Comments on: How are you currently dealing with compatibility issues of Excel 2003 / 2007?

By: J.Makinen

J.Makinen — Sun, 16 May 2010 09:28:19 +0000

Guite simple rule works around with this annoying compatibility issue.

If you use 2007, then save the files always in 2007 format, not in 2003 (unless you need to downgrade it to previous Excel vrs. Should this be required, just work in ‘2007′ format until you need to e.g. send to someone else who you does not have 2007). This may sound quite naive advice, but I’ve tried so many different ways to sort this out and none of them has provided 100% working solution except this very simple principle.

The compability problem can happen even with very simple spreadsheets (replacing ‘toolpak’ function with #NAs, lost of data etc). Sometimes just only thing to trigger the problem is to open the file and resave it with other name (why this should cause an issue, I don’t have a clue).

The compatibility is the only major problem with Excel 2007 as far as I’m aware.

By: John Graham

John Graham — Fri, 07 May 2010 06:48:34 +0000

I develop models using Excel 2003 and test them in Excel 2004 & Excel 2007 (but not Excel 2008 since it does not have VBA).

The critical compatibility issues we have run into are:
(1) loss of functions in the Analysis Toolpak
(2) having too many formats

Work arounds we use to avoid the issues:
(1) a VBA module to implement the Toolpak functions as user defined functions, eg globally replace EOMONTH by EOMONTH_vba, EDATE by EDATE_vba, etc.
(2) systematic removal of styles & conditional formatting

By: Adrian Miric

Adrian Miric — Mon, 08 Mar 2010 19:53:54 +0000

With regards the =#n/a issue it is a serious issue with regards spreadsheets built to be used in Excel 2003 and Excel 2007. Unfortunately it is not possible to replicate it, so presumably this is why it still exists.

We have tried to explain it at http://www.auditexcel.co.za/NA-Error-Excel-2003-to-Excel-2007.html but all comments and feedback welcome.We believe this could cause serious loss of money at certain clients.

By: Gavin Townshend

Gavin Townshend — Tue, 26 Jan 2010 16:11:29 +0000

On several occassion I have seen the compatability problem of formulae using Analysis Toolpak functions being converted to #REFs (or is it #NA?).

I heard a suggestion for a potential solution to this problem, so now I am trying to reproduce this problem in a brand new spreadsheet with only one or two test formulae. But so far I can’t reproduce the problem, so can’t test the potential solution.

Can anyone help me reproduce the problem in a newly created simple spreadsheet (rather than discuss potential solutions or workarounds)?

By: Luke

Luke — Thu, 21 Jan 2010 21:26:02 +0000

I work in a US investment bank and the IT department doens’t allow installations of both Excel 2003 and 2007 so I am stuck with Excel 2007 for all financial modelling tasks. The amount of frustration this is causing is unbelievable and my life is jam-packed with corrupted workbooks and confused and irritated clients. If only they would let me, and everyone else involved in financial modelling, go back to Excel 2003!!

By: Jon Peltier

Jon Peltier — Thu, 21 Jan 2010 13:54:11 +0000

I use 2003 for my own use and for development of workbooks and coded solutions for clients. Naturally many clients have now downgraded to 2007, so I test everything in 2007, and make the requisite adjustments prior to delivery.

Many coded solutions should actually be opened, saved, and closed in Excel 2000 to ensure compatibility, because a number of recent upgrades to 2003 and 2007 seem to cause issues with each other.

I keep 2007 on a separate machine (virtualization is a godsend) to avoid cross talk between versions.