aboutsummaryrefslogtreecommitdiff
path: root/src/stable/randomcmwc.cpp (unfollow)
AgeCommit message (Collapse)Author
2012-08-29More header/include fun on windows.Mike Buland
2012-08-28Made a function that should be private private, and documented yield. Also,Mike Buland
tested everything just to be safe :).
2012-08-28Loads of win32 compilation issues fixed. Most are fairly minor unsigned/signedMike Buland
mismatches because of socket handles, but there were also some order-of-definition issues that were fixed in the FD_SETSIZE definition code. Fixed a few things that just never worked on windows, like Bu::Thread::yield().
2012-08-28INI driver is loading and saving just fine. Ideally it would escape thingsMike Buland
like ']', '=', and possibly some others just to make all characters safe to use. Right now, just don't use ] or = in keys and you're good.
2012-08-24Ini driver is close, loading isn't working perfectly yet...Mike Buland
2012-08-24The basic registry interface works now.Mike Buland
2012-08-24Taf isn't really...the best backend for this now, I need some more fixes thereMike Buland
first, so next up we need the windows registry backend.
2012-08-24Adding new Bu::Settings system, which works more or less like QSettings from Qt,Mike Buland
only...without qt. It's the first real system in libbu++ that is going to be all unicode, which is exciting, but I'm having to actually finish the Bu::UtfString class while I work on it. Also exciting.
2012-08-18Added hashtable += support, optomized the random number base some?Mike Buland
2012-08-14stupid...spelling...Mike Buland
2012-08-08Had to make FD_SETSIZE much, much larger. It looks like the way to go in theMike Buland
future is using overlapped I/O in windows, and who knows in linux. I believe the select method is always a good fallback, but apparently linux doesn't support posix aio (asynchronous i/o) in the kernel, it emulates it with threads, so who knows what the right choice is right now.
2012-08-08Fixed FD_SETSIZE in libbu++ for windows...hopefully.Mike Buland
2012-08-07Nevermind, windows has no poll, and it's select sucks...Mike Buland
2012-08-07Adding filesystem access classes, and converting Bu::Server to using ppoll.Mike Buland
I have to test to see if it works with mingw though.
2012-07-31Uuid had an == operator but not !=Mike Buland
2012-07-30Added helpers: Bu::List::eraseFirst and Bu::List::eraseLast. Pretty selfMike Buland
explanatory.
2012-07-24Fixed long-long specifier on basic random number generator.Mike Buland
2012-07-24Apparently I misspelled something.Mike Buland
2012-07-19Included compat dir in makefile for windows.Mike Buland
2012-07-19Fix to the makefile for windows.Mike Buland
2012-07-18Fixed an issue in the seldomly used dequeue( bool ) function inMike Buland
Bu::SynchroQueue and also added the unblockAll function, which makes the forever-blocking dequeue function suddenly much more useful.
2012-07-18Bu::SynchroQueue had a bug, it seems to be all fixed now.Mike Buland
2012-07-09RandomCmwc was far too static before, now it's seeded with the Basic generator.Mike Buland
2012-06-26Changed the Bu::Uuid generate interface. It's now one function that takes aMike Buland
type parameter. I think it's much nicer overall, the old gen function is stil there for now, but marked deprecated.
2012-06-25Bu::Uuid has some more features for niceness. It's easier to set them fromMike Buland
strings and you can archive them.
2012-06-20Silly RandomCmwc, it was using a static variable somewhere. That's fixed now.Mike Buland
2012-06-20Whew, the cipher was writing blocks at close when no data had been written yet.Mike Buland
It only caused a problem when you stopped part way through reading something.
2012-06-19Changed the macro tests around, it wasn't working on win32...because I did itMike Buland
wrong :-P
2012-06-19Bu::Uuid now uses windows native uuid generation when on windows. Pretty cool.Mike Buland
It looks like they put some real effort into it too, it uses macs and everything. You have to link against Rpcrt4.lib/dll i.e. -lRpct4
2012-06-16Minor tweak to blowfish, it now flushes correctly on close, which makes stackingMike Buland
work better.
2012-06-16Cipher will fill with zeros when flushed. There seems to be some oddMike Buland
interaction still, here and there. I'll see if I can track it down.
2012-06-15Bu::Myriad now only uses BitString during initialization, and I'm going toMike Buland
replace that with just an array, no problem. It's many, many, many times faster while streams are growing, and it should be constant time, not linear like it was before. It also handles myriad files in excess of 2gb correctly now, at least, it seems to just fine :)
2012-06-04UtfString supports get and append now, but it still needs a lot of work.Mike Buland
2012-05-25Fixed a *very* rare steady-state issue in rehashing code. There's stillMike Buland
potential for probing to fail in a way that can't be fixed by rehashing, but it should be amazingly rare.
2012-05-24bin2cpp supports lzma compression. I need to modify the headers so it onlyMike Buland
includes headers for the things you use.
2012-05-23We have a nice selection of basic randomness functions now.Mike Buland
2012-05-22Moved random to stable, just needs some minor tweaks. But it's already in useMike Buland
in a couple of core components, including in tempFile name generation.
2012-05-22Better win_o ignores. The random number system is pretty much together.Mike Buland
We need a few extra helper functions to cover some other good things, like normalized floating point numbers, etc.
2012-05-22Added libbu++ endianness support. It's probably not as fast as systemMike Buland
endianness, but on gnu it'll use the libc versions, if you don't have those, at least libbu++ will compile.
2012-05-16Lzma was still printing debugging on "inflate".Mike Buland
2012-05-10Amazing this made it through for so long, but destroying a Bu::Buffer in writeMike Buland
mode didn't flush it's buffers.
2012-05-07Fixed the CMWC random number generator, it compiles now.Mike Buland
2012-05-04Two basic random number generators, Cmwc is supposed to be a pretty good one.Mike Buland
I need to get the base class and singleton interface in place.
2012-05-04Minor changes, mostly comments.Mike Buland
2012-05-03The cipher base class does nice buffering now based on the blocksize. ThisMike Buland
means that you can write odd numbers of bytes or read odd numbers of bytes from/to the ciphers and it'll just buffer until it gets a full block. Next up is adding padding support. Right now it doesn't flush the buffers with padding if you don't fill the last block. For that matter, it doesn't do any padding at all.
2012-05-03Rearranging some stuff.Mike Buland
2012-04-17Fixed bug in bzip2 double-stop.Mike Buland
2012-04-15Updated copyright, packaging, makefile, etc.Mike Buland
2012-04-14Added support for running subsets of unit tests to Bu::UnitSuite, now just listMike Buland
the names of the tests you want to run on the command line. Also, fixed some minor issues in two of the test suites.
2012-04-14The core of myriad is now thread-safe. It could use a little work on multipleMike Buland
streams accessing the same blocks at the same time. Right now it's safe, but a little strange, since both streams work from seperate buffers. Maybe that's ok, but it still needs some more investigation. I want to remove the BitString stuff entirely, it turns out it's the slowest part, which, upon actually looking at the code is completely obvious. This change shouldn't change the API at all, just make adding blocks to streams much, much faster.