C/++, you're on notice!

Post Reply New Topic

Oodain, I reviewed Arduino’s website and there is just so much great material. I beats just going through the booklet. I want to send my students there to show them a wealth of projects with pretty darn good explanations. I want them to take an interest in the project.

sliqua-jcooter, I don’t think I could build a calculator. I’m not a technology teacher, I’m just a student teacher assigned to a home room of kids that are having a difficult time in school. I got the technology lab, I thought it could be a fun project for the students. They did pretty good with the Arduinos Friday making all sorts of modifications on them like adding LED’s, making a flip flop flashing setup.

Besides going through various projects on Arduinos site, I do hope to give them an important lesson. “The language C, to retrieving stored information and do mathematics. As a programmer, you can make the program do what you choose but you have to let it know if you don’t need it to give you an answer for a math problem… Right now, you’re probably not going to need it to solve math for you, that’s why you right VOID.” I don’t want to mess them up if they choose to pursue this.

I also want to show the kids the push button, tomorrow. At the end of the week, I hope the students can partner up and create a mini project. There isn’t a grade for this. Most of the students really have a hard time with school so I hope that this can give them some inspiration and ownership.

By the way, I didn’t actually, put C ‘on notice’ one of my students did. The student had me laughing to death.

If anyone has any suggestion as for the next direction I should move the class to, I'd love to hear it, because I don't really know.
Thank you guys, you actually did help me quite a lot!

This is closer to true - but in actuality, you can do a lot more than simple math in a method that returns something. For instance, you could have a method that takes arbitrary text and strips out any text that is a URL, and just return that string. So the distinction isn't that it's giving you an answer to a math problem necessarily, it's that it's passing *something* back. Again - it's not something easy to verbally explain, but once you actually start using methods that return values it becomes pretty apparent.



It shouldn't be particularly difficult - I've got an electronics bench in my office that I use far too sparingly - I'll see if I can grab some time this week and come up with a sketch. The nice thing is it can be made very simply, and then expanded on to become extremely complex (you can start with outputting information to the PC, then add a 7-segment display later, and then add a matrix keypad after that - and from there you can get into shift registers and all kinds of crazy stuff).

If you're in a major city, you could also look into seeing if there are any hacker spaces in your area that would be willing to help, either with space/equipment or lesson development/etc.

It appears so. Hopefully it will be removed eventually.



There's no deep reason why you should to specify the return type void when the function doesn't return a value. It's just that a funcion has the form

[return type] [name of function]([arguments]) {
[code]
}

so if you were allowed to write

f()

for a function that doesn't return anything it would be inconsistent with the syntax used for functions that return a value which makes things messier. C actually defaults the return type to int when it isn't specified which is arguably even worse.

If you don't want to do math, C/C++ aren't for you, nor are any of the FORTRAN/Algol derived languages such as Pascal, Python, Java, etc. You might want to try Smalltalk, COBOL, or BASIC.

Here's a more accurate statement:

Being a proficient programmer requires a basic knowledge of mathematics no matter what language you use. It might be slightly less in some languages but often basic math is inherent to the problem you are trying to solve. In these cases, you can't avoid it by using a different language.

Also, if you have trouble with math, you'll probably also have problems with logic (which is even more important for programming).

Geez, if you like C names, check out SQL Server. There is nothing quite so much fun as trying to convert a C int to a SQL Server tinyint or bigint. SQL also has lovlies like the smalldatetime and the datetime. The difference between the two being what, thousanths of a second?

Oracle is great too. Varchar2. What was so tough about a Varchar that they had to make a Varchar2?

I know the difference and I use Varchar2 all the time. I just wish at least MS would standardize. Converting an int16 to a tinyint or the reverse in SSIS is a PITA.

datetime rounds to the nearest thousandth of a second, smalldatetime rounds to the nearest minute. The smaller one is 4 bytes, the larger one is 8. If you need precision, the choice is obvious. If you need to store years worth of logs with timestamps that only need to be accurate to the nearest hour, the choice is also obvious.


Maybe I'm missing something, since I've never actually done that, but what's so hard about it? Can't you just cast to uint8_t or int64_t?

Whicj areas of maths are you referring to? Do you really need an in-depth knowledge of algebra, trig etc.? Or just binary and hex?

I just meant basic properties of integers for the most part. For some programming tasks, more advanced knowledge is needed; however, this is sufficient in many cases. Of course more math is always better...

With particular emphasis on numerical methods. In the last analysis, all a computer can crank out are approximations. (Yes, I am aware that a limited amount of symbolic math can be done with computer).

ruveyn

My observation about smalldatetime and datetime is that with any MS Date, you have a timestamp. Creating a simple date is anything but as it requires a mask. Oracle splits these out into date and time stamps. You only use what you need. SQL server does not permit a date without a time.

In most of my processes I read from SQL Server in C# and reply back to SQL server from C#.

C# does not have bigint and tinyint. SQL Server does not have int16 or int32. You have to cast to and from the SQLDbTypes in the C#. This suks if you are using multiple recordsets prior to SQL Server 2005 and so are using the OLEDB connections instead of the SQL Server ones.

That's a reasonable criticism.


OK. Personally, I'd lean towards blaming C# for this one.