If you’re trying to do a SqlBulkCopy using a DataTable and getting lots of column type (integer, string) mismatches, it’s probably because you haven’t mapped not only the column types, but also the column names.
SqlBulkCopy assumes that your DataTable is going to have its columns in the same order as the database table, which will cause mismatches when it tries to send your (string) Name field into your (int) Age column.
Just having your DataTable column names match the table isn’t enough, either. There must be a manual mapping.
But if your DataTable column names are the same as the table column names, a simple method can take care of that mapping for any DataTable you send it.
So I’ve been working on the next major release of the D20 Spell Lists app, and have found myself in a code and UI reorganization/refactoring jungle as I’ve refined my feature set and how I want to handle things.
One of the common-enough cases that the current version doesn’t handle well that I think needs to be is multiple spellcaster classes. If I’m a Druid 5/Bard 6, I’m going to want to keep separate spell lists, and will have different DCs, spells known, and spells per day to contend with. With the current version, the best solution is probably to have two different character files, each with its own spell lists.
My attempts to smooth that out have resulted in a lot of UI revisions as I tried to find good ways for displaying and updating all this info without slowing down the UI to uselessness (buh-bye, Table Layout Panel). Here’s what I’m sitting with now:
It’s not pretty, and there are more tweaks in the queue (like just putting the DCs/per days right with each class, probably), although for the moment it works well enough (read: stuff doesn’t end up hidden and inaccessible) that I can get functionality working again.
I have a friend who’s looking to become a marketable developer fairly quickly, but is essentially trying to go from zero to hero. Despite starting a project in Python to randomize wedding slideshows, he’s never really done development, nor does he understand its core concepts (classes vs. objects, for instance).
He kinda just needs a job, plus the ability to make useful tools for himself.
Folks are telling him Java and *.NET, and I (mostly) agree for the simple purposes of 1) easy desktop or web development, and 2) using keyword-heavy languages to reinforce programming concepts. Not strong reasons, no. If he just wanted a job, I’d recommend UI development.
Plus, for marketability in Charlotte, NC–home of plenty of Big Banks–diving into quasi-enterprise development can get you a cube (albeit maybe shared) and a comfortable paycheck. Good luck finding a Python gig here.
So I’m in a D&D 3.5 campaign now. And I’m playing a Druid, which is kinda exciting–it’s the first time I’ve played a Druid at a high enough level that I can shapeshift, and I just tipped 5th level on Sunday. (Campaigns always fizzle out early…)
Anyway, the campaign is a hodgepodge of standard D&D and Sandstorm, and summoning restrictions by the GM mean that my spell list involves a fair bit of swapping out that’s a little annoying to manage. For instance, I’m using Sandstorm’s “Desiccate” instead of “Summon Nature’s Ally II”, since I can’t summon.
Since Druids are the type to prepare a few spells per day from a large list of available spells, I needed a quick way to see all of my available options without flipping through two sourcebooks and my swap list at the table. Then I wanted to avoid having the jotted down short-hand description of the spell that I refresh every “day” as I try out new spells.
Pain in the ass.
A bit of Googling lead me to conclude that with the advent of D&D 4e, many of the 3.5e resources… disappeared. And by disappeared, I mean:
All files withdrawn at the request of Wizards of the Coast.