I haven’t felt much like writing lately, as I don’t seem to have anything special going on in my world. I have wrapped up development on the Testing Center application I have been working on since April. A lot of new things were introduced to me that I never had to work with before. Let me sum up a few of the pros and cons of the tools I used to build this project:
C# / Object Oriented Programming
- Logic mistakes aside, OOP allows code to be written less faulty. Responsibility for this is largely due to auto namespace completion, and typed variables.
- User / Company created extensions can be included in your projects
- Typed variables means a lot of code to perform trivial actions. Specifically there is a lot of checking, catching for null strings being converted to integers.
- Even simple tasks become complex. Reading a file means instantiating numerous objects before you can begin.
- Sessions and hashtables are STILL the fastest way to deal with key / value pairs. Auto name completion wont take advantage of this.
- Data focused objects allow rapid development of basic functionality. Paging, sorting, inserting, deleting and updating are (almost painless)
- It is C#, so there is no crazy learning curve for syntax.
- Deviating from the default actions on data seems at times to almost cancel out the benefits offered by the data objects. Conditional formatting is often dozens of lines of extra code.
- Rapid application development seems to have killed my passion. I have never cared less about code I have written.
- Instead of focusing on application logic, much time was spent figuring out how to “trick” ASP into making it do what I wanted it to. (Some of this is because it is new to me)
- Compared to C#, ASP seemed to have fewer user / company created extensions.
- It is easily one of the more polished products Microsoft offers.
- It is clearly made BY developers for developers.
- It has a good implementation of auto name completion complete with defaulting to recently used items.
- Syntax highlighting, code folding, line numbering, matching bracket highlighting, smart indentions, and many other features you would expect from a code editor.
- The “designer” mode auto populates column names in data tables / data sources.
- Like all monolithic programs, it does many things mediocre.
- It isn’t specific to web applications, so many features clutter the interface that aren’t applicable.
- The 10% of the features that almost everyone use are hidden by the remaining 90% of things that you would never use
- The server browser is horrible.
- The dataset tool is a gift and a curse. I like the idea of abstracting SQL, but it reformats the queries I would write.
- It still has some bugs where things like changing a column’s data type is completely mishandled by the dataset.
- Database creation is completely visual. I didn’t write a single line of code.
- Seems fast, and reliable.
- Pretty much transparent. None of the major problems I had were because of SQL Server.
- Windows only.
- Many functions present in Oracle’s databases are missing
- The only “management tool” I saw was the built in Windows tools. It wasn’t pretty. Perhaps others exist.