Forwarded message: > From: ftr <[email protected]> > > In the interest of the software and its users you might be more specific. Sorry, I thought this would be well known in the gnumeric community. There is a whole literature on statistical problems with Excel (which would be the ones I would run across). Here is one place to start:
http://www.csdassn.org/software_reports/gnumeric.pdf This looks at a very narrow range of issues. Here https://oit.utk.edu/research/documentation/Documents/ExcelStatProbs.pdf is a paper discussing non-overlapping issues. There are plently more. > On 21/05/2015 15:25, Bob wrote: > > What atracted me to gnumeric is that it corrected some of the many > > errors in Excel. So, my main concern here is that we not keep Excel > > errors in the interests of compatibility. > > > > Forwarded message: > >> From [email protected] Wed May 20 22:42:04 2015 > >> From: Markus Mohrhard <[email protected]> > >> > >> Hey, > >> > >> I'm one of the LibreOffice Calc maintainers and one of the people > >> maintaining the calc automated tests. > >> > >> As you might know there is no official OpenFormula test suite which in my > >> opinion is one of the big issues with OpenFormula. To make matters worse > >> many corner cases are not specified and therefore implementation defined > >> resulting in problems around interoperability. > >> > >> Another point is that adding new features requires to add them into an own > >> namespace which is important for Excel interoperability. As more or less > >> every implementation is implementing its own names for these non-specified > >> functions it further causes interoperability issues. > >> > >> > >> I was wondering if there is interested in tackling these issues as part of > >> a cooperation between the two projects by creating a test suite for > >> OpenFormula functions and related functions. The idea would be that this > >> suite could be consumed by both projects in their automated tests to make > >> sure that the implementation is consistent and that names for new functions > >> are the same for both projects. > >> > >> Currently my idea would be to use flat odf for a document containing a > >> spreadsheet with the functions and csv for the expected results but I'm > >> open to other ideas that create easy to consume tests. > >> > >> I'd love to hear what you think of this proposal for collaboration. > >> > >> Regards, > >> Markus > >> > >> --047d7b66f8e5061bb205168e7c71 > >> Content-Type: text/html; charset=UTF-8 > >> Content-Transfer-Encoding: quoted-printable > >> > >> <div > >> dir=3D"ltr"><div><div><div><div><div><div><div><div><div>Hey,<br><br><= > >> /div>I'm one of the LibreOffice Calc maintainers and one of the people > >> = > >> maintaining the calc automated tests.<br><br></div>As you might know there > >> = > >> is no official OpenFormula test suite which in my opinion is one of the > >> big= > >> issues with OpenFormula. To make matters worse many corner cases are not > >> s= > >> pecified and therefore implementation defined resulting in problems around > >> = > >> interoperability.<br><br></div>Another point is that adding new features > >> re= > >> quires to add them into an own namespace which is important for Excel > >> inter= > >> operability. As more or less every implementation is implementing its own > >> n= > >> ames for these non-specified functions it further causes interoperability > >> i= > >> ssues.<br><br><br></div></div>I was wondering if there is interested in > >> tac= > >> kling these issues as part of a cooperation between the two projects by > >> cre= > >> ating a test suite for OpenFormula functions and related functions. The > >> ide= > >> a would be that this suite could be consumed by both projects in their > >> auto= > >> mated tests to make sure that the implementation is consistent and that > >> nam= > >> es for new functions are the same for both > >> projects.<br><br></div>Currently= > >> my idea would be to use flat odf for a document containing a spreadsheet > >> w= > >> ith the functions and csv for the expected results but I'm open to > >> othe= > >> r ideas that create easy to consume tests.<br><br></div>I'd love to > >> hea= > >> r what you think of this proposal for > >> collaboration.<br><br></div>Regards,<= > >> br></div>Markus<br></div> > >> > >> --047d7b66f8e5061bb205168e7c71-- > >> > >> --===============4383337659633115881== > >> Content-Type: text/plain; charset="us-ascii" > >> MIME-Version: 1.0 > >> Content-Transfer-Encoding: 7bit > >> Content-Disposition: inline > >> > >> _______________________________________________ > >> gnumeric-list mailing list > >> [email protected] > >> https://mail.gnome.org/mailman/listinfo/gnumeric-list > >> > >> --===============4383337659633115881==-- > >> > >> > > > > -------> First-time AP Stats. teacher? Help is on the way! See > > http://courses.ncssm.edu/math/Stat_Inst/Stats2007/Bob%20Hayden/Relief.html > > _ > > | | Robert W. Hayden > > | | 614 Nashua Street #119 > > / | Milford, New Hampshire 03055 USA > > | | > > | | email: bob@ the site below > > / x | website: http://statland.org > > | / > > '''''' > > > > > > _______________________________________________ > > gnumeric-list mailing list > > [email protected] > > https://mail.gnome.org/mailman/listinfo/gnumeric-list > > > > _______________________________________________ > gnumeric-list mailing list > [email protected] > https://mail.gnome.org/mailman/listinfo/gnumeric-list > > -------> First-time AP Stats. teacher? Help is on the way! See http://courses.ncssm.edu/math/Stat_Inst/Stats2007/Bob%20Hayden/Relief.html _ | | Robert W. Hayden | | 614 Nashua Street #119 / | Milford, New Hampshire 03055 USA | | | | email: bob@ the site below / x | website: http://statland.org | / '''''' _______________________________________________ gnumeric-list mailing list [email protected] https://mail.gnome.org/mailman/listinfo/gnumeric-list
