Re: Beta 2.084.1
On Tuesday, 5 February 2019 at 04:18:55 UTC, Martin Nowak wrote: Glad to announce the first beta for the 2.084.1 point release, ♥ to the 6 contributors. http://dlang.org/download.html#dmd_beta http://dlang.org/changelog/2.084.1.html As usual please report any bugs at https://issues.dlang.org -Martin This beta again has this issue: lld-link.exe: error: could not open msvcrt100.lib: no such file or directory lld-link.exe: error: could not open OLDNAMES.lib: no such file or directory (It is caused if you have a visual studio / build tools installation, then lld gets confused). It was already solved in the nightly build, I downloaded some days ago. Kind regards André
Re: Issues Donating or Registering for DConf via Flipcause
On Tuesday, 29 January 2019 at 02:29:08 UTC, Mike Parker wrote: Flipcause released a major upgrade to their platform last week that unified several different components (donations, events, online stores, etc) into a single system they call Universal Checkout. [...] No chance... I updated my Master Card with the Secure Code thing, but the same result again. I tried again 2 times: only result => Your session has expired. Flipcause has been a better usage experience before. Sure I can give my 25 bucks via Paypal, but this is not the idea behind, isn't it?
Re: Issues Donating or Registering for DConf via Flipcause
On Tuesday, 5 February 2019 at 11:17:01 UTC, Martin Tschierschke wrote: Flipcause has been a better usage experience before. Sure I can give my 25 bucks via Paypal, but this is not the idea behind, isn't it? Go ahead and go through PayPal if Flipcause doesn't work for you. They're working on the issues right now. They've credited our account with some money and have been running tests on our account and others. They'll work it out eventually, I'm sure.
a van Emde Boas tree
Hi all, my van Emde Boas tree finally reached an announceable state, at version 0.12.0. It operates on a closed universe, defined on construction. After that, the tree operates on unique key up to a certain capacity, which is at least as large as the universe size. All operations including insertion, removing, look up and next neighbor search are O(log(log(U)). min and max lookup are O(1). Some graphics are included on the front page for performance comparison with existent libraries. I relicensed the package to BSL-1.0 and moved it on github, so the documentation should work properly now. All tickets are welcome of course and will be managed in my spare time.
Re: a van Emde Boas tree
On Tuesday, 5 February 2019 at 15:28:04 UTC, Alex wrote: Hi all, my van Emde Boas tree finally reached an announceable state, at version 0.12.0. clip All tickets are welcome of course and will be managed in my spare time. Link?
Re: a van Emde Boas tree
On Tuesday, 5 February 2019 at 15:28:04 UTC, Alex wrote: Hi all, my van Emde Boas tree finally reached an announceable state, at version 0.12.0. vEB tree is an interesting data structure. Where is the implementation? - You did not provide any links...
Re: a van Emde Boas tree
On Tuesday, 5 February 2019 at 15:47:30 UTC, Dejan Lekic wrote: On Tuesday, 5 February 2019 at 15:28:04 UTC, Alex wrote: Hi all, my van Emde Boas tree finally reached an announceable state, at version 0.12.0. vEB tree is an interesting data structure. Where is the implementation? - You did not provide any links... Most likely the repo is this one: https://github.com/Sandman83/vebtree
Re: a van Emde Boas tree
Dne 5.2.2019 v 16:47 Dejan Lekic via Digitalmars-d-announce napsal(a): > On Tuesday, 5 February 2019 at 15:28:04 UTC, Alex wrote: >> Hi all, >> my van Emde Boas tree finally reached an announceable state, at >> version 0.12.0. > > vEB tree is an interesting data structure. Where is the implementation? > - You did not provide any links... I guess it is: https://code.dlang.org/packages/vebtree
Re: a van Emde Boas tree
On Tuesday, 5 February 2019 at 16:04:03 UTC, Petar Kirov [ZombineDev] wrote: On Tuesday, 5 February 2019 at 15:47:30 UTC, Dejan Lekic wrote: On Tuesday, 5 February 2019 at 15:28:04 UTC, Alex wrote: Hi all, my van Emde Boas tree finally reached an announceable state, at version 0.12.0. vEB tree is an interesting data structure. Where is the implementation? - You did not provide any links... Most likely the repo is this one: https://github.com/Sandman83/vebtree Yes... sorry. Too deep in the code... https://code.dlang.org/packages/vebtree
Re: Beta 2.084.1
On Tuesday, 5 February 2019 at 04:18:55 UTC, Martin Nowak wrote: Glad to announce the first beta for the 2.084.1 point release, Any chance of getting this in? https://github.com/dlang/phobos/pull/6814 Phobos std.file has become unusable after 2.082 because of [1] leading to memory corruption. It's pretty bad: dmd 2.082 and 2.083, and LDC 1.12 and 1.13 cannot be used for serious work (unless you prevent people from using std.file by deleting it?). Cheers, Johan [1] the trigger of the bug: https://github.com/dlang/phobos/pull/6584
Gillespie algorithm
Doing some physical simulations I could abstract the Gillespie algorithm, see https://code.dlang.org/packages/gillespied which may be useful. The number of algorithms able to model physical (continuous) time during an ongoing (reaction) process is rather limited. The announced algorithm is known to provide this feature. There are very much manifestations and improvements of the original Gillespie. In general, the improvements try to improve the speed of the algorithm either via - faster timing generation - faster next reaction look up - approximating the output by assumptions Compared to the standard Gillespie algorithm two enhancements were implemented: - for the case, reaction propensities are known, interarrival timings do not necessarily depend on a random number and its logarithm. This makes each time query faster. - If the governing simulation can handle non-nogc algorithm instantiation, a dedicated working array can be handled inside the algorithm. This provides faster next reaction search for all operations. All issues are of course welcome and will be tackled in my spare time. I hope, for this library, many further improvements, and applications could be found in future.
intel-intrinsics v1.0.0
"intel-intrinsics" is a DUB package for people interested in x86 performance that want neither to write assembly, nor a LDC-specific snippet... and still have fastest possible code. Available through DUB: http://code.dlang.org/packages/intel-intrinsics *** Features of v1.1.0: - All intrinsics in this list: https://software.intel.com/sites/landingpage/IntrinsicsGuide/#techs=MMX,SSE,SSE2 Use existing Intel documentation and syntax - write the same code for both DMD and LDC, in the last 6 versions for each. (Note that debug performance might suffer a lot when no inlining is activated.) - Use operators on SIMD vectors as if core.simd were implemented on DMD 32-bit - Introduces int2 and float2 because short SIMD vectors are useful - about 6000 LOC (for now! more to come) - Bonus: approximated pow/exp/log. Perform 4 approximated pow at once. The long-term goal for this library is to be _only about semantics_, and not particularly codegen(!). This is because LLVM IR is portable, so forcing a particular instruction is undoing this portability work. **This can seem odd** for an "intrinsics" library but this way exact codegen options can be choosen by the library user, and most intrinsics can gracefuly degrade to portable IR in theory. In the future, "magic" LLVM intrinsics will only be used when built for x86, but I think all of it can become portable and not x86-specific. Besides, there is a trend in LLVM to remove magic intrinsics once they are doable with IR only. tl;dr you can use "intel-intrinsics" today, and get quite-optimal code with LDC, without duplication. You may come across early bugs too. http://code.dlang.org/packages/intel-intrinsics (note: it's important to bench against vanilla D code or arrays ops too, in some case the vanilla code wins)
Re: Beta 2.084.1
On Tuesday, 5 February 2019 at 17:41:50 UTC, Johan Engelen wrote: On Tuesday, 5 February 2019 at 04:18:55 UTC, Martin Nowak wrote: Glad to announce the first beta for the 2.084.1 point release, Any chance of getting this in? https://github.com/dlang/phobos/pull/6814 -> https://github.com/dlang/phobos/pull/6855
Re: intel-intrinsics v1.0.0
On Wednesday, 6 February 2019 at 01:05:29 UTC, Guillaume Piolat
wrote:
"intel-intrinsics" is a DUB package for people interested in
x86 performance that want neither to write assembly, nor a
LDC-specific snippet... and still have fastest possible code.
Neat. Question: On Github it's stated that implicit conversions
aren't supported, with this example:
__m128i b = _mm_set1_epi32(42);
__m128 a = b; // NO, only works in LDC
Couldn't this be solved through something like this:
struct __m128 {
float4 value;
alias value this;
void opAssign(__m128i rhs) {
value = cast(float4)rhs.value;
}
}
--
Simen
