Looks great, I'm verifying the c/p now. Several dependencies on boost could be dropped, I assume you're leaving that for a follow-up?

Verfied the code movement with c/p, all is clean with 2 exceptions. I'm only mentioning these for the sake of being thorough, I assume they're harmless.

GetTimeMicros and MilliSleep have moved from inline->non-inline. I have no clue if they actually ended up being inlined for real when compiled before, but if so, that may mean a functional change here due to the timing/precision involved.

That aside, ACK.

If you see any further dependencies on boost - especially includes in headers - that can be removed let me know. This has the overall effect of including less boost headers per compilation unit which means faster compile times.

And indeed: GetTimeMillis/Micros are no longer inline because they pull in a compile-time dependency on boost::posix_time. Having them implemented in utiltime.cpp means that the boost usage is at least isolated. It could theoretically affect timings - but what's one extra function call in microseconds, who knows what boost::posixtime::* is doing for awful stuff internally. To reliably benchmark small times one should use CPU performance counters (and yes, inlined :-). But IIRC we only benchmark larger time-spans.

PASSED! Hah, take that @BitcoinPullTester (for some reason the pulltester environment requires some more explicit #includes of boost/thread.hpp, so had to add those one by one, sorry for the spam)

@laanwj re: inlines, sounds good.

As for the boost deps, I'm not sure if there are others, but I noticed this one right away in utilstring.cpp

bool IsHex(const string& str)
{
 BOOST_FOREACH(char c, str)
 {
 if (HexDigit(c) < 0)

->

bool IsHex(const string& str)
{
 for(std::string::const_iterator c(str.begin()); c != str.end(); ++c)
 {
 if (HexDigit(*c) < 0)

That would mean dropping boost completely for that object.

@theuni I thought about that, although I like keeping FOREACH, so that they can be easily replaced when the time for c++11 comes. boost/foreach.hpp is header-only.

I havent verified the c/p yet (will do that after squash before merge), but you've definitely got my conceptual ACK. I might consider changing utilmoney to uilmoneystr or just moneystr, utilstring to utilstrencodings or something and splitting the new util up further into something like arguments/env and logging (but that may be better in a separate pull because some of it isnt clear).

Untested ACK; I didn't verify it to be move-only.

@TheBlueMatt I'm not going to do any further splitting in this pull. Those renames sound fine to me though.

ut (but verified cp) ACK

ut ACK

"utilstrencodings" is a lame name. utilstrcode or strutil ?

Untested ACK, but verified move-only.

I will refrain from joining the filename bikeshedding.

@jgarzik It was just utilstring, then I renamed it after @TheBlueMatt complained about the name once already. I won't keep renaming them. Go fight it out with him first..

@jgarzik now its bikeshedding, I dont care either way, it was just a suggestion.

Automatic sanity-testing: PASSED, see http://jenkins.bluematt.me/pull-tester/p4748_ad49c256c33bfe4088fd3c7ecb7d28cb81a8fc70/ for binaries and test log.
This test script verifies pulls every time they are updated. It, however, dies sometimes and fails to test properly. If you are waiting on a test, please check timestamps to verify that the test.log is moving at http://jenkins.bluematt.me/pull-tester/current/
Contact BlueMatt on freenode if something looks broken.

Merging, as this will likely be outdated soon otherwise.

File renames can easily be done later if necessary.