Brian,
Thanks for your reply and for the articles you linked to – especially the last one is very interesting. I have been aware of that but not so well explained.
As I think you are very cunny both on spreadsheets and how to use them and also the mathematics, I am not able to argue.
But I am sure you know what I meant when I said that a spreadsheet is ‘totally’ worthless if you cannot trust it, if it -- the program, its features or functions -- does not work (behave) as supposed.
As an engineer I have been working with MSExcel since late 1980 and have since 10 years successively began to use OO/LO.
I’ve always tried to crosscheck important calculations and with calculation processes start with some few (random or logical) sample data and/or manual calcs to be sure that the formulas (& combinations of) work OK.
Whatever people say about Microsoft, I must say that I have never experienced any problems caused by the program MSExcel itself (perhaps I’ve been lucky) – all my problems have been caused by my typos or my bad planning of the calculation process.
But, especially with LibreOffice/Calc I have experienced issues that clearly depend on LO as a program and have caused me a lot of extra doings – the last couple of years LO has become better.
Regarding the articles. If the programming of the "very big" programs is purchased from a "third party" where the programmer -- the person -- does not have no own knowledge (understanding) about the customer's process, then it is an unavoidable necessity that there will be flaws and serious issues -- regarless of branch.
Anyhow, the fact is that Microsoft has a 90 percent market share and every MSO suite includes MSExcel.
Above this are e.g. AOO/OO/LOwith roughly 2 mega spreadsheet users and their applications.
That means that quite a lot of – some very important -- calculations are globally done with these MSO&AOO&OO&LO (..etc..) spreadsheets.
If the spreadsheets are that unreliable as these articles (and you) explain, should it not then be good - if not a necessity – to warn about it? Experts and “nerds” can manage but not ordinary users like me.
I like LibreOffice and respect its intensions and thus I want it to be better.
If LO really wants to increase its market share – to beat MSO, a benefit for all users – then it has to eliminate every possible cause of problems especially for the ‘ordinary user’ -- even the small ones like that sorting issue.
The ordinary user selects a range and pushes the sort button -- if the sort does not behave as expected, and there is no explanation why or how and the help is too difficult to find -- then he/she considers this a problem.
Too many problems is not good for LO.
It is the mass of ‘ordinary users’ that build most of the MSO’s market share.
The main and final responsibility lies – not on the (unpaid) programmer -- but on the controlling team in the LO organization.
It is better to fix problems at a stage of planning and production than after a global distribution(ref. to Quality Management Systems).
regards Pertti Rönnberg