[Kst] Re: Benefit vs risk of constant-width columns + GUI proposal

[Nicolas Brisset]

> Then I would say, when the user doesn't know his data he should not
> use same non default functions.
OK. That's why we should default to the less risky version (no assumption). That's what I did.

> Yes, please commit.
Just done. Now you need to connect the code to the GUI...
For units, feel free to add them if you want. I will avoid working on the ASCII source while you're also working on it. I have enough to do with icons for a while. I'll pick it up from where you leave it... if there is anything left!
 
> Why do you think we break old kst files, the only difference I see is
> the changed default comment delimiter, arne't they loaded from the kst
> file?
Maybe I'm wrong, but I thought some options are saved as an int/enum. If we split them into two enums or add numbers to the sequence, it can become tricky. But again, I don't really have the details in mind right now. I guess you'll see when you implement it. Jsut keep the issue in mind, should it turn out to exist...

> And I have the impressen the read code for ascii data is completely
> broken:
There I can't help you. I never completely understood why there are so many functions to read/save from/to .kst files. Maybe it's just old cruft that never gets called. Or maybe there really is a good reason. Maybe Barth knows more?
 
Nicolas