA Goddard Space Flight Center -- Greenbelt, Maryland USA Path: solitaire.cv.nrao.edu!hearst.acc.Virginia.EDU!news-server.ncren.net!taco.cc.ncsu.edu!gatech!newsfeed.internetmci.com!swrinde!sgigate.sgi.com!enews.sgi.com!ames!newsfeed.gsfc.nasa.gov!orpheus.nascom.nasa.gov!thompson References: Newsgroups: sci.astro.fits content-length: 1330 From: thompson at orpheus.nascom.nasa.gov (William Thompson) Sender: fitsbits-request at fits.cv.nrao.edu To: fitsbits at fits.cv.nrao.edu Subject: Re: fits ambiguities Date: 1 Dec 1995 20:34:19 GMT davis at space.mit.edu (John E. Davis) writes: >Hi, > In a previous article I mentioned that there are some ambiguities in the >fits definition and indicated a need for a formal grammar for fits keywords. >I believe that such problems need to be addressed before programs that >convert fits to other data formats can be constructed. > Here is an example of one ambiguity: > > 1 2 3 4 5 6 7 >1234567890123456789012345678901234567890123456789012345678901234567890 >AMBIG1 = 10 is more common >AMBIG2 = 1 0 is more common >AMBIG3 = 1 10 is less common >As far as I can tell, these lines are permitted by the fits standard >although they are strongly discouraged. Nevertheless, it is not clear how >they should be interpreted. >AMBIG1 is a keyword whose value is 1 with comment `0 is more common'. There >is no ambiguity here. ^^ I disagree. This is also ambiguous. It could be interpreted as a character string with the quotes left off--many poorly designed FITS writers do this. The only completely non-ambiguous form would be 1234567890123456789012345678901234567890123456789012345678901234567890 AMBIG0 = 1 /0 is more common Bill Thompson From fitsbits-request Fri Dec 8 12:26:43 1995 X-VM-v5-Data: ([nil nil nil nil nil nil nil nil nil] ["540" "Sun" "3" "December" "1995" "23:29:13" "GMT" "Kevin Kuehl" "kkuehl at ih4gp743.ih.att.com" "" "15" "Re: fits ambiguities" "^From:" nil nil "12" "1995120323:29:13" "fits ambiguities" (number " " mark " Kevin Kuehl Dec 3 15/540 " thread-indent "\"Re: fits ambiguities\"\n") ""] nil) Received: by fits.cv.nrao.edu (5.x/S2.3/NRAO-CV/2.4) id AA28644; Fri, 8 Dec 1995 12:26:43 -0500 Return-Path: Message-Id: Organization: AT&T Path: solitaire.cv.nrao.edu!hearst.acc.Virginia.EDU!portal.gmu.edu!newsfeed.internetmci.com!chi-news.cic.net!news.midplains.net!gw2.att.com!nntpa!ssbunews!not-for-mail References: <49nosb$t5u at post.gsfc.nasa.gov> Newsgroups: sci.astro.fits content-length: 538 From: kkuehl at ih4gp743.ih.att.com (Kevin Kuehl) Sender: fitsbits-request at fits.cv.nrao.edu To: fitsbits at fits.cv.nrao.edu Subject: Re: fits ambiguities Date: Sun, 3 Dec 1995 23:29:13 GMT In article , John E. Davis wrote: >I am more concerned about the informal nature of the NOST fits definition. >The descriptions should supplemented by something more formal and the >ambiguities that I described above should be resolved by the standard. I agree. I am new to FITs and I expected to see some grammar or other formalisms in the FITs definitions but I see only English text. Being from a CS background this gives me the creeps. :-) -- Kevin kevin.r.kuehl at att.com From fitsbits-request Fri Dec 8 12:26:41 1995 X-VM-v5-Data: ([nil nil nil nil nil nil nil nil nil] ["1652" "Sun" "3" "December" "1995" "23:26:44" "GMT" "Kevin Kuehl" "kkuehl at ih4gp743.ih.att.com" "" "40" "Re: fits ambiguities" "^From:" nil nil "12" "1995120323:26:44" "fits ambiguities" (number " " mark " Kevin Kuehl Dec 3 40/1652 " thread-indent "\"Re: fits ambiguities\"\n") "<49nosb$t5u at post.gsfc.nasa.gov>"] nil) Received: by fits.cv.nrao.edu (5.x/S2.3/NRAO-CV/2.4) id AA28642; Fri, 8 Dec 1995 12:26:41 -0500 Return-Path: Message-Id: Organization: AT&T Path: solitaire.cv.nrao.edu!hearst.acc.Virginia.EDU!portal.gmu.edu!newsfeed.internetmci.com!chi-news.cic.net!news.midplains.net!gw2.att.com!nntpa!ssbunews!not-for-mail References: <49nosb$t5u at post.gsfc.nasa.gov> Newsgroups: sci.astro.fits content-length: 1650 From: kkuehl at ih4gp743.ih.att.com (Kevin Kuehl) Sender: fitsbits-request at fits.cv.nrao.edu To: fitsbits at fits.cv.nrao.edu Subject: Re: fits ambiguities Date: Sun, 3 Dec 1995 23:26:44 GMT In article <49nosb$t5u at post.gsfc.nasa.gov>, William Thompson wrote: >davis at space.mit.edu (John E. Davis) writes: >> >> 1 2 3 4 5 6 7 >>1234567890123456789012345678901234567890123456789012345678901234567890 >>AMBIG1 = 10 is more common >>AMBIG2 = 1 0 is more common >>AMBIG3 = 1 10 is less common > >>As far as I can tell, these lines are permitted by the fits standard >>although they are strongly discouraged. Nevertheless, it is not clear how >>they should be interpreted. > >>AMBIG1 is a keyword whose value is 1 with comment `0 is more common'. There >>is no ambiguity here. > ^^ > >I disagree. This is also ambiguous. It could be interpreted as a character >string with the quotes left off--many poorly designed FITS writers do this. I don't think it can be interpreted as a character string. I have "Definition of the Flexible Image Transport System (FITS)" NOST 100-1.1 dated September 29, 1995 right here in front of me. In section 5.3.2.1 Character String, it says If the value is a character string, column 11 shall contain a single quote (hexadecimal code 27, "'"); the string shall follow, starting in column 12, followed by a closing single quote (also hexadecimal 27) that should not occur before column 20 and must occur in or before column 80. From this, I don't see how it could be properly interpreted as a string. While I am pretty new to FITS, the definition seems pretty clearly written. -- Kevin kevin.r.kuehl at att.com From fitsbits-request Sat Dec 9 06:43:52 1995 X-VM-v5-Data: ([nil nil nil nil nil nil nil nil nil] ["1491" "" "4" "December" "1995" "15:20:32" "GMT" "Anthony J. Ferro" "tferro at merlin.la.asu.edu" "<49v3k0$edm at news.ccit.arizona.edu>" "42" "Re: fits ambiguities" "^From:" nil nil "12" "1995120415:20:32" "fits ambiguities" (number " " mark " Anthony J. Ferro Dec 4 42/1491 " thread-indent "\"Re: fits ambiguities\"\n") ""] nil) Received: by fits.cv.nrao.edu (5.x/S2.3/NRAO-CV/2.4) id AA07674; Sat, 9 Dec 1995 06:43:52 -0500 Return-Path: Message-Id: <49v3k0$edm at news.ccit.arizona.edu> Organization: The University of Arizona Path: solitaire.cv.nrao.edu!hearst.acc.Virginia.EDU!maui.cc.odu.edu!news.larc.nasa.gov!lerc.nasa.gov!magnus.acs.ohio-state.edu!math.ohio-state.edu!howland.reston.ans.net!swrinde!newsfeed.internetmci.com!in1.uu.net!nntp.primenet.com!news.asu.edu!news.eas.asu.edu!noao!math.arizona.edu!news.Arizona.EDU!merlin!tferro References: <49nosb$t5u at post.gsfc.nasa.gov> Newsgroups: sci.astro.fits content-length: 1489 From: tferro at merlin.la.asu.edu (Anthony J. Ferro) Sender: fitsbits-request at fits.cv.nrao.edu To: fitsbits at fits.cv.nrao.edu Subject: Re: fits ambiguities Date: 4 Dec 1995 15:20:32 GMT Kevin Kuehl (kkuehl at ih4gp743.ih.att.com) wrote: : I don't think it can be interpreted as a character string. I have : "Definition of the Flexible Image Transport System (FITS)" NOST 100-1.1 : dated September 29, 1995 right here in front of me. In section : 5.3.2.1 Character String, it says : If the value is a character string, column 11 shall contain : a single quote (hexadecimal code 27, "'"); the string shall : follow, starting in column 12, followed by a closing single : quote (also hexadecimal 27) that should not occur before : column 20 and must occur in or before column 80. : From this, I don't see how it could be properly interpreted as a : string. While I am pretty new to FITS, the definition seems pretty : clearly written. : -- This brings up one of my "favorite" problems in the FITS definitions which I've read. Suppose one wants the equivalent of: Name: O'Malley As I've read the standards there are two possibilities: NAME = 'O''Malley ' NAME = 'O'Malley ' (Pardon if I misscount spaces, my point is the arrangment of "'".) For me, the first is ambigious, since it could be: Name: O''Malley The second is not ambigious, since the ending quote must be at column 20 or beyond. Of course, I don't know how many FITS readers would agree with me. --------- Anthony J. Ferro aferro at as.arizona.edu Steward Observatory, NICMOS Project Ph: (520) 621-8683 University of Arizona FAX: (520) 621-1891 Tucson, AZ 85721 --------- From fitsbits-request Sat Dec 9 09:50:45 1995 X-VM-v5-Data: ([nil nil nil nil nil nil nil nil nil] ["2935" "" "4" "December" "1995" "20:48:41" "GMT" "William Thompson" "thompson at achilles.nascom.nasa.gov" "<49vmr9$ptb at post.gsfc.nasa.gov>" "72" "Re: fits ambiguities" "^From:" nil nil "12" "1995120420:48:41" "fits ambiguities" (number " " mark " William Thompson Dec 4 72/2935 " thread-indent "\"Re: fits ambiguities\"\n") ""] nil) Received: by fits.cv.nrao.edu (5.x/S2.3/NRAO-CV/2.4) id AA09185; Sat, 9 Dec 1995 09:50:45 -0500 Return-Path: Message-Id: <49vmr9$ptb at post.gsfc.nasa.gov> Organization: NASA Goddard Space Flight Center -- Greenbelt, Maryland USA Path: solitaire.cv.nrao.edu!hearst.acc.Virginia.EDU!portal.gmu.edu!newsfeed.internetmci.com!news.kei.com!nntp.coast.net!lll-winken.llnl.gov!ames!newsfeed.gsfc.nasa.gov!achilles.nascom.nasa.gov!thompson References: <49nosb$t5u at post.gsfc.nasa.gov> Newsgroups: sci.astro.fits content-length: 2933 From: thompson at achilles.nascom.nasa.gov (William Thompson) Sender: fitsbits-request at fits.cv.nrao.edu To: fitsbits at fits.cv.nrao.edu Subject: Re: fits ambiguities Date: 4 Dec 1995 20:48:41 GMT kkuehl at ih4gp743.ih.att.com (Kevin Kuehl) writes: >In article <49nosb$t5u at post.gsfc.nasa.gov>, >William Thompson wrote: >>davis at space.mit.edu (John E. Davis) writes: >>> >>> 1 2 3 4 5 6 7 >>>1234567890123456789012345678901234567890123456789012345678901234567890 >>>AMBIG1 = 10 is more common >>>AMBIG2 = 1 0 is more common >>>AMBIG3 = 1 10 is less common >> >>>As far as I can tell, these lines are permitted by the fits standard >>>although they are strongly discouraged. Nevertheless, it is not clear how >>>they should be interpreted. >> >>>AMBIG1 is a keyword whose value is 1 with comment `0 is more common'. There >>>is no ambiguity here. >> ^^ >> >>I disagree. This is also ambiguous. It could be interpreted as a character >>string with the quotes left off--many poorly designed FITS writers do this. >I don't think it can be interpreted as a character string. I have >"Definition of the Flexible Image Transport System (FITS)" NOST 100-1.1 >dated September 29, 1995 right here in front of me. In section >5.3.2.1 Character String, it says > If the value is a character string, column 11 shall contain > a single quote (hexadecimal code 27, "'"); the string shall > follow, starting in column 12, followed by a closing single > quote (also hexadecimal 27) that should not occur before > column 20 and must occur in or before column 80. >From this, I don't see how it could be properly interpreted as a >string. While I am pretty new to FITS, the definition seems pretty >clearly written. Yes, but in section 5.3.1 it states The fixed format is required for values of mandatory keywords and recommended for values of all others. ^^^^^^^^^^^ The "recommended" is an example of the ambiguity that the FITS standard as written allows, and which the original author was (rightly) complaining about. I certainly don't condone or defend such practise--note that I said "poorly dessign"--but I've seen it done. In section 5.1.2.3 the standard also states Separation of the value and comments by a slash (hexadecimal 2F, "/"), and a space between the value and the slash are strongly recommended. (Another example of the ambiguity that the standard allows.) It was this that I was bringing into the debate. The standard would allow AMBIG1 = 10 is more common to be interpreted as either a character string or as a value and a comment, depending on whether you believe that the quotes or the / was left out, but I don't consider it as clear as AMBIG0 = 1/ 0 is more common If the argument is that "once FITS, always FITS" doesn't allow us to get rid of such ambiguities, then the answer is to use the deprecation mechanism already applied to random groups. Bill Thompson From fitsbits-request Sat Dec 9 11:51:49 1995 X-VM-v5-Data: ([nil nil nil nil nil nil nil nil nil] ["1430" "" "7" "December" "1995" "17:34:14" "GMT" "John E. Davis" "davis at space.mit.edu" "" "38" "Re: fits ambiguities" "^From:" nil nil "12" "1995120717:34:14" "fits ambiguities" (number " " mark " John E. Davis Dec 7 38/1430 " thread-indent "\"Re: fits ambiguities\"\n") "<49v3k0$edm at news.ccit.arizona.edu>"] nil) Received: by fits.cv.nrao.edu (5.x/S2.3/NRAO-CV/2.4) id AA10038; Sat, 9 Dec 1995 11:51:49 -0500 Return-Path: Message-Id: Organization: Center for Space Research Path: solitaire.cv.nrao.edu!hearst.acc.Virginia.EDU!caen!hookup!news.mathworks.com!news.kei.com!bloom-beacon.mit.edu!senator-bedfellow.mit.edu!davis References: <49nosb$t5u at post.gsfc.nasa.gov> <49v3k0$edm at news.ccit.arizona.edu> Reply-To: davis at space.mit.edu Newsgroups: sci.astro.fits content-length: 1428 From: davis at space.mit.edu (John E. Davis) Sender: fitsbits-request at fits.cv.nrao.edu To: fitsbits at fits.cv.nrao.edu Subject: Re: fits ambiguities Date: 7 Dec 1995 17:34:14 GMT On 4 Dec 1995 15:20:32 GMT, Anthony J. Ferro wrote: : As I've read the standards there are two possibilities: : : NAME = 'O''Malley ' : NAME = 'O'Malley ' : : (Pardon if I misscount spaces, my point is the arrangment of "'".) : For me, the first is ambigious, since it could be: Name: O''Malley : The second is not ambigious, since the ending quote must be at column : 20 or beyond. Of course, I don't know how many FITS readers would : agree with me. If you want O''Malley, you have to use O''''Malley. I believe that the fits definition paper is quite clear on this (Section 5.3.2.1). However, that paragraph confuses me with the statement: ``Reading data files in a fits file should not require decoding any more than the first eight characters of a character string value of a keyword.'' What exactly does this mean? My interpretation is that it implies that only the first eight characters are meaningful. I hope I am wrong about this because with my interpretation this fits card: RQFIL1 = './sisdata/s0c3p40e0/gau.s0.c3.g?.ecd' / 1st file -- branching ratio is worthless. I suspect that the statement ``once fits, always fits'' really means ``once a fits reader, always a fits reader''. -- John E. Davis Center for Space Research/AXAF Science Center 617-258-8119 MIT 37-662c, Cambridge, MA 02139 From fitsbits-request Sun Dec 10 13:49:38 1995 X-VM-v5-Data: ([nil nil nil nil nil nil nil nil nil] ["923" "" "5" "December" "1995" "09:22" "EDT" "Barry M. Schlesinger" "bschlesinger at nssdca.gsfc.nasa.gov" "<5DEC199509221915 at nssdca.gsfc.nasa.gov>" "29" "Re: fits ambiguities" "^From:" nil nil "12" "1995120513:22:00" "fits ambiguities" (number " " mark " Barry M. Schlesin Dec 5 29/923 " thread-indent "\"Re: fits ambiguities\"\n") "<49v3k0$edm at news.ccit.arizona.edu>"] nil) Received: by fits.cv.nrao.edu (5.x/S2.3/NRAO-CV/2.4) id AA04872; Sun, 10 Dec 1995 13:49:38 -0500 Return-Path: Message-Id: <5DEC199509221915 at nssdca.gsfc.nasa.gov> Organization: NASA Goddard Space Flight Center -- Greenbelt, Maryland USA Path: solitaire.cv.nrao.edu!hearst.acc.Virginia.EDU!portal.gmu.edu!newsfeed.internetmci.com!news.kei.com!nntp.coast.net!lll-winken.llnl.gov!ames!newsfeed.gsfc.nasa.gov!nssdca.gsfc.nasa.gov!bschlesinger References: <49nosb$t5u at post.gsfc.nasa.gov> <49v3k0$edm at news.ccit.arizona.edu> Newsgroups: sci.astro.fits content-length: 921 From: bschlesinger at nssdca.gsfc.nasa.gov (Barry M. Schlesinger) Sender: fitsbits-request at fits.cv.nrao.edu To: fitsbits at fits.cv.nrao.edu Subject: Re: fits ambiguities Date: 5 Dec 1995 09:22 EDT In article <49v3k0$edm at news.ccit.arizona.edu>, tferro at merlin.la.asu.edu (Anthony J. Ferro) writes... > This brings up one of my "favorite" problems in the FITS definitions > which I've read. Suppose one wants the equivalent of: > > Name: O'Malley > > As I've read the standards there are two possibilities: > > NAME = 'O''Malley ' > NAME = 'O'Malley ' > > (Pardon if I misscount spaces, my point is the arrangment of "'".) > For me, the first is ambigious, since it could be: Name: O''Malley > The second is not ambigious, since the ending quote must be at column > 20 or beyond. Of course, I don't know how many FITS readers would > agree with me. > From the Definition of FITS (NOST), section 5.3.2.1 (Character String) # A single quote is represented within a string as two successive # single quotes, e. g., O'HARA = {\tt 'O''HARA'}. Barry Schlesinger FITS Support Office GSFC/ADF From fitsbits-request Sun Dec 10 20:52:37 1995 X-VM-v5-Data: ([nil nil nil nil nil nil nil nil nil] ["1370" "" "7" "December" "1995" "03:44:20" "GMT" "John E. Davis" "davis at space.mit.edu" "" "43" "Re: fits ambiguities" "^From:" nil nil "12" "1995120703:44:20" "fits ambiguities" (number " " mark " John E. Davis Dec 7 43/1370 " thread-indent "\"Re: fits ambiguities\"\n") ""] nil) Received: by fits.cv.nrao.edu (5.x/S2.3/NRAO-CV/2.4) id AA05422; Sun, 10 Dec 1995 20:52:37 -0500 Return-Path: Message-Id: Organization: Center for Space Research Path: solitaire.cv.nrao.edu!hearst.acc.Virginia.EDU!maui.cc.odu.edu!news.larc.nasa.gov!night.primate.wisc.edu!newsspool.doit.wisc.edu!chi-news.cic.net!newsfeed.internetmci.com!bloom-beacon.mit.edu!senator-bedfellow.mit.edu!davis References: <49nosb$t5u at post.gsfc.nasa.gov> <49vmr9$ptb at post.gsfc.nasa.gov> Reply-To: davis at space.mit.edu Newsgroups: sci.astro.fits content-length: 1368 From: davis at space.mit.edu (John E. Davis) Sender: fitsbits-request at fits.cv.nrao.edu To: fitsbits at fits.cv.nrao.edu Subject: Re: fits ambiguities Date: 7 Dec 1995 03:44:20 GMT On Tue, 5 Dec 1995 13:08:55 GMT, Kevin Kuehl wrote: : >Yes, but in section 5.3.1 it states : > : > The fixed format is required for values of mandatory keywords and : > recommended for values of all others. : > ^^^^^^^^^^^ [...] : I guess I skipped over that part because I couldn't believe : that it meant what you are saying. I am not saying you are : wrong, I just couldn't believe that the standard said that : a FITS writer did not always have to adhere to the value : definitions. How are readers supposed to read FITS files : if writers don't have to adhere to the standard? Yuck! I honestly believe that many people treat fits files as glorified binary files where one already ``knows'' the format, e.g., skip_header (); skip_header (); while (NOT END OF FILE) { skip_32_bytes (); read_4_bytes (); skip_32_bytes (); read_4_bytes (); } Of course the ``reader'' needs modified when the format changes. For this reason fits files that do not follow standard conventions for all keywords are not uncommon. To paraphrase Forest Gump, ``Fits is is like a box of chocolates-- you never know what you're gonna get.'' -- John E. Davis Center for Space Research/AXAF Science Center 617-258-8119 MIT 37-662c, Cambridge, MA 02139 From fitsbits-request Mon Dec 11 17:07:36 1995 X-VM-v5-Data: ([nil nil nil nil nil nil nil nil nil] ["753" "" "9" "December" "1995" "20:40:08" "GMT" "John E. Davis" "davis at space.mit.edu" "" "18" "Re: FITS \"ambiguities\"" "^From:" nil nil "12" "1995120920:40:08" "FITS \"ambiguities\"" (number " " mark " John E. Davis Dec 9 18/753 " thread-indent "\"Re: FITS \"ambiguities\"\"\n") ""] nil) Received: by fits.cv.nrao.edu (5.x/S2.3/NRAO-CV/2.4) id AA01688; Mon, 11 Dec 1995 17:07:36 -0500 Return-Path: Message-Id: Organization: Center for Space Research Path: solitaire.cv.nrao.edu!hearst.acc.Virginia.EDU!uunet!in1.uu.net!news.mathworks.com!news.kei.com!bloom-beacon.mit.edu!senator-bedfellow.mit.edu!davis References: <9512061508.AA25155 at primate.cv.nrao.edu> Reply-To: davis at space.mit.edu Newsgroups: sci.astro.fits content-length: 751 From: davis at space.mit.edu (John E. Davis) Sender: fitsbits-request at fits.cv.nrao.edu To: fitsbits at fits.cv.nrao.edu Subject: Re: FITS "ambiguities" Date: 9 Dec 1995 20:40:08 GMT On Thu, 7 Dec 1995 15:07:12 GMT, Kevin Kuehl wrote: : But I think I have a better understanding of the intent of FITS : based on e-mail people have sent me. Basically it sounds like FITS : is intended to have a human help the computer out and it is not : intended to be automatically readable like GIF or JPEG. This is also the impression that I am getting. I would like to regard a fits file as a well-defined standalone object. However, it seems to only be well-defined within the context of a specific reader. This is not what I would call ``self-describing''. -- John E. Davis Center for Space Research/AXAF Science Center 617-258-8119 MIT 37-662c, Cambridge, MA 02139 From fitsbits-request Tue Dec 12 18:34:29 1995 X-VM-v5-Data: ([nil nil nil nil nil nil nil nil nil] ["1231" "Thu" "7" "December" "1995" "15:07:12" "GMT" "Kevin Kuehl" "kkuehl at ih4gp743.ih.att.com" "" "28" "Re: FITS \"ambiguities\"" "^From:" nil nil "12" "1995120715:07:12" "FITS \"ambiguities\"" (number " " mark " Kevin Kuehl Dec 7 28/1231 " thread-indent "\"Re: FITS \"ambiguities\"\"\n") "<9512061508.AA25155 at primate.cv.nrao.edu>"] nil) Received: by fits.cv.nrao.edu (5.x/S2.3/NRAO-CV/2.4) id AA05943; Tue, 12 Dec 1995 18:34:29 -0500 Return-Path: Message-Id: Organization: AT&T Path: solitaire.cv.nrao.edu!hearst.acc.Virginia.EDU!portal.gmu.edu!newsfeed.internetmci.com!news.akorn.net!news.his.com!news.cais.net!news.sprintlink.net!howland.reston.ans.net!vixen.cso.uiuc.edu!uwm.edu!fnnews.fnal.gov!gw1.att.com!nntpa!ssbunews!not-for-mail References: <9512061508.AA25155 at primate.cv.nrao.edu> Newsgroups: sci.astro.fits content-length: 1229 From: kkuehl at ih4gp743.ih.att.com (Kevin Kuehl) Sender: fitsbits-request at fits.cv.nrao.edu To: fitsbits at fits.cv.nrao.edu Subject: Re: FITS "ambiguities" Date: Thu, 7 Dec 1995 15:07:12 GMT In article <9512061508.AA25155 at primate.cv.nrao.edu>, Eric Greisen wrote: > >Reply to: kkuehl at ih4gp743.ih.att.com > >Are you suggesting in your recent posting that all FITS keywords >should be required to have some sort of fixed format no matter what >they actually mean and require? We thought long and hard about >requiring fixed format for ANY keywords and decided that the very >small number of required keywords - all logicals and modest integers - >would have a fixed format to help out stupid reading programs. What I am suggesting is that card images of the form 'keyword = value' should have a "keyword" field that follows a set format and a "value" field that follows a set format. I was under the impression that "value"s had to follow the definitions in the spec, but people have said that if the "keyword" string is not predefined then whatever is in the "value" area does not have to follow the definition. But I think I have a better understanding of the intent of FITS based on e-mail people have sent me. Basically it sounds like FITS is intended to have a human help the computer out and it is not intended to be automatically readable like GIF or JPEG. -- Kevin kevin.r.kuehl at att.com From fitsbits-request Tue Dec 12 19:18:26 1995 X-VM-v5-Data: ([nil nil nil nil nil nil nil nil nil] ["1946" "Thu" "7" "December" "1995" "15:34:20" "GMT" "Kevin Kuehl" "kkuehl at ih4gp743.ih.att.com" "" "43" "Re: fits ambiguities" "^From:" nil nil "12" "1995120715:34:20" "fits ambiguities" (number " " mark " Kevin Kuehl Dec 7 43/1946 " thread-indent "\"Re: fits ambiguities\"\n") "<199512062106.QAA18781 at tetra.gsfc.nasa.gov>"] nil) Received: by fits.cv.nrao.edu (5.x/S2.3/NRAO-CV/2.4) id AA06007; Tue, 12 Dec 1995 19:18:26 -0500 Return-Path: Message-Id: Organization: AT&T Path: solitaire.cv.nrao.edu!hearst.acc.Virginia.EDU!portal.gmu.edu!newsfeed.internetmci.com!news.compuserve.com!chi-news.cic.net!uwm.edu!fnnews.fnal.gov!gw1.att.com!nntpa!ssbunews!not-for-mail References: <199512062106.QAA18781 at tetra.gsfc.nasa.gov> Newsgroups: sci.astro.fits content-length: 1944 From: kkuehl at ih4gp743.ih.att.com (Kevin Kuehl) Sender: fitsbits-request at fits.cv.nrao.edu To: fitsbits at fits.cv.nrao.edu Subject: Re: fits ambiguities Date: Thu, 7 Dec 1995 15:34:20 GMT In article <199512062106.QAA18781 at tetra.gsfc.nasa.gov>, William Pence wrote: >On the subject of FITS keyword 'ambiguities' it might be of interest >to give a short explaination of how the FITSIO software package >(http://heasarc.gsfc.nasa.gov/docs/software /fitsio/fitsio.html) >interpretes header keywords. I believe this interpretation is >consistent with all the rules defined in the FITS standard. While the >FITS standard could perhaps be made clearer, I don't see any >significant ambiguities in how to interpret FITS keywords (except >perhaps for complex keywords as illustrated below). Thanks for posting this. You cleared up a lot of things that were confusing me (it doesn't sound like I was the only one either). Here is the root of my confusion: Unless otherwise specified, the value field must be written in a notation consistent with the list-directed read operations in ANSI FORTRAN-77. The structure shall be determined by the type of the variable. The fixed format is required for values of mandatory keywords and recommended for values of all others. This standard imposes no requirements on case sensitivity of character strings other than those explicitly specified. I understand now that "list-directed read operations" is something different from the fixed format. Part of the problem may be is that you are now getting a whole generation of people (of which I am part of) in this area who know nothing about FORTRAN-77. Saying it follows FORTRAN standards means as much to me as saying the fields must be in EBCDIC. The only thing I know about either is that they are part of the history of computing. :-) Luckily your post tells me what "list-directed read operations" means so that I can understand it. I still think a formal grammar in the definition would help FITS' younger users (as well as others it sounds like from the posts). -- Kevin kevin.r.kuehl at att.com From fitsbits-request Wed Dec 13 06:33:30 1995 X-VM-v5-Data: ([nil nil nil nil nil nil nil nil nil] ["460" "Thu" "7" "December" "1995" "22:53:37" "CST" "U39732 at uicvm.uic.edu" "U39732 at uicvm.uic.edu" "<95341.225337U39732 at uicvm.uic.edu>" "13" "DSS Viewers for Windows" "^From:" nil nil "12" "1995120804:53:37" "DSS Viewers for Windows" (number " " mark " U39732 at uicvm.uic. Dec 7 13/460 " thread-indent "\"DSS Viewers for Windows\"\n") nil] nil) Received: by fits.cv.nrao.edu (5.x/S2.3/NRAO-CV/2.4) id AA07671; Wed, 13 Dec 1995 06:33:30 -0500 Return-Path: Message-Id: <95341.225337U39732 at uicvm.uic.edu> Organization: University of Illinois at Chicago, ADN Computer Center Path: solitaire.cv.nrao.edu!hearst.acc.Virginia.EDU!portal.gmu.edu!newsfeed.internetmci.com!news.kei.com!nntp.coast.net!chi-news.cic.net!news.luc.edu!newsfeed.acns.nwu.edu!uicvm.uic.edu!u39732 Newsgroups: sci.astro.fits content-length: 458 From: Sender: fitsbits-request at fits.cv.nrao.edu To: fitsbits at fits.cv.nrao.edu Subject: DSS Viewers for Windows Date: Thu, 7 Dec 1995 22:53:37 CST Can anyone point me to a FITS viewer for Windows that understands header on Digitized Sky Survey files from STScI? I currently use FITS View for Windows by Bill Cotton on VLA files and I can find weak sources more easily than with AIPS TV tools (perhaps because the intensity and contrast controls are discrete), but it can't deal with the coordinate keywords in DSS files. Is there a Windows viewer that can, or a filter for DSS files? Thanks! Bob Gray From fitsbits-request Thu Dec 14 10:00:36 1995 X-VM-v5-Data: ([nil nil nil nil nil nil nil nil nil] ["977" "" "11" "December" "1995" "04:05:07" "GMT" "Gene Heskett" "gene_heskett at wvlink.mpl.com" "<974.6552T1375T1465 at wvlink.mpl.com>" "26" "Re: fits ambiguities" "^From:" nil nil "12" "1995121104:05:07" "fits ambiguities" (number " " mark " Gene Heskett Dec 11 26/977 " thread-indent "\"Re: fits ambiguities\"\n") ""] nil) Received: by fits.cv.nrao.edu (5.x/S2.3/NRAO-CV/2.4) id AA10347; Thu, 14 Dec 1995 10:00:36 -0500 Return-Path: Message-Id: <974.6552T1375T1465 at wvlink.mpl.com> Organization: MountainNet, Inc. Morgantown WV 800.444.1458 Path: solitaire.cv.nrao.edu!hearst.acc.Virginia.EDU!uunet!in1.uu.net!salliemae!newsfeed.internetmci.com!lamarck.sura.net!news.mountain.net!usenet References: Newsgroups: sci.astro.fits content-length: 975 From: gene_heskett at wvlink.mpl.com (Gene Heskett) Sender: fitsbits-request at fits.cv.nrao.edu To: fitsbits at fits.cv.nrao.edu Subject: Re: fits ambiguities Date: 11 Dec 1995 04:05:07 GMT KK> In article , KK> John E. Davis wrote: >>I am more concerned about the informal nature of the NOST fits >>definition. The descriptions should supplemented by something more formal >>and the ambiguities that I described above should be resolved by the >>standard. KK> I agree. I am new to FITs and I expected to see some grammar KK> or other formalisms in the FITs definitions but I see only KK> English text. Being from a CS background this gives me the KK> creeps. :-) KK> -- KK> Kevin KK> kevin.r.kuehl at att.com Have I not saw this name before, on a FalSoft Publishing (Rainbow) article? We won't discuss how many years back up the log that was though. /* Gene Heskett | These opinions are NOT to be */ /* CE at WDTV Weston/Clarksburg WV | confused with the official */ /* | WDTV managment views */ #include From fitsbits-request Thu Dec 14 10:12:40 1995 X-VM-v5-Data: ([nil nil nil nil nil nil nil nil nil] ["645" "" "8" "December" "1995" "17:59:20" "GMT" "John E. Davis" "davis at space.mit.edu" "<4a9udo$ppn at senator-bedfellow.MIT.EDU>" "19" "Re: FITS \"ambiguities\"" "^From:" nil nil "12" "1995120817:59:20" "FITS \"ambiguities\"" (number " " mark " John E. Davis Dec 8 19/645 " thread-indent "\"Re: FITS \"ambiguities\"\"\n") "<9512061508.AA25155 at primate.cv.nrao.edu>"] nil) Received: by fits.cv.nrao.edu (5.x/S2.3/NRAO-CV/2.4) id AA10385; Thu, 14 Dec 1995 10:12:40 -0500 Return-Path: Message-Id: <4a9udo$ppn at senator-bedfellow.MIT.EDU> Organization: Center for Space Research Path: solitaire.cv.nrao.edu!hearst.acc.Virginia.EDU!portal.gmu.edu!europa.chnt.gtegsc.com!news.kreonet.re.kr!news.dacom.co.kr!newsfeed.internetmci.com!news.kei.com!bloom-beacon.mit.edu!senator-bedfellow.mit.edu!davis References: <9512061508.AA25155 at primate.cv.nrao.edu> Reply-To: davis at space.mit.edu Newsgroups: sci.astro.fits content-length: 643 From: davis at space.mit.edu (John E. Davis) Sender: fitsbits-request at fits.cv.nrao.edu To: fitsbits at fits.cv.nrao.edu Subject: Re: FITS "ambiguities" Date: 8 Dec 1995 17:59:20 GMT On 6 Dec 1995 10:10:20 -0500, Eric Greisen wrote: : surrounded by ' marks. Earlier examples such as : : KEYWORD = 0 1 a sample : : are just erroneous. If a comment is desired, it must be preceded by a ^^^^ : / . KEYWORD can be either a 1-valued or a 2-valued numeric parameter, The fact is that the fits definition paper does NOT require a / to start a comment. This is the source of one ambiguity. -- John E. Davis Center for Space Research/AXAF Science Center 617-258-8119 MIT 37-662c, Cambridge, MA 02139 From fitsbits-request Thu Dec 14 17:59:34 1995 X-VM-v5-Data: ([nil nil nil nil nil nil nil nil nil] ["2444" "" "11" "December" "1995" "19:25:57" "GMT" "John E. Davis" "davis at space.mit.edu" "" "51" "Re: fits ambiguities" "^From:" nil nil "12" "1995121119:25:57" "fits ambiguities" (number " " mark " John E. Davis Dec 11 51/2444 " thread-indent "\"Re: fits ambiguities\"\n") ""] nil) Received: by fits.cv.nrao.edu (5.x/S2.3/NRAO-CV/2.4) id AA14509; Thu, 14 Dec 1995 17:59:34 -0500 Return-Path: Message-Id: Organization: Center for Space Research Path: solitaire.cv.nrao.edu!hearst.acc.Virginia.EDU!uunet!in2.uu.net!salliemae!newsfeed.internetmci.com!news.kei.com!bloom-beacon.mit.edu!senator-bedfellow.mit.edu!davis References: <199512062106.QAA18781 at tetra.gsfc.nasa.gov> Reply-To: davis at space.mit.edu Newsgroups: sci.astro.fits content-length: 2442 From: davis at space.mit.edu (John E. Davis) Sender: fitsbits-request at fits.cv.nrao.edu To: fitsbits at fits.cv.nrao.edu Subject: Re: fits ambiguities Date: 11 Dec 1995 19:25:57 GMT On Thu, 7 Dec 1995 15:34:20 GMT, Kevin Kuehl wrote: : In article <199512062106.QAA18781 at tetra.gsfc.nasa.gov>, : William Pence wrote: : >On the subject of FITS keyword 'ambiguities' it might be of interest : >to give a short explaination of how the FITSIO software package [...] : Thanks for posting this. You cleared up a lot of things that I also appreciate William's explanation of how fitsio handles parsing keywords. : were confusing me (it doesn't sound like I was the only one : either). Here is the root of my confusion: : : Unless otherwise specified, the value field : must be written in a notation consistent with the : list-directed read operations in ANSI FORTRAN-77. : The structure shall be determined by the type of : the variable. The fixed format is required for : values of mandatory keywords and recommended for : values of all others. This standard imposes no : requirements on case sensitivity of character : strings other than those explicitly specified. On the one hand, I would like to see the NOST fits definition document present discussion of FITS cards that are consistent with FORTRAN list-directed read operations. However, a large part of me does not want to see the discussion at all because it might encourage fits writers to use a format other than the fixed one. It is probably too late but I would like to have the standard only endorse the fixed format for ALL keywords and change all occurances of ``strongly suggested'' to ``required''. The problem with the wording above is that I have come across several people who believe that the value-format of non-mandatory keywords is arbitrary. This is why I originally posted the question about HEX values in the integer field-- the person who created such fits files believed it was fine because the HEX value was used with a non-mandatory keyword. None of the tools that he used complained about it (iraf) until some tool using my fitsio library encountered it. While my library is forgiving with poorly formatted cards (generates warnings), it interpreted the HEX value as a boolean (the value was F0F0-- interpreted as FALSE with comment 0F0). -- John E. Davis Center for Space Research/AXAF Science Center 617-258-8119 MIT 37-662c, Cambridge, MA 02139 From fitsbits-request Fri Dec 15 13:16:00 1995 X-VM-v5-Data: ([nil nil nil nil nil nil nil nil nil] ["1613" "Fri" "15" "December" "1995" "13:14:32" "-0500" "William Pence" "pence at tetra.gsfc.nasa.gov" "<199512151814.NAA24354 at tetra.gsfc.nasa.gov>" "34" "Re: fits ambiguities" "^From:" nil nil "12" "1995121518:14:32" "fits ambiguities" (number " " mark " William Pence Dec 15 34/1613 " thread-indent "\"Re: fits ambiguities\"\n") nil] nil) Received: by fits.cv.nrao.edu (5.x/S2.3/NRAO-CV/2.4) id AA23235; Fri, 15 Dec 1995 13:16:00 -0500 Return-Path: Message-Id: <199512151814.NAA24354 at tetra.gsfc.nasa.gov> content-length: 1611 From: William Pence Sender: fitsbits-request at fits.cv.nrao.edu To: fitsbits at fits.cv.nrao.edu Cc: pence at tetra.gsfc.nasa.gov Subject: Re: fits ambiguities Date: Fri, 15 Dec 1995 13:14:32 -0500 On 11 Dec 1995 19:25:57 GMT davis at space.mit.edu (John E. Davis) wrote: > (...) This is why I originally posted the question about HEX values > in the integer field-- the person who created such fits files believed it > was fine because the HEX value was used with a non-mandatory keyword. > None of the tools that he used complained about it (iraf) until > some tool using my fitsio library encountered it. While my library is > forgiving with poorly formatted cards (generates warnings), it > interpreted the HEX value as a boolean (the value was F0F0-- interpreted > as FALSE with comment 0F0). There are only 2 possibilities for the interpretation of a FITS keyword like: KEYNAME = F0F0 It is either an boolean false value, or it is a completely invalid FITS keyword. Personally, I believe it is correct to interpret this as a boolean FALSE because list directed FORTRAN reads will interprete any value beginning with the letter 'T' as a boolean true, and any value beginning with the letter 'F' as a boolean false. Other characters may follow the initial T or F, but they do not take part in determining the logical value. Fortran also allows an optional decimal point to preceed the T or F, and allows lower case t or f logical values, but I don't think this added generality should be allowed in FITS keywords. The FITS standard is explicit about requiring an uppercase T or F for fixed format logical values, so it seems resonable to also require this for free format logical values. The FITS Standard should be made clearer on this point. -Bill Pence From fitsbits-request Sun Dec 17 08:35:02 1995 X-VM-v5-Data: ([nil nil nil nil nil nil nil nil nil] ["377" "" "13" "December" "1995" "22:04:58" "GMT" "Tom Nicinski" "nicinski at sdsshost.fnal.gov" "<4anima$cjt at fnnews.fnal.gov>" "20" "Re: FITS Browser test page" "^From:" nil nil "12" "1995121322:04:58" "FITS Browser test page" (number " " mark " Tom Nicinski Dec 13 20/377 " thread-indent "\"Re: FITS Browser test page\"\n") "<4amql2$orp at vixen.cso.uiuc.edu>"] nil) Received: by fits.cv.nrao.edu (5.x/S2.3/NRAO-CV/2.4) id AA16168; Sun, 17 Dec 1995 08:35:02 -0500 Return-Path: Message-Id: <4anima$cjt at fnnews.fnal.gov> Organization: FERMILAB, Batavia, IL Path: solitaire.cv.nrao.edu!hearst.acc.Virginia.EDU!caen!msunews!uwm.edu!fnnews.fnal.gov!sdsshost.fnal.gov!nicinski References: <4amql2$orp at vixen.cso.uiuc.edu> Reply-To: nicinski at fnal.gov Newsgroups: sci.astro.fits,sci.data.formats content-length: 375 From: nicinski at sdsshost.fnal.gov (Tom Nicinski) Sender: fitsbits-request at fits.cv.nrao.edu To: fitsbits at fits.cv.nrao.edu Subject: Re: FITS Browser test page Date: 13 Dec 1995 22:04:58 GMT I've quickly tried your FITS browser on the Web. The first suggestion I have is that your URL permit HTML FORMs to be used. Your current URLs end .../files/?fits_file where .../files?file=fits_file would be more useful as
could be used to generate URLs interactively. Tom Nicinski From fitsbits-request Mon Dec 18 21:07:23 1995 X-VM-v5-Data: ([nil nil nil nil nil nil nil nil nil] ["726" "" "13" "December" "1995" "15:14:42" "GMT" "Xinjian Lu" "xlu at solitaire.cv.nrao.edu" "<4amql2$orp at vixen.cso.uiuc.edu>" "33" "FITS Browser test page" "^From:" nil nil "12" "1995121315:14:42" "FITS Browser test page" (number " " mark " Xinjian Lu Dec 13 33/726 " thread-indent "\"FITS Browser test page\"\n") nil] nil) Received: by fits.cv.nrao.edu (5.x/S2.3/NRAO-CV/2.4) id AA03170; Mon, 18 Dec 1995 21:07:23 -0500 Return-Path: Message-Id: <4amql2$orp at vixen.cso.uiuc.edu> Organization: University of Illinois at Urbana Path: solitaire.cv.nrao.edu!hearst.acc.Virginia.EDU!portal.gmu.edu!solaris.cc.vt.edu!newsfeed.internetmci.com!vixen.cso.uiuc.edu!usenet Newsgroups: sci.astro.fits,sci.data.formats content-length: 724 From: Xinjian Lu Sender: fitsbits-request at fits.cv.nrao.edu To: fitsbits at fits.cv.nrao.edu Subject: FITS Browser test page Date: 13 Dec 1995 15:14:42 GMT Hi,there The NCSA HDF group has set up a FITS Browser test page at: http://hdf.ncsa.uiuc.edu:4321/fits This FITS Browser will let you: .Retrieve Images from a FITS file .View a 2-d image .View a group of planes from a 3-d image .Convert a FITS file to an HDF file so that you can use our HDF Browser to see the FITS data. .Get summary information from the FITS header or the full FITS header. .Read data from a FITS TABLE or binTABLE. The FITS Browser creates a nice HTML table .And more. This FITS browser is a prototype. Any coments or suggestions for additional features would be appreciated. Thanks. -xinjian From fitsbits-request Sat Dec 23 08:36:34 1995 X-VM-v5-Data: ([nil nil nil nil nil nil nil nil nil] ["6324" "" "21" "December" "1995" "13:39" "EDT" "Barry M. Schlesinger" "bschlesinger at nssdca.gsfc.nasa.gov" "<21DEC199513394652 at nssdca.gsfc.nasa.gov>" "178" "Sources of FITS Information" "^From:" nil nil "12" "1995122117:39:00" "Sources of FITS Information" (number " " mark " Barry M. Schlesin Dec 21 178/6324 " thread-indent "\"Sources of FITS Information\"\n") nil] nil) Received: by fits.cv.nrao.edu (5.x/S2.3/NRAO-CV/2.4) id AA07405; Sat, 23 Dec 1995 08:36:34 -0500 Return-Path: Message-Id: <21DEC199513394652 at nssdca.gsfc.nasa.gov> Organization: NASA Goddard Space Flight Center -- Greenbelt, Maryland USA Path: solitaire.cv.nrao.edu!hearst.acc.Virginia.EDU!news-server.ncren.net!concert!gatech!psuvax1!news.eecs.nwu.edu!newsfeed.acns.nwu.edu!math.ohio-state.edu!howland.reston.ans.net!nntp.crl.com!decwrl!lll-winken.llnl.gov!ames!newsfeed.gsfc.nasa.gov!nssdca.gsfc.nasa.gov!bschlesinger Reply-To: fits at nssdca.gsfc.nasa.gov Newsgroups: sci.astro.fits,news.answers,sci.answers content-length: 6322 From: bschlesinger at nssdca.gsfc.nasa.gov (Barry M. Schlesinger) Sender: fitsbits-request at fits.cv.nrao.edu To: fitsbits at fits.cv.nrao.edu Subject: Sources of FITS Information Date: 21 Dec 1995 13:39 EDT Date: 21 Dec 1995 13:35 EDT Expires: 23 Jan 1996 17:00 EST Distribution: world Keywords: FITS, application, image, table, astronomy Archive-name: astronomy/fits/info-sources Last-modified: 1995/12/21 No revisions this month. -------- Sources of FITS Information Preface This material on sources of Flexible Image Transport System (FITS) information is posted and updated periodically by the FITS Support Office at the NASA Goddard Space Flight Center (GSFC). It discusses where general FITS information, including some answers to frequently asked questions, can be found, and provides sources for detailed information on FITS software and documentation. ------- FITS Support Office The FITS Support Office maintains a library of FITS information accessible from http://www.gsfc.nasa.gov/astro.fits/fits_home.html or ftp://nssdc.gsfc.nasa.gov/pub/fits/. The material available includes o "Definition of FITS," a codification of FITS for the NASA/Science Office of Standards and Technology (NOST), available in LaTeX, uncompressed PostScript, compressed PostScript and (usually) ASCII text (The ASCII text version may not be available for a short while after the announcement of a new version.) o "A User's Guide to FITS", published by the FITS Support Office, in LaTex, and compressed and uncompressed PostScript o Revisions to version 1.0 of the "Definition of FITS" covering the specification of units, that were incorporated into version 1.1 (text) o A current list of the extension type (structure) names registered with the International Astronomical Union FITS Working Group (IAUFWG) (text) o Rules for physical blocking on various media adopted by the IAUFWG (text) In the same directory, but accessible directly via http://www.gsfc.nasa.gov/astro/fits/basics_info.html is the FITS Basics and Information that used to be regularly posted to sci.astro.fits and sci.data.formats under the heading of FITS Basics and Information. It continues to be revised to reflect current FITS developments. It contains the following material: o An overview of FITS o A list of FITS documents o A list of software packages that support FITS, including FITS-image converters for various platforms o A list of on-line FITS resources o A description of the FITS Support Office The hypertext version provides links to many of the documents, software, and network locations listed. The text version provides information on how to obtain much of this material. There is also a hypertext version of the List of Registered Extensions. Links from the Web page and subdirectories of the ftp directory contain o Software developed by the FITS Support Office. o Error test files, primary HDUs useful for testing the ability of software designed to read FITS files to cope with files that have errors or are non-standard. These files should be downloaded in binary form. Printed copies of the material in the FITS directory can be obtained from the Coordinated Request and User Support Office (CRUSO): (Postal) Coordinated Request and User Support Office Code 633 National Space Science Data Center NASA Goddard Space Flight Center Greenbelt MD 20771 USA (Electronic mail) request at nssdca.gsfc.nasa.gov (Telephone) +1-301-286-6695 8:00 A. M. - 4:30 P.M. U. S. Eastern Time (-0500 from the last Sunday in October through the first Saturday in April; -0400 the remainder of the year) When no one is available, messages can be left on voice mail. (FAX) +1-301-286-1635 ------- National Radio Astronomy Observatory (NRAO) A FITS Archive can be found at URL http://fits.cv.nrao.edu/ or at ftp://fits.cv.nrao.edu/fits, located at NRAO. This machine supports a WAIS server named nrao-fits which has an index of all of the FITS-related text files in the archive; the file nrao-fits.src is available at think.com and at ftp://fits.cv.nrao.edu/fits/wais-sources/nrao-fits.src. Some of the more noteworthy materials in this archive are o Drafts of proposed additions to the FITS standard and other drafts that may in the future be formally proposed o Text of any detailed proposals currently being discussed by the FITS committees o A collection of documents on World Coordinate Systems, including the current draft proposal o Conventions specific to particular projects or disciplines o Some code for various environments and Usenet postings about code o Sample data and special test files designed to measure the ability of a FITS reader to handle a wide variety of FITS files o Archives of traffic on FITS-related newsgroups and exploders ------- HEASARC The NASA/Goddard High Energy Astrophysics Science Archive Research Center (HEASARC) Web server at http://heasarc.gsfc.nasa.gov/docs/heasarc/fits.html and the anonymous ftp access through ftp://heasarc.gsfc.nasa.gov/fits_info/ provide FITS material. HEASARC has developed a the FITSIO package of software routines for easily reading and writing FITS files, in FORTRAN with a C interface available, portable to a wide variety of machines. There is also the FTOOLS collection of software tools and the VERIFITS FITS conformance verifier. HEASARC software is available directly through http://heasarc.gsfc.nasa.gov/docs/heasarc/tech_res_software.html or ftp://heasarc.gsfc.nasa.gov/fits_info/software/ . The HEASARC server also provides information from the OGIP/HEASARC FITS Working Group, (HFWG) the internal legislative body on FITS-related matters within the Office of Guest Investigator Programs (OGIP) at NASA/GSFC, at http://heasarc.gsfc.nasa.gov/heasarc/ofwg/ofwg_intro.html or ftp://heasarc.gsfc.nasa.gov/fits_info/ . The HFWG has developed a number of FITS conventions that are more specific than the requirements of the FITS standards. Proposed conventions are publicized to the FITS community as a whole, with the goal of collaborative development of a set of conventions that will be accepted throughout the community as well as within OGIP/HEASARC. ------- Direct questions about this posting to Barry M. Schlesinger Coordinator, FITS Support Office Electronic mail: fits at nssdca.gsfc.nasa.gov Telephone: +1-301-441-4205 The FITS Support Office is operated under the guidance of the NASA/GSFC Astrophysics Data Facility. From fitsbits-request Sat Dec 23 11:20:39 1995 X-VM-v5-Data: ([nil nil nil nil nil nil nil nil nil] ["420" "" "22" "December" "1995" "11:45:12" "GMT" "jesus rodriguez" "rodri12 at ibm.net" "<4be5o8$km6 at news-s01.ny.us.ibm.net>" "13" "opening Fits files in Visual Basic" "^From:" nil nil "12" "1995122211:45:12" "opening Fits files in Visual Basic" (number " " mark " jesus rodriguez Dec 22 13/420 " thread-indent "\"opening Fits files in Visual Basic\"\n") nil] nil) Received: by fits.cv.nrao.edu (5.x/S2.3/NRAO-CV/2.4) id AA08504; Sat, 23 Dec 1995 11:20:39 -0500 Return-Path: Message-Id: <4be5o8$km6 at news-s01.ny.us.ibm.net> Path: solitaire.cv.nrao.edu!hearst.acc.Virginia.EDU!caen!newsxfer.itd.umich.edu!tank.news.pipex.net!pipex!news.mathworks.com!newsfeed.internetmci.com!in1.uu.net!news-m01.ny.us.ibm.net!usenet Newsgroups: sci.astro.fits content-length: 418 From: rodri12 at ibm.net (jesus rodriguez) Sender: fitsbits-request at fits.cv.nrao.edu To: fitsbits at fits.cv.nrao.edu Subject: opening Fits files in Visual Basic Date: 22 Dec 1995 11:45:12 GMT Hello, i'm trying to develope a processing software in VB and the first step i need is the code for opening a Fits file, transform it into an array (pixel(x,y)) and visualize the image. I wonder if someone could help me in this task. If you can please e-mail me. If this is not the forum to ask it, please help me to find the right way. Thank you for your time, Clear skies, Jesus Rodriguez, MADRID rodri12 at ibm.net From fitsbits-request Fri Dec 29 05:03:35 1995 X-VM-v5-Data: ([nil nil nil nil nil nil nil nil nil] ["2314" "" "27" "December" "1995" "16:52:17" "GMT" "William Thompson" "thompson at orpheus.nascom.nasa.gov" "<4brtk1$obl at post.gsfc.nasa.gov>" "56" "CDF question" "^From:" nil nil "12" "1995122716:52:17" "CDF question" (number " " mark " William Thompson Dec 27 56/2314 " thread-indent "\"CDF question\"\n") nil] nil) Received: by fits.cv.nrao.edu (5.x/S2.3/NRAO-CV/2.4) id AA07101; Fri, 29 Dec 1995 05:03:35 -0500 Return-Path: Message-Id: <4brtk1$obl at post.gsfc.nasa.gov> Organization: NASA Goddard Space Flight Center -- Greenbelt, Maryland USA Path: solitaire.cv.nrao.edu!hearst.acc.Virginia.EDU!uunet!in1.uu.net!usc!elroy.jpl.nasa.gov!ames!newsfeed.gsfc.nasa.gov!orpheus.nascom.nasa.gov!thompson Newsgroups: sci.astro.fits,comp.lang.idl-pvwave,gsfc.general,gsfc.software.unix content-length: 2312 From: thompson at orpheus.nascom.nasa.gov (William Thompson) Sender: fitsbits-request at fits.cv.nrao.edu To: fitsbits at fits.cv.nrao.edu Subject: CDF question Date: 27 Dec 1995 16:52:17 GMT Please excuse me if this is a little outside the scope of this newsgroup, but I couldn't find one specifically for Common Data Format. I'm reading some FITS binary table files that were converted from CDF format. The original CDF files have descriptions such as !------------------------------------------------------------- ! Variable Data Number Record Dim ! Name Type Elements Variance Variance ! ------- ---- -------- -------- -------- "LAT_SPACE" CDF_REAL8 1 T F ! Attribute Data Value ! Name Type !----------- ----- ------ "FIELDNAM" CDF_CHAR { "Heliographic Lat of the Craft" } "VALIDMIN" CDF_REAL8 { -1.570796327 } "VALIDMAX" CDF_REAL8 { 1.570796327 } "SCALEMIN" CDF_REAL8 { -1.570796327 } "SCALEMAX" CDF_REAL8 { 1.570796327 } "LABLAXIS" CDF_CHAR { "Helio Lat" } "UNITS" CDF_CHAR { "rad" } "FORMAT" CDF_CHAR { "F7.3" } "DEPEND_0" CDF_CHAR { "Epoch" } "FILLVAL" CDF_REAL8 { -1.0E31 } "VAR_TYPE" CDF_CHAR { "data" } "DICT_KEY" CDF_CHAR { " " }. It appears that the data in the FITS binary table file has the same resolution as that given in the FORMAT statement, rather than the full resolution of the Real*8 data type. In the above example, the numbers are quantized to 0.001 radians, even though the double precision numbers are theoretically much more precise than this. I'm not very familiar with CDF, so I have some simple questions: 1. Are all CDF files written out to the resolution given by the FORMAT statement? Or is there a mode of writing CDF files that automatically applies this quantization? 2. If not, is it possible that the process of reading the CDF file unintentionally quantized the values? In other words, is there a mode of reading CDF files that automatically applies this quantization? 3. Failing the above two possibilities, is it standard practice for CDF files written by some of the groups here at Goddard to write out data to a quantization given by the FORMAT specification. Please excuse these possibly naive questions. Thank you, Bill Thompson From fitsbits-request Fri Dec 29 13:09:18 1995 X-VM-v5-Data: ([nil nil nil nil nil nil nil nil nil] ["100" "" "28" "December" "1995" "08:24:55" "GMT" "Brett Collars" "brettc at aloha.com" "<4btk8n$b6c at news.aloha.com>" "8" "Windows FITS freeware other than FITSVIEW ??" "^From:" nil nil "12" "1995122808:24:55" "Windows FITS freeware other than FITSVIEW ??" (number " " mark " Brett Collars Dec 28 8/100 " thread-indent "\"Windows FITS freeware other than FITSVIEW ??\"\n") nil] nil) Received: by fits.cv.nrao.edu (5.x/S2.3/NRAO-CV/2.4) id AA09178; Fri, 29 Dec 1995 13:09:18 -0500 Return-Path: Message-Id: <4btk8n$b6c at news.aloha.com> Organization: FlexNet Inc, HAWAII Path: solitaire.cv.nrao.edu!hearst.acc.Virginia.EDU!news-server.ncren.net!news.duke.edu!news.mathworks.com!newsfeed.internetmci.com!in2.uu.net!coconut!usenet Newsgroups: sci.astro.fits content-length: 98 From: brettc at aloha.com (Brett Collars) Sender: fitsbits-request at fits.cv.nrao.edu To: fitsbits at fits.cv.nrao.edu Subject: Windows FITS freeware other than FITSVIEW ?? Date: 28 Dec 1995 08:24:55 GMT Is there any other FITS freeware package for Windows besides the NRAO FITSview 4.0 ?? TIA Brett From fitsbits-request Fri Dec 29 13:41:30 1995 X-VM-v5-Data: ([nil nil nil nil nil nil nil nil nil] ["672" "" "28" "December" "1995" "10:15:40" "GMT" "Gerard Farrell" "gerard at mercury.ee.man.ac.uk" "<4btqoc$7uq at yama.mcc.ac.uk>" "35" "Re: Windows FITS freeware other than FITSVIEW ??" "^From:" nil nil "12" "1995122810:15:40" "Windows FITS freeware other than FITSVIEW ??" (number " " mark " Gerard Farrell Dec 28 35/672 " thread-indent "\"Re: Windows FITS freeware other than FITSVIEW ??\"\n") "<4btk8n$b6c at news.aloha.com>"] nil) Received: by fits.cv.nrao.edu (5.x/S2.3/NRAO-CV/2.4) id AA09253; Fri, 29 Dec 1995 13:41:30 -0500 Return-Path: Message-Id: <4btqoc$7uq at yama.mcc.ac.uk> Organization: Manchester Scientific Instruments Ltd Path: solitaire.cv.nrao.edu!hearst.acc.Virginia.EDU!news-server.ncren.net!news.duke.edu!agate!sunsite.doc.ic.ac.uk!yama.mcc.ac.uk!usenet References: <4btk8n$b6c at news.aloha.com> Newsgroups: sci.astro.fits content-length: 670 From: gerard at mercury.ee.man.ac.uk (Gerard Farrell) Sender: fitsbits-request at fits.cv.nrao.edu To: fitsbits at fits.cv.nrao.edu Subject: Re: Windows FITS freeware other than FITSVIEW ?? Date: 28 Dec 1995 10:15:40 GMT In article <4btk8n$b6c at news.aloha.com>, brettc at aloha.com says... > >Is there any other FITS freeware package for Windows besides the >NRAO FITSview 4.0 ?? > >TIA > >Brett There is a freeware version of our FITS viewer for Windows on our web site, http://mercury.ee.man.ac.uk Just follow the services & products links. Alternatively use anonymous ftp to mercury.ee.man.ac.uk/pub/fits. Regards, Gerard -- Dr. Gerard Farrell Manchester Scientific Instruments Ltd. Campus Ventures Centre University of Manchester Oxford Rd Manchester M13-9PL UK Tel: 0161 273 5110 Fax: 0161 275 4546 e-mail: gerard at mercury.ee.man.ac.uk WWW: http://mercury.ee.man.ac.uk