On Tue, 5 Sep 2006 20:54:46 +0200, Uri David Akavia wrote (in article <[EMAIL PROTECTED]>):
> If you want to assign an identifier to each cell of a sheet, what's > wrong with using the identifiers you already have - that is, the cell > address? There are a _lot_ of thing that are "wrong" with those stone-age A1-style hard-coded cell addresses dating back to Visicalc more than 25 years ago. Only a few of the fundamental flaws are: - They make formulas un-readable for human beings - They make it impossible to track errors in formulas - They break as soon as someone moves cell contents around - They make sheets essentially un-reusable, which was originally _the_ basic reason of existence of so-called "office applications" - the posiisibility to re-use results of work already done All this has been well known to all people who have ever used a spreadsheet for any kind of application case with a complexity that goes beyond ridiculous demos at trade show booths. And more than 15 years ago, there already was an application that showed how to get this right (or at least better), Lotus Improv. And now please don't tell me that the Gnumeric developers, just like those of OO Calc and Kspread, are too ignorant and incompetent to know this? And that they expect their users remain stuck at an (un)productivity level that was already obsolete more than 15 years ago? And that they can't do any better than simply implement a clone of this ridiculous Excel crap? Say it ain't so, Joe, please, say it ain't so-oh... Sincerely, Wolfgang Keller -- My email-address is correct. Do NOT remove ".nospam" to reply. _______________________________________________ gnumeric-list mailing list [email protected] http://mail.gnome.org/mailman/listinfo/gnumeric-list
