"Date and Time" Extents and Preference Saves

raswaim · Post by **raswaim** » Thu Feb 03, 2005 5:13 pm

I've experienced two problems with using a Date and Time format X-axis in version 1.9.8.1:

1) With a "Date and Time" formatted x-axis, if I manually specify extents, close the extents dialog (hit "OK") and subsequently go back to the extents dialog (to change the y extents, for example), I find that the X-extents have been changed by 1 day. Example: I set one extent entry to "2/3/2005 16:00"; this works, but when I come back to the dialog, I find the field has "2/4/2005 16:00" in it... so I have to restore the date everytime I subsequently want to change any other extent/tick mark/size setting.

2) When I store a preference for a plot with HH:mm:ss time formats, applying that preference set to a new plot leaves me with the default format (AM/PM).

Post by **DPlotAdmin** » Thu Feb 03, 2005 6:20 pm

1) Even better, with 'Calendar date' formatting the upper extent increases by 1 month with every usage.

2) Time format isn't saved to the preference file. Just an oversight.

Thanks for pointing out the problems. These are both easy and will definitely be fixed next time.

Post by **DPlotAdmin** » Thu Feb 03, 2005 7:38 pm

1) With a "Date and Time" formatted x-axis, if I manually specify extents, close the extents dialog (hit "OK") and subsequently go back to the extents dialog (to change the y extents, for example), I find that the X-extents have been changed by 1 day. Example: I set one extent entry to "2/3/2005 16:00"; this works, but when I come back to the dialog, I find the field has "2/4/2005 16:00" in it... so I have to restore the date everytime I subsequently want to change any other extent/tick mark/size setting.

After further review, it looks to me like everything works as expected if you also specify the tick mark interval. If this isn't what you're seeing then please save and send me a DPlot file. If you aren't forcing the tick mark interval, the problem is competing goals + roundoff error - your selected extents vs. a tick mark interval that DPlot thinks "makes sense". It may be that I can improve on this anyway, but... maybe not.