I could be going mad, or becoming incompetent in my old age, but it seems ever since the recent update I have been experiencing problems I have never come across before.
The first problem I encountered was where I exported some tables to MIF to share with our client. To check if the exports were correct, I re-imported one back into MapInfo and was greeted with the error "unsupported character set: UTF-16". For starters MapInfo exported in this charset, how can it not be supported? Second, I didn't recall specifying such a charset so went back through the process – seems I chose to export in the source's charset. Right, so the source turns out to be UTF-16 for some reason. Again, if it's IN UTF-16 how is this not supported when I'm pulling it back in? More to the point, WHY is the source file suddenly in this charset when I'm pretty sure windowsLatin1 has been the standard.
But then next I'm pulling in a CSV of float values, some of which are over 999, as a lookup table to update another table. Confidently using the add column function to do this, I run it my program without error, search for 0s, and was satisfied everything had populated as it should. But then I notice several records have values of 1 in them. That's not right. I check the lookup and the value that is supposed to being pulled is 1,876. Whatever I do mapinfo interprets this as 1. Both source and destination fields are floats. Both accept vales of over 999 if I type directly into them, and MapInfo even adds commas to separate them by thousands. But pulling the number from one table into the other and it truncates it at 1: at the comma.
So I look at the original CSV and obviously as its comma delimited, numerical values WITH commas in them have escape quotes around them. Right. Clearly this value in escape quotes is being interpreted differently or something. Which is utterly baffling, as how on earth have I never experienced this problem in 20 years of MapInfo use? I must have imported a billion CSVs into MapInfo over my lifetime, and certainly a significant proportion of those will have contained values over 999. I work with coordinates ffs! Not once have I ever noticed an instance of MapInfo changing these values to 1. I SURELY would have noticed.
So I grab an old CSV at random which I know contains floats of 1,000+. And the pop-up I get is defaulting to UTF-8 charset. Eh? Since when? And not even UTF-16? But 8? So I venture into Options and system settings say UTF-16 for new tables. Hmm. It may be just one of those things I have been incredibly unobservant about but that doesn't seem typical to me. Could this be why the initial table in my first encounter was in this format? And if MapInfo is literally defaulting to that format, how come it says it is not supported when I'm trying to open a MIF? And are all these problems such as the truncating of 1,000 to 1 related?
Am I having a breakdown, or is something going on here? I mean, are we supposed to remove all thousand separators from our raw data now? How long has this been a thing? Dazed and confused.
------------------------------
Ryan Cook
Knowledge Community Shared Account
------------------------------