Le jeudi 26 janvier 2017 07:42:55 UTC+1, JohnGB a écrit :
>
> Thanks Pierre, that looks like a simple and quite clean implementation.
> Although it's easier to use an iota for the numbering, does that open up
> any issues with keeping the documentation up to date? I know it's easier
> to pro
On Fri, Jan 27, 2017 at 6:25 AM Aram Hăvărneanu wrote:
> On Solaris the syscall interface is a private, unstable interface
(just like on Windows), and the only supported way to interact with
the system is through the vendor-supplied libc. Go on Solaris always
uses libc, it doesn't make system cal
Arafath, that's kind of what I did (except I replaced the git command with
my setversion.exe). It worked as expected... thank you!
On Friday, January 27, 2017 at 12:43:01 AM UTC-6, Arafath M wrote:
>
> Hi Eric,
>
> Under Windows, create a bat file with following contents. It uses latest
> tag f
Arafath, that's kind of what I did (expect I replaced the git command with
my setversion.exe). It worked as expected... thank you!
On Friday, January 27, 2017 at 12:43:01 AM UTC-6, Arafath M wrote:
>
> Hi Eric,
>
> Under Windows, create a bat file with following contents. It uses latest
> tag f
Hi Eric,
Under Windows, create a bat file with following contents. It uses latest
tag from git repo as version.
echo Rebuilding App...
for /f %%i in ('git describe --tags ') do set version=%%i
go install -a -v -ldflags "-X main.versionBuild=%version%"
Sincerely,
Arafath M
On Fri, Jan 27, 2017
Got it to work as intended under Linux; however, compiling on Windows is a
no-go. Perhaps I have to create a script, pass the output to a variable,
and insert the variable into the build command? Hopefully not, so if
anybody has any suggestions... it would be highly appreciated. Thanks!
On T
Hi Jakob.
> I don't know the OP's reason, but environment variables are somewhat
> more cumbersome on Windows. I don't think there is a way (in cmd) to
> do both in one command, for example
I believe that this might work. Modifying Dave's example:
Set (GOOS=Darwin) & go build
The () p
On Thu, Jan 26, 2017 at 9:24 PM, Aram Hăvărneanu wrote:
> On Thu, Jan 26, 2017 at 7:00 PM, Jan Mercl <0xj...@gmail.com> wrote:
>> or any POSIX-like OS for that matter
>
> On Solaris the syscall interface is a private, unstable interface
> (just like on Windows), and the only supported way to inter
On Thu, Jan 26, 2017 at 7:00 PM, Jan Mercl <0xj...@gmail.com> wrote:
> or any POSIX-like OS for that matter
On Solaris the syscall interface is a private, unstable interface
(just like on Windows), and the only supported way to interact with
the system is through the vendor-supplied libc. Go on So
How often your application changes GOMAXPROCS at runtime?
What is the reason of this change?
What patterns of this change exist (ie is it bounded, is it frequent, etc)?
--
You received this message because you are subscribed to the Google Groups
"golang-nuts" group.
To unsubscribe from this grou
On Thu, Jan 26, 2017 at 7:53 PM, T L wrote:
>
> On Friday, January 27, 2017 at 4:05:03 AM UTC+8, Ian Lance Taylor wrote:
>>
>> On Thu, Jan 26, 2017 at 11:10 AM, T L wrote:
>> >
>> > I still don't understand what are implicit memory allocations, could you
>> > make an explanation?
>>
>> An example
Using -ldflags -X, is there a way to set the internal value as the output
of an executable console file?
I've seen examples like this:
"-X main.githash=`git rev-parse HEAD`" (which would put the result of the git
command w/ flags into it)
however, I just can't get it to work (if it's even pos
On Friday, January 27, 2017 at 3:38:12 AM UTC+8, Jakob Borg wrote:
>
> It depends. A call to new can (and will, often) give you a pointer to a
> stack allocated object. If you pass that pointer to something like
> fmt.Println() you'll see the object escape to the heap, get flagged by -m,
> and
On Friday, January 27, 2017 at 4:05:03 AM UTC+8, Ian Lance Taylor wrote:
>
> On Thu, Jan 26, 2017 at 11:10 AM, T L >
> wrote:
> >
> > I still don't understand what are implicit memory allocations, could you
> > make an explanation?
>
> An example of an implicit memory allocation:
>
> var gl
Hello,
I'm trying to call Linux SCSI generic ioctls from Go, ideally without
resorting to cgo, and have hit a few snags.
I have ported the sg_io_hdr_t struct () to Go, and according to
http://golang-sizeof.tips, it is the same size as sizeof(sg_io_hdr_t) in C,
and does not contain any unexpect
I reproduced this issue several times on Jan 24, but for some reason
that still eludes me right now, I can’t reproduce it anymore. Must be
PEBCAK. I’ll report back if I can reproduce it again and have some
evidence to share.
On Thu, Jan 26, 2017 at 1:27 PM, Aurélien Bombo wrote:
> Any update on
Any update on this? I'm curious.
Aurélien Bombo
2017-01-24 23:20 GMT+01:00 tsuna :
> Hi there,
> I have a linux/386 production binary that I compiled and deployed with
> go1.8rc2 and I noticed that the RSS is 2.5x of the same binary doing
> the same thing but compiled with go1.7.3. Anybody e
Hello gophers,
We have just released go1.7.5, which includes fixes to the compiler,
runtime, and the crypto/x509 and time packages.
https://golang.org/doc/devel/release.html#go1.7.minor
You can download binary and source distributions from the Go web site:
https://golang.org/dl/
We also released
Pro tip: optimise for correctness before performance.
--
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 email
to golang-nuts+unsubscr...@googlegroups.com.
For more options,
Thanks all, I don't use defer in a lot of the places to optimise and
minimise the lock time. I suppose looking at those points is probably the
safest place to start!
On Thursday, January 26, 2017 at 8:07:41 PM UTC, Dave Cheney wrote:
>
> Start with the basics like go vet which will spot a lock
Start with the basics like go vet which will spot a lock being copied by value.
Then check for each place you call Lock, there is a defer statement on th next
line.
--
You received this message because you are subscribed to the Google Groups
"golang-nuts" group.
To unsubscribe from this group
SET GOOS=Darwin
go build
--
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 email
to golang-nuts+unsubscr...@googlegroups.com.
For more options, visit https://groups.google.c
On Thu, Jan 26, 2017 at 11:10 AM, T L wrote:
>
> I still don't understand what are implicit memory allocations, could you
> make an explanation?
An example of an implicit memory allocation:
var globalVar *int
func f() {
var i int
globalVar = &i
}
Another one:
var globalVar interface{}
It depends. A call to new can (and will, often) give you a pointer to a stack
allocated object. If you pass that pointer to something like fmt.Println()
you'll see the object escape to the heap, get flagged by -m, and presumably
forbidden by -+.
//jb
> On 26 Jan 2017, at 20:10, T L wrote:
>
I don't know the OP's reason, but environment variables are somewhat more
cumbersome on Windows. I don't think there is a way (in cmd) to do both in one
command, for example. Maybe in Powershell, I don't speak it, but commands tend
to look quite verbose there as well.
//jb
> On 26 Jan 2017, at
On Friday, January 27, 2017 at 2:05:02 AM UTC+8, Ian Lance Taylor wrote:
>
> On Thu, Jan 26, 2017 at 9:18 AM, T L >
> wrote:
> >
> > On Wednesday, January 25, 2017 at 5:16:07 AM UTC+8, Ian Lance Taylor
> wrote:
> >
> >> By not allocating memory. It is helped by a "secret" compiler option,
"BSD" isn't a valid OS. It's a family of OSes.. but why don't you just
set the environment variable while running the command?
Such as:
GOOS=dragonfly go build prog
On Thu, Jan 26, 2017 at 1:10 PM, wrote:
> Is there any way to cross compile without setting GOOS environment variable?
>
> i.e
On Thu, Jan 26, 2017 at 7:10 PM wrote:
> go build os=bsd prog
> Or is it by design that you cannot just compile using a compiler switch,
you must set GOOS environment variable?
$ GOOS=bsd go build prog
$ go build -os bsd prog
The difference in length is only one letter anyway ;-)
--
-j
--
Is there any way to cross compile without setting GOOS environment variable?
i.e. something like
go build os=bsd prog
Or is it by design that you cannot just compile using a compiler switch,
you must set GOOS environment variable?
Thanks
--
You received this message because you are subscribe
On Thu, Jan 26, 2017 at 9:18 AM, T L wrote:
>
> On Wednesday, January 25, 2017 at 5:16:07 AM UTC+8, Ian Lance Taylor wrote:
>
>> By not allocating memory. It is helped by a "secret" compiler option,
>> -+. When that compiler option is used, the compiler gives an error
>> for any implicit memory
Hi.
I'm looking at the source of mobile/gl.
If I understand correctly, it has a mechanism to reduce the number of cgo
calls: the gl calls are accumulated, until a "blocking" call is reached;
when that happens, all the calls with associated arguments are passed at
once to a C function, which exe
On Thu, Jan 26, 2017 at 6:46 PM wrote:
> "If you want to use any C libraries" Don't all programs use C
libraries... When I use sysinternals procmon a windows Go exe uses tens
(almost hundreds) of OS c libraries. On linux/bsd I have not checked, but
you cannot possibly get away with doing anyt
On Thursday, January 26, 2017 at 11:15:39 PM UTC+8, andrey mirtchovski
wrote:
>
> I bet you get wildly differing results with each run. That's an
> indicator that your sample size is not statistically relevant (there's
> randomness introduced both in the select, as well as in the OS and Go
>
On Thursday, January 26, 2017 at 11:15:19 PM UTC+8, Jesper Louis Andersen
wrote:
>
> Roughly, select proceeds in two passes[0]:
>
> 1. Figure out what select-cases have events that can fire.
> 2. Pick fairly among the eligible events. Fairly means pseudo-random.
>
> What likely happens in your c
On this article"
https://inconshreveable.com/04-30-2014/cross-compiling-golang-programs-with-native-libraries/
it says "If you want to use any C libraries"
But, all operating systems today (relevant ones) are written in C and their
tools are in C (DSO's, DLL's) so as far as I can see, you are alw
On Thursday, January 26, 2017 at 7:33:42 PM UTC+2, David Sofo wrote:
>
>
> Hi,
>
> I am using this Go library https://github.com/tealeg/xlsx to to read
> excel file . Is there a way to get all sheets names? Thank you.
>
I have not used it myself but looking at the doc:
xlFile, err := xlsx.O
https://play.golang.org/p/yvr04Jhu8c
See the docs here:
https://godoc.org/github.com/tealeg/xlsx
I'm also a maintainer of this project, so let me know if you have any more
questions. ^_^
--
You received this message because you are subscribed to the Google Groups
"golang-nuts" group.
To unsub
Sorry I got it : sheet.Name is ok.
David
Le jeudi 26 janvier 2017 18:33:42 UTC+1, David Sofo a écrit :
>
>
> Hi,
>
> I am using this Go library https://github.com/tealeg/xlsx to to read
> excel file . Is there a way to get all sheets names? Thank you.
>
> Regards
> David
>
--
You received this
Hi,
I am using this Go library https://github.com/tealeg/xlsx to to read excel
file . Is there a way to get all sheets names? Thank you.
Regards
David
--
You received this message because you are subscribed to the Google Groups
"golang-nuts" group.
To unsubscribe from this group and stop rec
On Wednesday, January 25, 2017 at 5:16:07 AM UTC+8, Ian Lance Taylor wrote:
>
> On Tue, Jan 24, 2017 at 12:46 PM, josvazg >
> wrote:
> >
> > Golang runtime has been fully translated to Go for a while now. I know I
> > could just read the source code directly but...
> >
> > Is there any know
I have recently been reflecting about OrangeFS and AMD's HSA and how they
eventually reach golang users. That implies somewhere there will be a
cgogen generated wrapper for them somehow, then they will be used.
What if OrangeFS' and HSA were deeply embedded into the go language
themselves in a
It's because strings.Join accepts slice while *s* is array. You should use
*s[:]* to get slice:
fmt.Println(strings.Join(s[:], ","))
On Thursday, January 26, 2017 at 7:06:23 PM UTC+3, Joey Miller wrote:
>
> var s [2]string
> s[0] = "PASS"
> s[1] = "FAIL"
> fmt.Println(strings.Join(s, ","))
>
>
On Thu, Jan 26, 2017 at 7:45 AM, wrote:
> var s [2]string
> s[0] = "PASS"
> s[1] = "FAIL"
> fmt.Println(strings.Join(s, ","))
>
>
> The following code fails, I know this works with a slice, But I am a newbie
> to go and I am trying to figure out the reason why this fails? Does it have
> something
On Thu, Jan 26, 2017 at 2:21 AM, josvazg wrote:
> Thanks a lot Ian!
>
> My mind picture of the runtime is more detailed now. I guess for more I
> should now read the code itself.
>
> Just one side issue. You said:
> "
> Memory allocated on the stack in function F can not be used by callees
> of F
var s [2]string
s[0] = "PASS"
s[1] = "FAIL"
fmt.Println(strings.Join(s, ","))
The following code fails, I know this works with a slice, But I am a newbie
to go and I am trying to figure out the reason why this fails? Does it have
something to do with cap or len?
--
You received this message
Hi,
I feel quite stupid already, but I can not figure out how to use signed
URLs with the go storage client. I have found lots of examples on how to
generate signed URLs as well as a documentation on how to use them with
plain HTTP requests (e.g. initial Post request and subsequent PUT requests
I bet you get wildly differing results with each run. That's an
indicator that your sample size is not statistically relevant (there's
randomness introduced both in the select, as well as in the OS and Go
schedulers).
Increase the iterations until your results stabilize, then you'll see
what the r
Roughly, select proceeds in two passes[0]:
1. Figure out what select-cases have events that can fire.
2. Pick fairly among the eligible events. Fairly means pseudo-random.
What likely happens in your case is that your consumer is able to empty out
one of the channels which then means the other ch
2017. január 26., csütörtök 10:20:07 UTC+1 időpontban l...@pinkfroot.com a
következőt írta:
>
> Ok, why do you think I'm looking at too low a level, interested to
> understand!
>
> When I look at http://127.0.0.1/debug/pprof/goroutine?debug=2 most of
> the threads are in the semacquire state wa
Il giorno giovedì 26 gennaio 2017 07:37:49 UTC+1, JohnGB ha scritto:
>
> Since it is an HTTP Api, what about returning a standard HTTP status line,
>> maybe with some non standard status code) and additional details about the
>> error in the response body, JSON encoded?
>
>
> That is exactly what
On Thursday, January 26, 2017 at 8:08:28 PM UTC+8, Volker Dobler wrote:
>
> [...] Fewer select-block usages means more efficient code.
>>
>
> What if the priority-enabled-select is much complexer and
> massively less efficient than two all-equal-selects?
> It might be less lines of code but not i
>
> [...] Fewer select-block usages means more efficient code.
>
What if the priority-enabled-select is much complexer and
massively less efficient than two all-equal-selects?
It might be less lines of code but not in itself more efficient
(for whatever definition of efficiency).
V.
--
You rec
package main
import "fmt"
func main() {
c0, c1 := make(chan int, 10), make(chan int, 10)
go func() {
for {
c0 <- 0
}
}()
go func() {
for {
c1 <- 1
}
}()
var n0, n1 int
f0 := func() {
n
On Thursday, January 26, 2017 at 6:29:54 PM UTC+8, dja...@gmail.com wrote:
>
>
>
> On Wednesday, January 25, 2017 at 3:14:27 PM UTC+2, T L wrote:
>>
>> sometimes, I do want one case on a select block get selected even if
>> there are multiple unblocked cases.
>> For example, I don't want the dat
Replying to myself. Here is a must read to understand the go runtime better and
its source code:
https://github.com/golang/go/blob/master/src/runtime/HACKING.md
--
You received this message because you are subscribed to the Google Groups
"golang-nuts" group.
To unsubscribe from this group and s
Also, any recommended entry point or guide to read the runtime source code?
--
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 email
to golang-nuts+unsubscr...@googlegroups.
On Wednesday, January 25, 2017 at 3:14:27 PM UTC+2, T L wrote:
>
> sometimes, I do want one case on a select block get selected even if there
> are multiple unblocked cases.
> For example, I don't want the dataCh case is selected if the stopCh and
> the timer channel are unblocked:
>
> select {
Thanks a lot Ian!
My mind picture of the runtime is more detailed now. I guess for more I should
now read the code itself.
Just one side issue. You said:
"
Memory allocated on the stack in function F can not be used by callees
of F. After F returns, the memory is gone.
"
I don't quite get it.
Ok, why do you think I'm looking at too low a level, interested to
understand!
When I look at http://127.0.0.1/debug/pprof/goroutine?debug=2 most of the
threads are in the semacquire state waiting for a RWMutex lock to free up.
I'm trying to track down what/where it is stuck inside a current
On Thu, 26 Jan 2017 07:51:11 +
Lee Armstrong wrote:
> > Send HUP (kill -HUP pid or pressing Ctrl+\ on the terminal), when
> > the program is locked up, save the resulting stack trace, and
> > inspect. It'll list all running/blocked goroutines, and the point
> > of the wait, too.
> Thanks, I j
60 matches
Mail list logo
