Re: [ANN] Neanderthal 0.6.0: new support for AMD, Nvidia, and Intel GPUs on Linux, Windows and OS X (fast matrix library)

Neanderthal 0.7.0 released: Nvidia, AMD, and Intel GPU on LInux, Windows, and OS X http://neanderthal.uncomplicate.org#matrix #Clojure #GPU #GPGPU On Monday, May 23, 2016 at 9:57:00 PM UTC+2, Dragan Djuric wrote: > > This is a major release of Neanderthal, a fast native & GPU matrix libra

Re: Can't extract multipart-params from POST request in compojure

You are right! I didn't set the boundary in my client code. Thanks a lot! 在 2016年7月7日星期四 UTC+8下午12:46:04，James Reeves写道： > > By "application" I mean "Ring handler". > > Your request looks wrong. The boundary you set in your header > (--BoUnDaRyFF) doesn't match the boundary I see in the b

Re: Can't extract multipart-params from POST request in compojure

By "application" I mean "Ring handler". Your request looks wrong. The boundary you set in your header (--BoUnDaRyFF) doesn't match the boundary I see in the body (--2n639DJeu...). - James On 7 July 2016 at 05:29, wrote: > The request header: > > > >

Re: Can't extract multipart-params from POST request in compojure

The request header: The request body:

Re: Can't extract multipart-params from POST request in compojure

In all your examples except for possibly the last, you have the multipart middleware added twice over to your application. Also, are you sure it's not a problem with your client code? One thing to try is recording your HTTP request by pointing your client at a netcat listener (e.g. "nc -l 3000").

Re: Can't extract multipart-params from POST request in compojure

This is the link to the question in stackoverflow: http://stackoverflow.com/questions/38226276/cant-extract-multipart-params-from-post-request-in-compojure 在 2016年7月7日星期四 UTC+8上午10:13:25，aaartu...@gmail.com写道： > > When I try to extract the multipart-params from a POST request like this: > > (defro

Can't extract multipart-params from POST request in compojure

When I try to extract the multipart-params from a POST request like this: (defroutes upload-routes (POST "/upload" {params :params} (println params)) I got {}. Then I tried like this: (defroutes upload-routes (POST "/upload" {multipart-params :multipart-params} (println multipart-params))

Re: Get Lein to run 32-bit JVM on Windows

My solution was to dual-boot linux. I think it's about time anyway. On Tuesday, 5 July 2016 12:17:02 UTC-7, JvJ wrote: > > I think when I wrote LEIN_CMD, I meant LEIN_JAVA_CMD. I've already set it > to the x86 JVM path, with no luck so far. > > (System/getProperty "os.arch") still returns "amd6

Re: [ANN] akar 0.1.0

Hi Rahul, Akar seems very nice! It works in Clojurescript? Saludos, Nahuel Greco. On Tue, Jul 5, 2016 at 11:38 AM, Rahul Goma Phulore < rahul.phulore@gmail.com> wrote: > Hi, all. > > I am quite excited to announce my project Akar – > https://github.com/missingfaktor/akar. It's a pattern matc

Re: [ANN] Clojure 1.9.0-alpha9

There will soon be a spec for let that will fail in this circumstance, letting you know. On Wednesday, July 6, 2016 at 7:42:22 AM UTC-5, Leon Grapenthin wrote: > > 1.9 could still support the behavior that worked in 1.8 and in CLJS, too, > or throw a compile time Exception. > > If we had known

Re: [ANN] Clojure 1.9.0-alpha9

1.9 could still support the behavior that worked in 1.8 and in CLJS, too, or throw a compile time Exception. If we had known this change in behavior, fixing it in the codebase wouldn't have been much trouble. But a missing default can lead to errors far away so it takes time to get to the grou

Re: [ANN] Clojure 1.9.0-alpha9

This *is* different than 1.8 (and prior) because it has been buggy since namespaced keys were added (due to a lack of tests - entirely my fault). Nothing was ever documented about this behavior - it was purely accidental and I would consider it in the realm of "unspecified". In 1.9 I added addi

Re: [ANN] Clojure 1.9.0-alpha9

I'd say that's a bug with 1.8 rather than 1.9 > On 6 Jul 2016, at 13:01, Leon Grapenthin wrote: > > @Alex Miller: Please reconsider. What you state does not apply to 1.8, i. e. > the ns is required in :or and such code is broken in 1.9 > > 1.8/1.9 differences: > (let [{:keys [a/b] :or {b 42}}

Re: [ANN] Clojure 1.9.0-alpha9

@Alex Miller: Please reconsider. What you state does not apply to 1.8, i. e. the ns is required in :or and such code is broken in 1.9 1.8/1.9 differences: (let [{:keys [a/b] :or {b 42}} {}] b) => 1.8: nil 1.9: 42(let [{:keys [a/b] :or {a/b 42}} {}] b) => 1.8: 42 1.9: nil On Wednesday, J

Re: [ANN] Clojure 1.9.0-alpha9

Thanks! If I could trouble you to make a jira for that, would love to get it fixed. > On Jul 6, 2016, at 4:09 AM, Leon Grapenthin wrote: > > Small bug report > > Throwable->map in core_print.clj doesn't handle Throwable.getCause returning > null in L463. This results in a NPE in StrackTraceEl

Re: [ANN] Clojure 1.9.0-alpha9

This is the correct behavior. The :or map keys are always the local names being bound (not the keys being looked up in the map). > On Jul 6, 2016, at 4:40 AM, Leon Grapenthin wrote: > > Another small bug report: > > (let [{:keys [a/b] :or {a/b 42}} {}] > b) > > Evaluates to nil, but should

Re: [ANN] Clojure 1.9.0-alpha9

Another small bug report: (let [{:keys [a/b] :or {a/b 42}} {}] b) Evaluates to nil, but should evaluate to 42 (let [{:keys [a/b] :or {b 42}} {}] b) Evaluates to 42, but should evaluate to nil. On Tuesday, July 5, 2016 at 9:50:47 PM UTC+2, Alex Miller wrote: > > Clojure 1.9.0-alpha9 is n

Re: [ANN] Clojure 1.9.0-alpha9

Small bug report Throwable->map in core_print.clj doesn't handle Throwable.getCause returning null in L463. This results in a NPE in StrackTraceElement->vec in the same file in some cases, so printing a stacktrace results in a new Exception which is a bit confusing. On Tuesday, July 5, 2016 at

Re: Get Lein to run 32-bit JVM on Windows

If system/getProperty "os.arch" says it's 64 bits, it means the processor is a 64 bits compatible processor, not that the Java process is 64 or 32 bits. If you want to see what the JVM data model is you need to use (System/getPropery "sun.arch.data.model") and see what it returns (either "32"