>That, or check for 1900-01-01 in the cursors and blank those values. That will change the update status on those records, but then there's a trick - you set the cursor to non-updatable, do a tableupdate(), change back to updatable. Easy to do if you're using a cursoradapter.
I can see that working for something with just one record or a few - but when doing reporting, this app has date fields that also work like a flag e.g. field Funding_Date not only gives the date of funding for their product but if it is empty, it means the product has not been funded yet. Where it comes into play on reports is running a report of unfunded files e.g. WHERE EMPTY(Funding_Date); so would I have to bring down the whole table first, nuke the Funding_Date for empty dates and then run another query against the local cursor? Or I guess re-reading your answer, I would have to put in AND Funding_Date # {^1990-01-01}?
>That's the limit on string value in a variable (perhaps not even there, but many string functions may chop it) but not on a memo field. You can still do strtofile() and filetostr() straight with a memo field. I saw it work for a few hundred megabytes.
I think someone did point out that StrToFile()/FileToStr() can handle larger strings - in this case, the string has to be run through STRCONV() because of encoding and STREXTRACT() to pull out data between tags so from what I can tell, we are bumping up against 16MB limits once we get to those other functions. Mind you, if I rewrote the entire app myself to do a away with it going through this document app, maybe I would not need any of that.
>In SQL 2016 the allegedly doomed types of text and image (which wear "to be deprecated" tag for a number of years now) are still there and working as advertised. I use them for memo and binary memo, and have experienced no problems. As for Postgress, see
https://www.postgresql.org/docs/9.5/static/datatype.htmlI looked at the Postegresql docs on data types - do you know is there really any difference between varchar and text?
Albert
