libcurl.m4 defines LIBCURL_CPPFLAGS and CFLAGS would be something completely different as far as I understand it, so you're missing the m4 file from some package.

Ok, configure scripts are witchcraft to me.

I'm having trouble finding any interest in supporting older distributions and libraries in light of that knowledge limitation of mine. If you have patches, I'd be more than happy to accept them 

I'd say the cryptopp_asm32 on windows is likely broken.

Not only is it still there, but they even managed to port the bug to the linux driver as well!

Anyone on linux, do NOT move away from 11.6.

None of the variables from other miners are directly portable to cgminer. Presumably what you want is the dynamic mode OFF which happens whenever you set an intensity with -I (number).

Why 9? I thought I explained that. I get more mhashes now with 9 than 8. With previous versions I didn't.

Normally work gets leaked to backup pools if the primary pool is detected as lagging before it has actually failed. Enabling this new flag means no work will go to backup pools unless the primary one has failed completely. This was a requested feature by a donor.

Just as a quick data point, I am now recommending dedicated mining rigs use -I 9 instead of -I 8. Changes in the last few versions seem to make for a very small rise in throughput on 69xx cards with the higher intensity. As always, test for yourself as your mileage may vary. 

Interesting. The queued value is truly the "number of items this specific device queued" and there are other ways the queue is kept full that mean this device may not ever really request work on its own. So, strictly speaking, it's still correct under that description, but I can see why it looks odd.

Indeed as ancow said, I completely ignored that fact. bash DOES cache the location and if you remove it after you've already used it, it will stupidly continue to try and use that one.

Because the least work and most portable was to add the --prefix value to the directories cgminer checks. If you grab the precompiled binary it will be set to /usr, if you don't set --prefix it will be set to /usr, if you set --prefix it will be wherever you choose and it ends up when you do 'make install'. The last part was the most important.

That's because it has to have some default prefix if none is specified, and (perhaps foolishly) I chose /usr since most distros install everything there (even though most configure scripts default to /usr/local, I know). Ideally it should have a custom location for kernels, in something like $prefix/lib/cgminer or share/cgminer or some other crap, but I couldn't decide where would work cross platform and my care factor dropped to negative levels once I got it working at all.

This can probably be alleviated by increasing the -Q value (default is 1). This can be done while it's still running in ettings -> [Q]ueue. Experiment and see.

You must specify your install directory with --prefix= and compile it yourself if you want to install it somewhere specific.

Okay that is much crappier than I'd even imagined it would be on windows.

New release: Version 1.6.1, Links in first post.

Executive summary of user visible changes:
Hopefully fixed the zero sized binary bug \o/
Hopefully fixed the (old ncurses) resize crash
Hopefully fixed the memory leak in cpu mining and the inappropriately high rejects
Tidier build and smarter config and new --disable-opencl option
Less false positives for "unable to provide work fast enough"
New --shares option (mine X shares and exit)
New --enable-cpu and --disable-gpu options
Version info


Full changelog:
- Copy cgminer path, not cat it.
- Switching between redrawing windows does not fix the crash with old
libncurses, so redraw both windows, but only when the window size hasn't
changed.
- Reinstate minimum 1 extra in queue to make it extremely unlikely to ever have
0 staged work items and any idle time.
- Return -1 if no input is detected from the menu to prevent it being
interpreted as a 0.
- Make pthread, libcurl and libcurses library checks mandatory or fail.
- Add a --disable-opencl configure option to make it possible to override
detection of opencl and build without GPU mining support.
- Confusion over the variable name for number of devices was passing a bogus
value which likely was causing the zero sized binary issue.
- cgminer no longer supports default url user and pass so remove them.
- Don't show value of intensity since it's dynamic by default.
- Add options to explicitly enable CPU mining or disable GPU mining.
- Convert the opt queue into a minimum number of work items to have queued
instead of an extra number to decrease risk of getting idle devices without
increasing risk of higher rejects.
- Statify tv_sort.
- Check for SSE2 before trying to build 32 bit SSE2 assembly version. Prevents
build failure when yasm is installed but -msse2 is not specified.
- Add some defines to configure.ac to enable exporting of values and packaging,
and clean up output.
- Give convenient summary at end of ./configure.
- Display version information and add --version command line option, and make
sure we flush stdout.
- Enable curses after the mining threads are set up so that failure messages
won't be lost in the curses interface.
- Disable curses after inputting a pool if we requested no curses interface.
- Add an option to break out after successfully mining a number of accepted
shares.
- Exit with a failed return code if we did not reach opt_shares.
- The cpu mining work data can get modified before we copy it if we submit it
async, and the sync submission is not truly sync anyway, so just submit it sync.


---
If you're wondering why I include the entire build in the linux binary tarball, it's so you can do 'make install' to install it into a system location. 

That's the new 32 bit assembly cpu mining code which is only being half compiled. A workaround for the moment is to add -msse2 to your CFLAGS till I can figure out how to fix it.

It only uses a total amount of CPU that amounts to 100% of one CPU core due to the strange way the ATI driver is written, and binding the application to one core just means all the work will go to that one core; it does not actually decrease CPU usage.

I believe slush doesn't have longpoll. cgminer detects it by default and tries to enable it. Luckily cgminer is very good at minimising rejects even without longpoll.

Best kernel is likely phatk, and I'm pretty sure it will use that anyway by default.

One by one every mining software is being tagged as a virus by the stupid virus software makers because the mining software is being packaged with a trojan set of parameters. Obviously that's not the mining software's fault, so the virus software makers can go and get fffff

There is no official terminology for any of this stuff, but I'm assuming what ask-rate is, I guess the answer is yes.