Re: [PATCH] APR on WinCE 5 - trunk

[Ioan Popescu <ip...@dataq.com>]

William A. Rowe, Jr. wrote:
> Ioan Popescu wrote:
>> Why is "WINNT" defined in the project file? What does it "mean"? It 
>> wasn't explicitly defined before, why is it now?
>> 
>> Just checking whether I need this for CE...
> 
> I don't think you do; we look at each.
> 
> WINNT triggers some shorthands to melt away Win9x code from the build 
> when it's contained within the various ELSE_WIN_OS_IS_ANSI blocks, and 
> eliminating the conditional evaluation behind IF_WIN_OS_IS_UNICODE, but I
> believe those macros are already triggered by _WINCE.

You're right. I don't need to worry about WINNT.

I've checked the libapriconv and libaprutil patches and both apply cleanly
to trunk. Not all patches to libapr are clean, but the ones that are, I've
reattached. I've also found a couple errors in my own previous patches to
libapr and updated the use of _WIN32_WCE. APIs existing before or including
CE 3 default to checking whether it's defined. APIs existing afterwards are
used only if they exist (depends on _WIN32_WCE version).

Still working on the files that won't merge because they're too different. I
have to do them by hand. Probably won't have anything until next week.

Re: [PATCH] APR on WinCE 5 - trunk

[Ioan Popescu <ip...@dataq.com>]

Ioan Popescu wrote:
> I've checked the libapriconv and libaprutil patches and both apply cleanly
> to trunk. Not all patches to libapr are clean, but the ones that are, I've
> reattached. I've also found a couple errors in my own previous patches to
> libapr and updated the use of _WIN32_WCE. APIs existing before or including
> CE 3 default to checking whether it's defined. APIs existing afterwards are
> used only if they exist (depends on _WIN32_WCE version).

Here's the rest.