Re: [go-nuts] go runtime in DLL not getting intialized(?) on some windows hosts

On Fri, May 10, 2019 at 1:43 PM Jason E. Aten wrote: > > I wondering if anyone can shed some light on > > a) how I might understand when the Go runtime inside a DLL is supposed to be > initialized; In c-shared mode the function _rt0_amd64_windows_lib is supposed to be called when the DLL is

[go-nuts] go runtime in DLL not getting intialized(?) on some windows hosts

I'm stumped on why the Go runtime would not be initialized inside my c-shared DLL. I'm building a mixed Go and C DLL for windows, using go build -buildmode=c-shared On one instance of Windows 2016, everything runs fine, both with go1.12.5 and with go1.10.8. However, on an instance of Windows

Re: [go-nuts] Using archive ".a" at time of build

On Fri, May 10, 2019 at 11:35 AM wrote: > > Thanks for update. > But than it is confusing, what is use of ".a" files ? > Where we can use it ? Especially when using module mode with Go 1.12 and future releases, ".a" files are an internal detail of how Go programs are built. You aren't expected

Re: [go-nuts] Using archive ".a" at time of build

Hi Ian, Thanks for update. But than it is confusing, what is use of ".a" files ? Where we can use it ? Regards, Rajesh -- You received this message because you are subscribed to the Google Groups "golang-nuts" group. To unsubscribe from this group and stop receiving emails from it, send an

[go-nuts] Re: Yet Another Generics Proposal...

So, my point is that operator naming should be standardized, or we might've introduce an additional construct like ``` type Neg operator - func() Neg type Compare operator <=> func(Compare) int ``` On Friday, May 10, 2019 at 9:15:30 PM UTC+3, Yuriy Yarosh wrote: > > So, there were plenty of

[go-nuts] Yet Another Generics Proposal...

So, there were plenty of generics proposal, and hopefully I'll get some constructive feedback about this one. Syntax Example: ``` type MyGeneric interface { int32 uint32 float32 float64 ANumber } type ANumber interface { Plus(ANumber) ANumber Minus(ANumber) Compare(ANumber) int

Re: [go-nuts] Using archive ".a" at time of build

On Fri, May 10, 2019 at 10:56 AM wrote: > > I have certain utility package which can be used by multiple projects. > Now in place of copying complete utility code in all of these projects at > time of building that project, I need to use archive file . > > So, I generated and it is placed in

[go-nuts] Using archive ".a" at time of build

Hi, I have certain utility package which can be used by multiple projects. Now in place of copying complete utility code in all of these projects at time of building that project, I need to use archive file . So, I generated and it is placed in $GOPATH/pkg/linux_amd64 [Note $GOROOT and

[go-nuts] Merging package files

Hello, I was wondering if there was an easy way to merge package files into a single one, like the bundle tool does but as a library. My goal is to take a package, transform some of its types (change their name) and output a file which

Re: [go-nuts] Timers and implementation of a timeout system on requests

Oh we are definitely not using a single core, haha. Every request is its own set of routines, being careful to not overspawn them though. Le vendredi 10 mai 2019 15:13:45 UTC+2, Robert Engels a écrit : > > One other point, at 5k req/s sustained each request must on average > complete in 200us

Re: [go-nuts] Timers and implementation of a timeout system on requests

The server has 12 cores (2 cpu*6 cores). The tolerance would be about +-10ms, I'd probably check how much time I have on the fly on every request, and remove 10-15ms just "for safety" if anything hangs for a bit too long. I just ran a few tests with a collegue using N simultaneous

Re: [go-nuts] Timers and implementation of a timeout system on requests

One other point, at 5k req/s sustained each request must on average complete in 200us (if using a single core) otherwise you will not make your deadlines and you will run out of memory. So 100 ms is way outside the threshold needed. Clearly by upping the parallelism you can increase the time

Re: [go-nuts] Timers and implementation of a timeout system on requests

That being said, 5000 requests per second is pretty low on any reasonable hardware. You can review github/robaho/go-trader - it does 30k requests per sec on desktop machines. > On May 10, 2019, at 7:58 AM, Robert Engels wrote: > > I don’t think your requirements are completely specified. For

Re: [go-nuts] Timers and implementation of a timeout system on requests

I don’t think your requirements are completely specified. For example, you say the timeout is 100ms - nothing is ever exact - what is the tolerance in the delay before it is cancelled ? Are the calls in the handler even cancelable? What type of hardware (64+ cores?) I think this is why you are

Re: [go-nuts] Re: http.ListenAndServe holding on to port after Process.Kill()

Coming to see in May 2019. Thank you for your help very much! 2012년 8월 23일 목요일 오전 10시 40분 53초 UTC+9, andrey mirtchovski 님의 말: > > Jesse is right. When you kill the 'go' command you don't kill the child. > Observe on OSX: > > $ go run server.go & > [1] 17246 > 2012/08/22 19:28:02 Listening on

Re: [go-nuts] Timers and implementation of a timeout system on requests

Good point on the implementing side of things, it's cleaner. I'm still really curious of the limits and implementation details - There has to be some kind of limit where things start to become erratic. If anyone wants to chime in :) Le jeudi 9 mai 2019 17:23:27 UTC+2, Burak Serdar a écrit : >

Re: [go-nuts] How to run a go method on multiple cores ?

Apologies Marvin. On Thu, May 9, 2019 at 6:53 PM Marvin Renich wrote: > * Nitish Saboo [190508 05:48]: > > Please remove me from the CC on this thread; I am subscribed to the > list. I explicitly set Reply-To, which was not respected, and now I am > getting duplicates of every mail, with one