[UPHPU] priorities was : using $_SESSION to store a user object

Kyle Waters unum at unum5.org
Tue May 31 09:49:06 MDT 2011


  On 05/27/2011 06:11 PM, Aaron Luman wrote:
> Hey all,
>
> I posted this question a while ago to stackoverflow but didn't get any bites.  I'm curious about security, performance, etc issues in how I have planned out a site I am putting together.  Background, the project is to create something similar to blackboard (minus all the suck).  This is primarily being done as a learning experience, but if we end up liking the project enough we might try to convince some of the teachers here to adopt it.
>

Most of this question was already answered in an other reply, but I want 
to mention something here that I think is a common mistake among 
developers.  Premature optimization.  As my CS 171 instructor put it, 
the definition of successful software product is one that is done on 
time and under budget.  Here are my priorities when coding:

1. Project Objectives
2. Security
3. My time(short term)
4. My time(long term)
5. Ease of use(unless this is a project objective, which in my case it 
usually is)
6. Computer Resources

I answer most of my questions like this by going down this list.  Most 
are answered at #3. With your question I'd invoke priority #4(less code 
to maintain), but it doesn't make that big of a difference.

There has been twice in my admittedly short(5 years) career where I've 
had to go back and rewrite a process because it wasn't efficient enough, 
and it wasn't that big of deal either time because I didn't spend a lot 
of time designing it the first time.

Now it is a good idea to ask about best practices and problems that 
other developers have ran into, but just a word of warning to new 
developers out there.  No matter how fast your product runs, and how few 
servers you need to serve millions of people. If you don't finish your 
product, it doesn't matter.


Kyle


More information about the UPHPU mailing list