Re: Re: Case sensitivity: Conclusion(?)

At 09:06 PM 2/8/2002 +0900, Yasuo Ohgaki wrote:

Marko Karppinen wrote:

Yasuo:

Hmm. I vote -1 for this. It just does not make sense to store original(case sensitive) names while langage ignores case. It's also confusing, lead to case sensitivity BC problem anyway just like with case sensitive function/names.

Case preservation makes very much sense in a case-insensitive environment. Just look at file system implementations from Apple and Microsoft, where, believe it or not, a lot of thought has gone into this issue. I had a private chat with Jason Greene about this, and I came out with the understanding that the .NET/SOAP interop issues mentioned here earlier by Markus and Jason are actually issues with PHP not preserving case rather than with PHP not being fully case-sensitive.

It will be confusing PHP preserve case for other while internally case insensitve... Therefore, -1 for this :)

It won't be confusing. You are just -1 because you want case sensitivity. How will it confuse you? The only thing it will allow is to support interoperability. It wouldn't confuse people writing PHP code. I don't mind you being against case insensitivity. You can have whatever opinion you want but it doesn't mean you need to find bad reasons :)

And since Andi promised to take a shot at making case preservation happen, it seems like we are on our way to better interoperability without too much BC hassles.

If get_class() or like returns names with case preserved, it breaks BC. I expect additional parameter for all of these functions if it really implemented.

Stig answered this one.

I hope not, since it's a additional memory and performance loss...

There is no chance in hell you'd notice the additional memory :) And the performance difference would be negligible because what can be done at compile-time is already done in ZE2. Andi