> At 09:44 28/02/2010 -0800, Brewster Gillett wrote: <snip> > >This is an incomprehensible error, and calls into question the > >integrity of the code, big-time ...
Brian Barker wrote: > Either that or the care taken by the user? Let's be fair: you need > to test and reproduce such a problem before you can rightly assume > that it is a bug. Right you are, Brian - of course it's more a matter of ignorance than of lack of care.... > bg: > >... because I cannot come up with any form of operator error that > >could have caused such a thing. brian: > Perhaps I can help: may I please suggest such an error? > > Are these values numbers? As far as I've ever been aware, most any spreadsheet defaults to numbers if that's the first thing you type into it - only formatting as text if the first-typed character is non-numeric. Have I been misinformed? brian: <snip> > If so, is > it possible that some of your text values have leading blanks? Not so far as I can tell. brian: > Alternatively, are some of your values numbers and others text? Try > View | Value Highlighting: are they all black (text) or all blue > (numbers) - or a mixture? bg: Thank you for a very useful tip, which I may never have discovered otherwise. Sure enough, the ones that aren't included in the sort are showing as text. And, fool that I am, I should have seen it - every single one of them has a leading single quote - an artifact whose presence is a complete mystery to me. Let me elaborate on one of the still-puzzling aspects of this - I only have two ways I can have a ZIP code end up in this spreadsheeet. Either the name comes in as part of a .CSV file downloaded from our National Office website, or it comes directly to me from an inquiry by phone or email. In the former case, the .CSV file gets saved as an OOCalc file. In the latter, I would have manually typed in the ZIP. But here's the part that is bizarre; the split between those with leading single quotes and those without does not even close to line up with the split of sources just described! IOW Some of the ones I manually typed show up as numbers, and some as text. And likewise, some of the .CVS imports show one way, and some the other. I submit to you that that makes no sense at all.... .... and I still have no idea how those with the leading single quote got that way. Now here's a whole new question that this has triggered - I have read completely through the Help screens regarding Find and Replace, and nothing speaks to why Find & Replace is unable to process these entries. So thus far it looks like I am obliged to strip off the leading single quotes manually, cell by cell. Apparently Find & Replace cannot "see" a text entry like '97103 . This is not how my mentors described "automation" to me :-) brian: (At least, try this before dismissing the problem as a bug.) bg: What led me to believe it might be a bug is that it has never done this in previous versions....only since the 3.1 upgrade... ... a natural assumption, on some levels. brian: > I trust this helps. bg: It helps a great deal, thanks. Now if you have any ideas as to why Find & Replace seems to be so limited, that would also help a lot..... Thanks, Brewster -- *********************************************************************** Embrace a sharing community of sustainable justice low-carbon diversity *********************************************************************** W. Brewster Gillett [email protected] Portland, OR USA *********************************************************************** Simply because you don't like to hear it, that doesn't make it untrue. *********************************************************************** --------------------------------------------------------------------- To unsubscribe, e-mail: [email protected] For additional commands, e-mail: [email protected]
