Let me share an experience. We are in the middle of a migration project and under the pump. I needed to get a CSV file into a SQL 2000 database and a SQL 2005 database, quickly. I used DTS on the run to copy the file into the SQL 2000 database - it took about 5 minutes.
Then the fun started. I tried to do the same thing in SSIS. SSIS decided that varchar(50) was a good edefault rather than varchar(255) which DTS defaulted to. So an extra couple of minutes manually redesigning the table. Add a couple of connections, a data flow task, notice validation warnings and hit execute. A veritable encyclopedia of error messages mostly about obscure code pages which frankly I don't give a monkey's about. I could not get it tow work (even after it offered to "fix" the problem. I ended up using SSIS to import the table from the SQL 2000 database!! Time to achieve in SSIS - about 25 minutes (and only by "cheating").
This is woeful. Yes code pages may be important to some users some of the time, but surely as a default SSIS should be able to do better than that. This is just wasting time and money.
What should I learn from this? Was there a wizard I should have used? Do I need to become a code page guru? Is SQL 2008 improved? Should I not use SSIS for a simple ad hoc task like this? (and if not, then what?)
Thanks for any advice - I promise to treat responses with respect (despite the tone of my post above).
Les