[go-nuts] Re: Where should interfaces live?
On Monday, 21 November 2016 17:30:21 UTC+2, Vorn Mom wrote:
> Sorry if it was asked before, but where should interfaces live?
>
>- In the package that contains an implementation of it.
>- In its own package.
>- or in the package that needs it.
>
> The quick answer here is "yes to all", but most commonly it should be in
"the package that needs it".
The interface can be seen as an evolution of indirection:
*#1 - no interface*
type memorydb struct { people []*Person }
func (db *memorydb) PersonByName(name string) (*Person, error) { ... }
type Server struct {
db *memorydb
}
*#2 - private interface*, introduced to allow multiple db or to separate
code more
type memorydb struct { ... }
...
type sqllitedb struct { ... }
...
type peopledb interface { PersonByName(name string) (*Person, error) }
type Server struct {
db peopledb
}
*#3 - public interface* - introduced to allow sub-packages access/see the
same interface, this usually happens together with moving the
memorydb/sqllitedb into a separate package. It can also be used to break
dependency cycles.
...
type PeopleDB interface { PersonByName(name string) (*Person, error) }
type Server struct {
db PeopleDB
}
*#4 - common interface* - introduced to allow wide compatibility between
things; you only end-up here when you are implementing a widely used
package, e.g. "database/sql/driver.Driver", "io.Reader", "io.Writer",
"context.Context".
Try to stop at the "smallest number" that solves your problem. Each step
adds extra indirection and more artifacts, so making code harder to
understand. Of course, you should keep in mind not to make each
file/struct/package/interface too large, which also makes code harder to
read. It's a balancing act . Most web applications will end up at #3...
frameworks/libraries end up at #3 or #4. Anything else can vary from #1 -
#4.
+ Egon
The standard library is not consistent. For example:
>
>- io.Writer is defined in a package that also has an implementation,
>io.PipeWriter.
>- Encoding is all interfaces, defering implementation to sub-packages.
>- database/sql returns a DB struct on Open() instead of an interface.
>
> Because Go interfaces are decoupled from implementation, I think it should
> be defined where it's needed. So I like having my own interface for DB
> that may only define a subset of what behaviors it provides.
>
> However, doesn't defining an interface that depends on other packages
> decrease cohesion and leaks the abstraction? For example,
>
> type interface DbQuery {
> QueryRow(query string, args ...interface{}) *sql.Row
> }
>
> This interface depends on sql.Row. This decreases cohesion because the
> interface is now defined across packages?
>
> -Vorn
>
>
>
--
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.com/d/optout.
[go-nuts] Re: Why there is a so great performance penalty when the usage combines LockOSThread and channel?
Please try profiling your application, if you are on Linux perf(1) works very well for tracing user and system time. -- 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.com/d/optout.
[go-nuts] Re: Why there is a so great performance penalty when the usage combines LockOSThread and channel?
On Tuesday, 22 November 2016 17:46:18 UTC+8, Dave Cheney wrote: > > Please try profiling your application, if you are on Linux perf(1) works > very well for tracing user and system time. Thanks for mention that :) Here is the first step result of *perf stat*: Performance counter stats for './penalty_with_LockOSThread': 4893.495113 task-clock (msec) #0.583 CPUs utilized *2,112,910 context-switches #0.432 M/sec* 80 cpu-migrations#0.016 K/sec 2,377 page-faults #0.486 K/sec cycles stalled-cycles-frontend stalled-cycles-backend instructions branches branch-misses 8.389103849 seconds time elapsed Performance counter stats for './penalty_without_LockOSThread': 930.442000 task-clock (msec) #0.978 CPUs utilized 2,947 context-switches #0.003 M/sec 41 cpu-migrations#0.044 K/sec 2,389 page-faults #0.003 M/sec cycles stalled-cycles-frontend stalled-cycles-backend instructions branches branch-misses 0.951572314 seconds time elapsed It seems there is something wrong with the *context-switches *for penalty_without_LockOSThread at first sight. -- 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.com/d/optout.
[go-nuts] Re: Why there is a so great performance penalty when the usage combines LockOSThread and channel?
Perf record / report will highlight the problem. -- 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.com/d/optout.
[go-nuts] Re: Why there is a so great performance penalty when the usage combines LockOSThread and channel?
Related, you don't need a pointer to a chan, channels are already pointers to the private runtime channel type. -- 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.com/d/optout.
[go-nuts] Re: Reflect Struct Comparison
Please do not reopen to a four year old thread. Instead please start a new thread describing the problem you have, what you tried, and what happened when you tried. Thanks -- 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.com/d/optout.
[go-nuts] Re: Why there is a so great performance penalty when the usage combines LockOSThread and channel?
On Tuesday, 22 November 2016 18:06:50 UTC+8, Dave Cheney wrote: > > Perf record / report will highlight the problem. Thanks Dave.And here is the output of the perf report: Samples: 48K of event 'cpu-clock', Event count (approx.): 1202025 Overhead Command Shared Object Symbol 59.40% penalty [kernel.kallsyms] [k] _raw_spin_unlock_irqrestore 13.22% penalty [kernel.kallsyms] [k] finish_task_switch 2.13% penalty [kernel.kallsyms] [k] get_user_pages_fast 1.87% penalty penalty[.] runtime.mallocgc 1.05% penalty [kernel.kallsyms] [k] futex_wake 0.75% penalty penalty[.] runtime.findrunnable 0.72% penalty penalty[.] runtime.futex 0.71% penalty penalty[.] runtime/internal/atomic.Xchg 0.67% penalty [kernel.kallsyms] [k] unlock_page 0.61% penalty [kernel.kallsyms] [k] wake_q_add 0.56% penalty penalty[.] runtime.schedule 0.55% penalty penalty[.] runtime.runqgrab 0.52% penalty [kernel.kallsyms] [k] entry_SYSCALL_64_after_swapgs 0.50% penalty penalty[.] runtime.lock 0.46% penalty [kernel.kallsyms] [k] __wake_up_bit 0.43% penalty [kernel.kallsyms] [k] smp_call_function_single 0.39% penalty penalty[.] main.main.func1 0.37% penalty [kernel.kallsyms] [k] get_futex_key 0.37% penalty penalty[.] runtime.notesleep 0.35% penalty [kernel.kallsyms] [k] do_futex 0.32% penalty [kernel.kallsyms] [k] hash_futex 0.32% penalty [kernel.kallsyms] [k] sys_futex 0.32% penalty penalty[.] runtime.chanrecv 0.32% penalty penalty[.] runtime.startm 0.31% penalty penalty[.] runtime/internal/atomic.Load 0.30% penalty penalty[.] runtime/internal/atomic.Cas 0.30% penalty [kernel.kallsyms] [k] do_select 0.28% penalty [kernel.kallsyms] [k] core_sys_select 0.28% penalty [kernel.kallsyms] [k] futex_wait_queue_me 0.27% penalty penalty[.] runtime.chansend 0.25% penalty penalty[.] runtime.casgstatus -- 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.com/d/optout.
Re: [go-nuts] Re: Reflect Struct Comparison
Hi Dave,
I need to create a common utility which checks whether the struct is empty
or not else i can throw error.
for example:
type Employee struct{
EmpId string
EmpName string
Department []String
Projects map[string]string
}
i need to check whether the object coming from rest api i hosted has empty
value or it is filled with something. and i want to make the utility not
limited to Employee but to use the same for other structs as well.
how can i do it?
Regards,
Parveen Kumar
On Tue, Nov 22, 2016 at 3:39 PM, Dave Cheney wrote:
> Please do not reopen to a four year old thread. Instead please start a new
> thread describing the problem you have, what you tried, and what happened
> when you tried.
>
> Thanks
>
> --
> You received this message because you are subscribed to a topic in the
> Google Groups "golang-nuts" group.
> To unsubscribe from this topic, visit https://groups.google.com/d/
> topic/golang-nuts/neGz_Rxtxxw/unsubscribe.
> To unsubscribe from this group and all its topics, send an email to
> golang-nuts+unsubscr...@googlegroups.com.
> For more options, visit https://groups.google.com/d/optout.
>
--
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.com/d/optout.
[go-nuts] Re: Why there is a so great performance penalty when the usage combines LockOSThread and channel?
On Tuesday, 22 November 2016 18:08:37 UTC+8, Dave Cheney wrote: > > Related, you don't need a pointer to a chan, channels are already pointers > to the private runtime channel type. Thanks for telling me that ☺ The codes will be pretty cleaner without the pointer '*' notation. Is there something wrong with the scheduler? Or maybe a really deeply hidden bug in the codes. (Doubting about 1st one honestly) -- 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.com/d/optout.
[go-nuts] Re: Memory usage and go tool pprof
Hello! "Real" is what I see in atop/htop/ps, as i mentioned erlier. Yep, about asked memory from OS is in my mind too, but how can I find situation when/why it's occur? To be more precise, my app load N merabytes from DB, and put them into map. So, I fully understand when N megabytes transform to N gigabytes in memory in map, in the next iteration, after some time, my app doing load and transform again, so, I can imagine situation, when in exist two maps, N gigabytes each, but after that old map is gone, and my app start using new map! So, in my mind GC shoud kill old map, is I'm right? I How can I tune memory allocator? May be I can tune it to allocate memory more tight? Why my app don't return back allocated memory? вторник, 22 ноября 2016 г., 0:10:20 UTC+3 пользователь Tamás Gulácsi написал: > > What is "real"? > 65Gb has been asked from the OS, and it gave that much to the runtime. But > it used only 25Gb of it. > Maybe even released some to the OS, but maybe that did not took that. > > Is the 25Gb really needed and used for good? The OOM suggests that for > moments your program allocated much more > -- 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.com/d/optout.
Re: [go-nuts] Re: Reflect Struct Comparison
If you want to write a general function you will need to use the reflect package, specifically reflect.Value.IsValid to check each field to see if it contains the zero value or not. -- 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.com/d/optout.
[go-nuts] Re: Memory usage and go tool pprof
Please turn on gc debugging with end GODEBUG=gctrace=1 and check that your applications heap usage is behaving as you believe. -- 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.com/d/optout.
[go-nuts] Re: Why there is a so great performance penalty when the usage combines LockOSThread and channel?
Why do you want to use LockOSthread, you've proved it has a significant performance cost for your application. -- 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.com/d/optout.
Re: [go-nuts] Re: Reflect Struct Comparison
i am new to Golang. can you just share some piece of code just to check each field in struct for empty or nil or zero value. On Tue, Nov 22, 2016 at 4:19 PM, Dave Cheney wrote: > If you want to write a general function you will need to use the reflect > package, specifically reflect.Value.IsValid to check each field to see if > it contains the zero value or not. > > -- > You received this message because you are subscribed to a topic in the > Google Groups "golang-nuts" group. > To unsubscribe from this topic, visit https://groups.google.com/d/ > topic/golang-nuts/neGz_Rxtxxw/unsubscribe. > To unsubscribe from this group and all its topics, send an email to > golang-nuts+unsubscr...@googlegroups.com. > For more options, visit https://groups.google.com/d/optout. > -- 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.com/d/optout.
[go-nuts] Re: Why there is a so great performance penalty when the usage combines LockOSThread and channel?
On Tuesday, 22 November 2016 18:51:46 UTC+8, Dave Cheney wrote: > > Why do you want to use LockOSthread, you've proved it has a significant > performance cost for your application. *Cgo* is the answer. (Actually I want to use a C library(which needs LockOSthread to reach local constant thread execution model) through Cgo in the Go application.And one of the biggest benefits of Go is the Channel's event notification mechanism which is on the user space and powerful.It could be used to create a sync & nonblock application easily.) Is this a expected performance penalty when we use channel with LockOSthread?(Maybe due to a design tradeoff or anything) (: Just want to figure it out and be sure the exclusion of the probability of some bugs maybe hidden on this test codes here :) -- 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.com/d/optout.
Re: [go-nuts] Interface vs first-class function
>From my point of view, the main difference is that interfaces are strictly
more powerful, because you can't dynamically type convert a function
type into something different.
You can always make an interface from a function (e.g. http.HandlerFunc
and the like) but not the other way around.
This is a real issue in some places - for example http.Handler is an interface,
but the popular github.com/julienschmidt/httprouter package's Handle
type is a function, which means that you can't use the same router to
(for example) store metadata about routes, because all you've got is
a function.
Another consideration is that a function is only one word, so if you're
likely to be using stateless non-closure functions, and putting them
into interface values, you'll incur less allocations.
On 20 November 2016 at 05:22, Henry wrote:
> Hi,
>
> I am wondering from best practice point of view whether it is better to use
> interface or first-class function, especially considering that Go encourages
> the use of small interfaces. Here is an example:
>
> type SomethingDoer interface{
>DoSomething(data Data)
> }
>
> func PerformWork(doer SomethingDoer){
>//...
> }
>
> Or
>
> type SomethingDoer func(Data)
>
> func PerformWork(doer SomethingDoer){
>//...
> }
>
> Is there some sort of a guideline when to use one vs the other? They seem
> like redundant features to me.
>
> Thanks
>
> Henry
>
> --
> 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.com/d/optout.
--
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.com/d/optout.
Re: [go-nuts] Re: Reflect Struct Comparison
This should get you started. https://play.golang.org/p/fKGkJPMToy On Tuesday, 22 November 2016 21:58:31 UTC+11, iAdvice Edge wrote: > > i am new to Golang. > > can you just share some piece of code just to check each field in struct > for empty or nil or zero value. > > On Tue, Nov 22, 2016 at 4:19 PM, Dave Cheney > wrote: > >> If you want to write a general function you will need to use the reflect >> package, specifically reflect.Value.IsValid to check each field to see if >> it contains the zero value or not. >> >> -- >> You received this message because you are subscribed to a topic in the >> Google Groups "golang-nuts" group. >> To unsubscribe from this topic, visit >> https://groups.google.com/d/topic/golang-nuts/neGz_Rxtxxw/unsubscribe. >> To unsubscribe from this group and all its topics, send an email to >> golang-nuts...@googlegroups.com . >> For more options, visit https://groups.google.com/d/optout. >> > > -- 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.com/d/optout.
[go-nuts] Re: Why there is a so great performance penalty when the usage combines LockOSThread and channel?
The cgo call is going to cost you a lot more than this. Can you write up your real code and profile it. On Tuesday, 22 November 2016 22:11:34 UTC+11, Cia wrote: > > > > On Tuesday, 22 November 2016 18:51:46 UTC+8, Dave Cheney wrote: >> >> Why do you want to use LockOSthread, you've proved it has a significant >> performance cost for your application. > > > *Cgo* is the answer. > > (Actually I want to use a C library(which needs LockOSthread to reach > local constant thread execution model) through Cgo in the Go > application.And one of the biggest benefits of Go is the Channel's event > notification mechanism which is on the user space and powerful.It could be > used to create a sync & nonblock application easily.) > > Is this a expected performance penalty when we use channel with > LockOSthread?(Maybe due to a design tradeoff or anything) > > (: Just want to figure it out and be sure the exclusion of the probability > of some bugs maybe hidden on this test codes here :) > > -- 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.com/d/optout.
[go-nuts] Re: Why there is a so great performance penalty when the usage combines LockOSThread and channel?
I have been tested Cgo's perf and here is the result below: chan: 1000, op/s (buffer 100) 400, op/s (no buffer,blocked) cgo: 554,4729 op/s It shows that the speed of Cgo is acceptable,and even without thinking about the batch optimization method when using it *:)* On Tuesday, 22 November 2016 19:30:55 UTC+8, Dave Cheney wrote: > > The cgo call is going to cost you a lot more than this. Can you write up > your real code and profile it. > > On Tuesday, 22 November 2016 22:11:34 UTC+11, Cia wrote: >> >> >> >> On Tuesday, 22 November 2016 18:51:46 UTC+8, Dave Cheney wrote: >>> >>> Why do you want to use LockOSthread, you've proved it has a significant >>> performance cost for your application. >> >> >> *Cgo* is the answer. >> >> (Actually I want to use a C library(which needs LockOSthread to reach >> local constant thread execution model) through Cgo in the Go >> application.And one of the biggest benefits of Go is the Channel's event >> notification mechanism which is on the user space and powerful.It could be >> used to create a sync & nonblock application easily.) >> >> Is this a expected performance penalty when we use channel with >> LockOSthread?(Maybe due to a design tradeoff or anything) >> >> (: Just want to figure it out and be sure the exclusion of the >> probability of some bugs maybe hidden on this test codes here :) >> >> -- 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.com/d/optout.
[go-nuts] stupid question: how much processors/ram a twitter like web app (doing just CRUD operations) need to serve 1k request/s ?
i know this depends on a lot of variables, but i don't need something accurate if from your experience you could guess how much an app like this would need? -- 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.com/d/optout.
Re: [go-nuts] stupid question: how much processors/ram a twitter like web app (doing just CRUD operations) need to serve 1k request/s ?
To answer this kind of question, i normally write a simple app, then run it on localhost, then use ab tool to test. e.g to send 100,000 requests 100 at a time, ab -n 10 -c 100 http://localhost:3000/ Then i use 'top' to see how much cpu/ram its using. On Tue, Nov 22, 2016 at 3:27 PM, Marwan abdel moneim wrote: > i know this depends on a lot of variables, but i don't need something > accurate > if from your experience you could guess how much an app like this would > need? > > -- > 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.com/d/optout. > -- 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.com/d/optout.
[go-nuts] Client access to TLS Session-ID
Hi,
I'd like to implement a Client app in go which accesses to a Java HTTPS
server and it uses the TLS Session-ID to protect against Man-in-the-middle
attack and bind to one single TLS Session. When I use the Java Servlet the
specification
http://download.oracle.com/otn-pub/jcp/servlet-3.0-fr-eval-oth-JSpec/servlet-3_0-final-spec.pdf
expose the "javax.servlet.request.ssl_session_id" property and I have
access to TLS Session-ID something like
"032554E059DB27BF8CD87EBC53E9FF29376265F0BBFDBBFB7773D2277E5559F5"
Also when I use the $ openssl s_client -connect www.google.com:443 I get
the Session-ID too
SSL-Session:
Protocol : TLSv1.2
Cipher: ECDHE-RSA-AES128-GCM-SHA256
Session-ID:
38950231484EDAA2E55951FB86CBCED2A2A6074B6E1C5285B354626434397694
How can I access to the same information in Go?
I'm working with this code to access to Session-ID somehow but it looks
like inaccessible
from
https://github.com/golang/go/blob/master/src/crypto/tls/handshake_client.go
I'd appreciate any help or guidance about how can I fetch the Session-ID of
a TLS connection.
Thank you in advance.
func main() {
log.SetFlags(log.Lshortfile)
tr := &http.Transport{
DialTLS: dialTLSDefault,
TLSClientConfig: &tls.Config{InsecureSkipVerify: true},
}
httpClient := &http.Client{Transport: tr}
res, err := httpClient.Get("https://www.google.com/robots.txt";)
if err != nil {
log.Fatal(err)
}
defer res.Body.Close()
robots, err := ioutil.ReadAll(res.Body)
if err != nil {
log.Fatal(err)
}
fmt.Printf("%s", robots)
}
func dialTLSDefault(network, addr string) (net.Conn, error) {
fmt.Printf("Connecto to: %s %s\n ", network, addr)
cfg := &tls.Config{
InsecureSkipVerify: true,
ClientSessionCache: tls.NewLRUClientSessionCache(2),
ServerName:"www.google.com",
}
cn, err := tls.Dial(network, addr, cfg)
if err != nil {
return nil, err
}
if err := cn.Handshake(); err != nil {
return nil, err
}
if !cfg.InsecureSkipVerify {
if err := cn.VerifyHostname(cfg.ServerName); err != nil {
return nil, err
}
}
state := cn.ConnectionState()
if !state.NegotiatedProtocolIsMutual {
return nil, errors.New("http: could not negotiate protocol
mutually")
}
fmt.Printf("NegotiatedProtocol: %s\n ", state.NegotiatedProtocol)
fmt.Printf("TLSUnique: %s\n ", hex.EncodeToString(state.TLSUnique))
fmt.Printf("Session-ID: %s\n ", )
return cn, nil
}
--
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.com/d/optout.
Re: [go-nuts] stupid question: how much processors/ram a twitter like web app (doing just CRUD operations) need to serve 1k request/s ?
i will try that, thanks On Tuesday, November 22, 2016 at 3:24:19 PM UTC+2, Josh Kamau wrote: > > To answer this kind of question, i normally write a simple app, then run > it on localhost, then use ab tool to test. > e.g to send 100,000 requests 100 at a time, ab -n 10 -c 100 > http://localhost:3000/ > > Then i use 'top' to see how much cpu/ram its using. > > On Tue, Nov 22, 2016 at 3:27 PM, Marwan abdel moneim > wrote: > >> i know this depends on a lot of variables, but i don't need something >> accurate >> if from your experience you could guess how much an app like this would >> need? >> >> -- >> 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...@googlegroups.com . >> For more options, visit https://groups.google.com/d/optout. >> > > -- 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.com/d/optout.
Re: [go-nuts] Timezone parsing discrepancy
Thanks Shawn & Andrew for your replies. Ignoring the TZ abbreviation when it doesn't happen to be the local TZ seems like somewhat undesirable behavior. Unfortunately I can't work around this by switching to numeric zones, since I'm not in control of the input I'm parsing. Is there a fundamental reason time.Parse() couldn't interpret all the TZ codes standardized in RFC-822? It seems time.LoadLocation(..) knows about them. I'll make this suggestion on the issue you filed. Regards, David On Monday, November 21, 2016 at 9:18:58 PM UTC-8, andrey mirtchovski wrote: > > > The last sentence of the time.Parse function is "To avoid such problems, > > prefer time layouts that use a numeric zone offset, or use > ParseInLocation." > > It specifically refers to using "named" time zones (such as PST) instead > of > > offsets, like -0700. > > The whole paragraph there is important to read but it's terse and > needs elaboration. time.Parse works off the local timezone. If the > abbreviation specified in the data matches an abbreviation with a UTC > offset in the current local zone then that offset will be used. When > you run the documentation example in a different timezone the package > tries to lookup PST in the local zone file, it doesn't find it and > then creates a fake PST timezone with offset of zero. This means that > to have "PST Feb 3, 2013 at 7:54pm (PST)" work correctly on your > computer and to have time.Parse print "... -0800 PST" you must be in > the PST timezone or set your local timezone to it. > > I created this little example to illustrate it: > https://play.golang.org/p/giQoGlik9m; below is a sample run on my > machine: > > $ date > Mon 21 Nov 2016 21:59:23 MST > $ for i in CET PST PDT MST MDT GMT; do ./t $i; done > 2013-02-03 19:54:00 + CET > 2013-02-03 19:54:00 + PST > 2013-02-03 19:54:00 + PDT > 2013-02-03 19:54:00 -0700 MST > 2013-02-03 18:54:00 -0700 MST > 2013-02-03 19:54:00 + GMT > > I've created an issue to gauge devs' opinion: > https://github.com/golang/go/issues/18012 > > The two workarounds there should work for you (numeric timezones and > ParseInLocation). > -- 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.com/d/optout.
Re: [go-nuts] Timezone parsing discrepancy
On Tue, Nov 22, 2016 at 10:30 AM, wrote: > > Ignoring the TZ abbreviation when it doesn't happen to be the local TZ seems > like somewhat undesirable behavior. Unfortunately I can't work around this > by switching to numeric zones, since I'm not in control of the input I'm > parsing. Is there a fundamental reason time.Parse() couldn't interpret all > the TZ codes standardized in RFC-822? It seems time.LoadLocation(..) knows > about them. I'll make this suggestion on the issue you filed. I replied on the issue. I note that if you know that the RFC 822 abbreviations are correct for your use case, despite the fact that they are ambiguous internationally, then you can use a simple string substitution to replace those abbreviations with the equivalent numeric timezone. Ian -- 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.com/d/optout.
[go-nuts] Re: Why there is a so great performance penalty when the usage combines LockOSThread and channel?
Thinking about this overnight I think this benchmark https://play.golang.org/p/5d4MUKqPYd Shows the issue you are having. On my machine % go test -bench=. -benchtime=5s BenchmarkWithoutLockOSThread-4 5000 1827196 ns/op BenchmarkWithLockOSThread-4 200 31506649 ns/op My guess is the LockOSThread version causes a context switch because the thread the sending goroutine runs on must be parked, and the receiving thread locked to the receiving goroutine must be woken to drain the value. I don't see an easy workaround for this slowdown. On Tuesday, 22 November 2016 22:43:34 UTC+11, Cia wrote: > > I have been tested Cgo's perf and here is the result below: > > chan: > 1000, op/s (buffer 100) > 400, op/s (no buffer,blocked) > > cgo: > 554,4729 op/s > > It shows that the speed of Cgo is acceptable,and even without thinking > about the batch optimization method when using it *:)* > > On Tuesday, 22 November 2016 19:30:55 UTC+8, Dave Cheney wrote: >> >> The cgo call is going to cost you a lot more than this. Can you write up >> your real code and profile it. >> >> On Tuesday, 22 November 2016 22:11:34 UTC+11, Cia wrote: >>> >>> >>> >>> On Tuesday, 22 November 2016 18:51:46 UTC+8, Dave Cheney wrote: Why do you want to use LockOSthread, you've proved it has a significant performance cost for your application. >>> >>> >>> *Cgo* is the answer. >>> >>> (Actually I want to use a C library(which needs LockOSthread to reach >>> local constant thread execution model) through Cgo in the Go >>> application.And one of the biggest benefits of Go is the Channel's event >>> notification mechanism which is on the user space and powerful.It could be >>> used to create a sync & nonblock application easily.) >>> >>> Is this a expected performance penalty when we use channel with >>> LockOSthread?(Maybe due to a design tradeoff or anything) >>> >>> (: Just want to figure it out and be sure the exclusion of the >>> probability of some bugs maybe hidden on this test codes here :) >>> >>> -- 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.com/d/optout.
Re: [go-nuts] Re: Why there is a so great performance penalty when the usage combines LockOSThread and channel?
On Tue, Nov 22, 2016 at 12:37 PM, Dave Cheney wrote: > Thinking about this overnight I think this benchmark > > https://play.golang.org/p/5d4MUKqPYd > > Shows the issue you are having. On my machine > > % go test -bench=. -benchtime=5s > BenchmarkWithoutLockOSThread-4 5000 1827196 ns/op > BenchmarkWithLockOSThread-4 200 31506649 ns/op > > My guess is the LockOSThread version causes a context switch because the > thread the sending goroutine runs on must be parked, and the receiving > thread locked to the receiving goroutine must be woken to drain the value. > > I don't see an easy workaround for this slowdown. I also have come to think it may be due to thread context switching. Still, I also wrote up a benchmark, and I filed https://golang.org/issue/18023. Ian -- 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.com/d/optout.
[go-nuts] Receiving results from a database
I just started leaning Go and I've been stuck on this for a couple of days
Running on an App Engine dev server
The idea is to get an object I can work with with a format like this:
{
"Results": [
{
"username": "Test username",
"username_sani": "testusername",
"dob": "1992-09-10",
"email": "",
"gender": "M",
"account_type": "user",
"account_status": "enabled",
"activated": 1,
"country": "US",
"weekly_playtime": "6",
"game": [
],
"primary_game": "",
"activation_code": "BOxc9P2FomaaL4qKkQ2JMUIa1kgaYCJUlkfUqTs9",
"password": "",
"last_active": "2016-11-20 20:27:38",
"lol": {
"verify_string": "ymn8Hm9ojfCzLqgZaZDi"
},
"login_history": [
{
"date": "2016-11-20 20:03:46",
"result": "login_ok",
"forward": "",
"timezone": -300
}
],
"current_login_att": 0,
"reset_password_entries": [
{
"code": "sZln9bCoLYVrPXErKes5",
"valid_till": "2016-11-22 01:10",
"result": "cancelled"
}
}
],
"Includes": [],
"IsStale": false,
"IndexTimestamp": "2016-11-21T16:21:06.5157187Z",
"TotalResults": 1,
"SkippedResults": 0,
"IndexName": "users_search",
"IndexEtag": "0100--0069--0001AED9",
"ResultEtag": "A85E1A89-3C87-F9AA-B0BA-41EC39CD225A",
"Highlightings": {},
"NonAuthoritativeInformation": false,
"LastQueryTime": "2016-11-22T19:19:39.0784570Z",
"DurationMilliseconds": 0,
"ScoreExplanations": {},
"TimingsInMilliseconds": {},
"ResultSize": 0
}
My current code :
ctx := appengine.NewContext(req)
client := urlfetch.Client(ctx)
//
resp,_ := client.Get(ip+"/indexes/"+ query)
//
doc,_ := ioutil.ReadAll(resp.Body)
//
resp.Body.Close();
//
var index_results IndexResults
err := json.Unmarshal(doc,&index_results)
if err != nil {
log.Println(err)
}
log.Println(index_results)
Struct
type IndexResults struct{
Results []byte `json:"Results"`
}
Which gives me an error : json: cannot unmarshal object into Go value of
type uint8
With result {[0]}
How can I receive json results with dynamic fields that may or not be
there? Do I define a struct for every possible result?
Thank you
--
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.com/d/optout.
Re: [go-nuts] Re: Why there is a so great performance penalty when the usage combines LockOSThread and channel?
Thanks Ian. On Wed, 23 Nov 2016, 07:47 Ian Lance Taylor wrote: > On Tue, Nov 22, 2016 at 12:37 PM, Dave Cheney wrote: > > Thinking about this overnight I think this benchmark > > > > https://play.golang.org/p/5d4MUKqPYd > > > > Shows the issue you are having. On my machine > > > > % go test -bench=. -benchtime=5s > > BenchmarkWithoutLockOSThread-4 5000 1827196 ns/op > > BenchmarkWithLockOSThread-4 200 31506649 ns/op > > > > My guess is the LockOSThread version causes a context switch because the > > thread the sending goroutine runs on must be parked, and the receiving > > thread locked to the receiving goroutine must be woken to drain the > value. > > > > I don't see an easy workaround for this slowdown. > > I also have come to think it may be due to thread context switching. > > Still, I also wrote up a benchmark, and I filed > https://golang.org/issue/18023. > > Ian > -- 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.com/d/optout.
[go-nuts] Re: Receiving results from a database
First of all, please NEVER ignore the returned errors!
Then, print out "doc" to be sure.
In the concrete case, let Results be []json.RawMessage, or simply marshal
into a map[string]interface{}.
--
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.com/d/optout.
[go-nuts] Thinking OO virtual function in Go
Hi,
How to architect the OO's virtual function in Go?
Please take a look at this (not working) Go program
https://play.golang.org/p/qrBX6ScABp
Please think of the "func Output()" as a very complicated function that I
only want to define *once *at the base level, not to duplicate into each
sub classes.
How can I make it works so that the last output statement, instead of
being,
fmt.Printf("[%v] %s: [%0.2f]\n", k, v.Name(), v.Area())
will be this instead:
fmt.Printf("[%v] %s\n", k, v.Output())
Thanks
--
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.com/d/optout.
Re: [go-nuts] Thinking OO virtual function in Go
On Wed, Nov 23, 2016 at 8:16 AM, Tong Sun wrote:
> Hi,
>
> How to architect the OO's virtual function in Go?
>
> Please take a look at this (not working) Go program
> https://play.golang.org/p/qrBX6ScABp
>
> Please think of the "func Output()" as a very complicated function that I
> only want to define once at the base level, not to duplicate into each sub
> classes.
>
> How can I make it works so that the last output statement, instead of being,
>
> fmt.Printf("[%v] %s: [%0.2f]\n", k, v.Name(), v.Area())
>
>
> will be this instead:
>
> fmt.Printf("[%v] %s\n", k, v.Output())
>
You define a function:
func Output(s Shape) string {
return s.Name() + "'s area size is " + s.Area()
}
Go uses interfaces for polymorphism.
Other OOP languages can use inheritance for polymorphism too, but Go
doesn't have inheritance.
--
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.com/d/optout.
Re: [go-nuts] Thinking OO virtual function in Go
On Tue, Nov 22, 2016 at 10:16 PM, Tong Sun wrote:
> Hi,
>
> How to architect the OO's virtual function in Go?
>
> Please take a look at this (not working) Go program
> https://play.golang.org/p/qrBX6ScABp
>
> Please think of the "func Output()" as a very complicated function that I
> only want to define *once *at the base level, not to duplicate into each
> sub classes.
>
> How can I make it works so that the last output statement, instead of
> being,
>
> fmt.Printf("[%v] %s: [%0.2f]\n", k, v.Name(), v.Area())
>
>
> will be this instead:
>
> fmt.Printf("[%v] %s\n", k, v.Output())
>
>
what about this:
https://play.golang.org/p/DdZBiOgF2w
-s
--
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.com/d/optout.
Re: [go-nuts] Thinking OO virtual function in Go
On Tue, Nov 22, 2016 at 4:29 PM, Seb Binet wrote:
>
>
> On Tue, Nov 22, 2016 at 10:16 PM, Tong Sun wrote:
>
>> Hi,
>>
>> How to architect the OO's virtual function in Go?
>>
>> Please take a look at this (not working) Go program
>> https://play.golang.org/p/qrBX6ScABp
>>
>> Please think of the "func Output()" as a very complicated function that I
>> only want to define *once *at the base level, not to duplicate into each
>> sub classes.
>>
>> How can I make it works so that the last output statement, instead of
>> being,
>>
>> fmt.Printf("[%v] %s: [%0.2f]\n", k, v.Name(), v.Area())
>>
>>
>> will be this instead:
>>
>> fmt.Printf("[%v] %s\n", k, v.Output())
>>
>>
> what about this:
> https://play.golang.org/p/DdZBiOgF2w
>
Aha, that way. I've been banning my head for a while & you saved my day.
Thanks a lot!
--
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.com/d/optout.
Re: [go-nuts] Thinking OO virtual function in Go
On Tue, Nov 22, 2016 at 4:29 PM, Jesse McNelis wrote:
> On Wed, Nov 23, 2016 at 8:16 AM, Tong Sun wrote:
> > Hi,
> >
> > How to architect the OO's virtual function in Go?
> >
> > Please take a look at this (not working) Go program
> > https://play.golang.org/p/qrBX6ScABp
> >
> > Please think of the "func Output()" as a very complicated function that I
> > only want to define once at the base level, not to duplicate into each
> sub
> > classes.
> >
> > How can I make it works so that the last output statement, instead of
> being,
> >
> > fmt.Printf("[%v] %s: [%0.2f]\n", k, v.Name(), v.Area())
> >
> >
> > will be this instead:
> >
> > fmt.Printf("[%v] %s\n", k, v.Output())
> >
>
> You define a function:
>
> func Output(s Shape) string {
>return s.Name() + "'s area size is " + s.Area()
> }
>
> Go uses interfaces for polymorphism.
> Other OOP languages can use inheritance for polymorphism too, but Go
> doesn't have inheritance.
>
Thanks Jesse. That works.
However I just realized that it is only part of my problem.
I have a huge list of common variables that I defined in my "base class",
changing it from a member function to a pure function causes almost every
single variable now undefined.
I can demonstrate the problem with this code
https://play.golang.org/p/QjCtD9rGpa
So, once again, thinking in OO, I'll define all of the common variables in
base class, and common functionalities in virtual functions. How to make
that idea work in Go?
For the above specific code, how to easily make "func Output" works?
Thanks
--
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.com/d/optout.
Re: [go-nuts] Thinking OO virtual function in Go
On 23 Nov. 2016 9:03 am, "Tong Sun" wrote: > > So, once again, thinking in OO, I'll define all of the common variables in base class, and common functionalities in virtual functions. How to make that idea work in Go? > > For the above specific code, how to easily make "func Output" works? > You can use functions and embedding for code reuse and interfaces for polymorphism. In your example you've implemented Speak() method for each type and defined a Speaker interface but then in Output() you don't call Speak(). If you have fields you want available through the Speaker interface then you need to define getter methods for those fields and add them to the interface. You could also add a Speak() method to the Animal type which would then be available on any type that embeds an Animal and would have direct access to any fields on Animal. You could define a Speak() function that all the Speak() methods call for some common functionality. How you structure it depends on what real world problem you're trying to solve. -- 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.com/d/optout.
Re: [go-nuts] Thinking OO virtual function in Go
Hi,
There is no direct mapping of what you can do with virtual functions in
other OO languages, and Go. There are different compromises you have to
make; because of this, synthetic examples will probably not help much.
That being said, in the case of your last example I would make Output() a
method of Animal.
The Speak() methods of the specific animals would then call Output().
/npat
On Wednesday, November 23, 2016 at 12:03:54 AM UTC+2, Tong Sun wrote:
>
>
>
> On Tue, Nov 22, 2016 at 4:29 PM, Jesse McNelis wrote:
>
>> On Wed, Nov 23, 2016 at 8:16 AM, Tong Sun wrote:
>> > Hi,
>> >
>> > How to architect the OO's virtual function in Go?
>> >
>> > Please take a look at this (not working) Go program
>> > https://play.golang.org/p/qrBX6ScABp
>> >
>> > Please think of the "func Output()" as a very complicated function that
>> I
>> > only want to define once at the base level, not to duplicate into each
>> sub
>> > classes.
>> >
>> > How can I make it works so that the last output statement, instead of
>> being,
>> >
>> > fmt.Printf("[%v] %s: [%0.2f]\n", k, v.Name(), v.Area())
>> >
>> >
>> > will be this instead:
>> >
>> > fmt.Printf("[%v] %s\n", k, v.Output())
>> >
>>
>> You define a function:
>>
>> func Output(s Shape) string {
>>return s.Name() + "'s area size is " + s.Area()
>> }
>>
>> Go uses interfaces for polymorphism.
>> Other OOP languages can use inheritance for polymorphism too, but Go
>> doesn't have inheritance.
>>
>
> Thanks Jesse. That works.
>
> However I just realized that it is only part of my problem.
>
> I have a huge list of common variables that I defined in my "base class",
> changing it from a member function to a pure function causes almost every
> single variable now undefined.
>
> I can demonstrate the problem with this code
> https://play.golang.org/p/QjCtD9rGpa
>
> So, once again, thinking in OO, I'll define all of the common variables in
> base class, and common functionalities in virtual functions. How to make
> that idea work in Go?
>
> For the above specific code, how to easily make "func Output" works?
>
> Thanks
>
>
--
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.com/d/optout.
Re: [go-nuts] Thinking OO virtual function in Go
interfaces only work on methods.
https://play.golang.org/p/o6Ot4IdJZ1
Name, Age , ... are not methods they are fields.You need to make them part
of Speaker interface by using methods instead of fields.
Le mardi 22 novembre 2016 23:03:54 UTC+1, Tong Sun a écrit :
>
>
>
> On Tue, Nov 22, 2016 at 4:29 PM, Jesse McNelis wrote:
>
>> On Wed, Nov 23, 2016 at 8:16 AM, Tong Sun wrote:
>> > Hi,
>> >
>> > How to architect the OO's virtual function in Go?
>> >
>> > Please take a look at this (not working) Go program
>> > https://play.golang.org/p/qrBX6ScABp
>> >
>> > Please think of the "func Output()" as a very complicated function that
>> I
>> > only want to define once at the base level, not to duplicate into each
>> sub
>> > classes.
>> >
>> > How can I make it works so that the last output statement, instead of
>> being,
>> >
>> > fmt.Printf("[%v] %s: [%0.2f]\n", k, v.Name(), v.Area())
>> >
>> >
>> > will be this instead:
>> >
>> > fmt.Printf("[%v] %s\n", k, v.Output())
>> >
>>
>> You define a function:
>>
>> func Output(s Shape) string {
>>return s.Name() + "'s area size is " + s.Area()
>> }
>>
>> Go uses interfaces for polymorphism.
>> Other OOP languages can use inheritance for polymorphism too, but Go
>> doesn't have inheritance.
>>
>
> Thanks Jesse. That works.
>
> However I just realized that it is only part of my problem.
>
> I have a huge list of common variables that I defined in my "base class",
> changing it from a member function to a pure function causes almost every
> single variable now undefined.
>
> I can demonstrate the problem with this code
> https://play.golang.org/p/QjCtD9rGpa
>
> So, once again, thinking in OO, I'll define all of the common variables in
> base class, and common functionalities in virtual functions. How to make
> that idea work in Go?
>
> For the above specific code, how to easily make "func Output" works?
>
> Thanks
>
>
--
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.com/d/optout.
Re: [go-nuts] Dynamic linking against individual golang standard libraries
Alright, I will take a look. Thanks for the response. ~Parker On Monday, November 21, 2016 at 7:00:01 PM UTC-5, Ian Lance Taylor wrote: > > On Sun, Nov 20, 2016 at 1:22 PM, Parker Evans > wrote: > > > > In order to do this kind of thing, do I need to manually figure out > > dependencies and then setup appropriate linker flags? > > Yes, pretty much. > > It may help a bit to look at go/build/deps_test.go. > > Ian > -- 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.com/d/optout.
Re: [go-nuts] Thinking OO virtual function in Go
On Tue, Nov 22, 2016 at 6:23 PM, Jesse McNelis wrote: > On 23 Nov. 2016 9:03 am, "Tong Sun" wrote: > > > > So, once again, thinking in OO, I'll define all of the common variables > in base class, and common functionalities in virtual functions. How to make > that idea work in Go? > > > > For the above specific code, how to easily make "func Output" works? > > > > You can use functions and embedding for code reuse and interfaces for > polymorphism. > > In your example you've implemented Speak() method for each type and > defined a Speaker interface but then in Output() you don't call Speak(). > Yeah, that Output() is the virtual function I'm implementing. In my real code it has nothing to do with Speak() method, only to use those common variables. > If you have fields you want available through the Speaker interface then > you need to define getter methods for those fields and add them to the > interface. > Got it. > You could also add a Speak() method to the Animal type which would then be > available on any type that embeds an Animal and would have direct access to > any fields on Animal. > However, different animals speak differently. I.e., it is not the fields on Animal that matter but what outside it matter to Speak(), right? So adding a Speak() method to the Animal doesn't make much sense, right? > You could define a Speak() function that all the Speak() methods call for > some common functionality. > > How you structure it depends on what real world problem you're trying to > solve. > Probably, but just thinking in abstract terms, I thought there are some basic principles that people can follow. Anyway, I think paraiso.marc's methodology might be the only way. > -- 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.com/d/optout.
Re: [go-nuts] Thinking OO virtual function in Go
No Nick, making Output() a member method won't work.
See my OP and Jesse's answer. I.e., I have to change it from a member
function to a pure function.
On Tue, Nov 22, 2016 at 6:25 PM, Nick Patavalis
wrote:
> Hi,
>
> There is no direct mapping of what you can do with virtual functions in
> other OO languages, and Go. There are different compromises you have to
> make; because of this, synthetic examples will probably not help much.
>
> That being said, in the case of your last example I would make Output() a
> method of Animal.
> The Speak() methods of the specific animals would then call Output().
>
> /npat
>
> On Wednesday, November 23, 2016 at 12:03:54 AM UTC+2, Tong Sun wrote:
>>
>>
>>
>> On Tue, Nov 22, 2016 at 4:29 PM, Jesse McNelis wrote:
>>
>>> On Wed, Nov 23, 2016 at 8:16 AM, Tong Sun wrote:
>>> > Hi,
>>> >
>>> > How to architect the OO's virtual function in Go?
>>> >
>>> > Please take a look at this (not working) Go program
>>> > https://play.golang.org/p/qrBX6ScABp
>>> >
>>> > Please think of the "func Output()" as a very complicated function
>>> that I
>>> > only want to define once at the base level, not to duplicate into each
>>> sub
>>> > classes.
>>> >
>>> > How can I make it works so that the last output statement, instead of
>>> being,
>>> >
>>> > fmt.Printf("[%v] %s: [%0.2f]\n", k, v.Name(), v.Area())
>>> >
>>> >
>>> > will be this instead:
>>> >
>>> > fmt.Printf("[%v] %s\n", k, v.Output())
>>> >
>>>
>>> You define a function:
>>>
>>> func Output(s Shape) string {
>>>return s.Name() + "'s area size is " + s.Area()
>>> }
>>>
>>> Go uses interfaces for polymorphism.
>>> Other OOP languages can use inheritance for polymorphism too, but Go
>>> doesn't have inheritance.
>>>
>>
>> Thanks Jesse. That works.
>>
>> However I just realized that it is only part of my problem.
>>
>> I have a huge list of common variables that I defined in my "base class",
>> changing it from a member function to a pure function causes almost every
>> single variable now undefined.
>>
>> I can demonstrate the problem with this code
>> https://play.golang.org/p/QjCtD9rGpa
>>
>> So, once again, thinking in OO, I'll define all of the common variables
>> in base class, and common functionalities in virtual functions. How to make
>> that idea work in Go?
>>
>> For the above specific code, how to easily make "func Output" works?
>>
>> Thanks
>>
>> --
> You received this message because you are subscribed to a topic in the
> Google Groups "golang-nuts" group.
> To unsubscribe from this topic, visit https://groups.google.com/d/
> topic/golang-nuts/f_62HEOIBV4/unsubscribe.
> To unsubscribe from this group and all its topics, send an email to
> golang-nuts+unsubscr...@googlegroups.com.
> For more options, visit https://groups.google.com/d/optout.
>
--
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.com/d/optout.
Re: [go-nuts] Thinking OO virtual function in Go
Thanks a lot for your explicit example. Much more helpful to me than merely
saying define getters. Much appreciate it!
On Tue, Nov 22, 2016 at 6:27 PM, wrote:
> interfaces only work on methods.
>
> https://play.golang.org/p/o6Ot4IdJZ1
>
> Name, Age , ... are not methods they are fields.You need to make them part
> of Speaker interface by using methods instead of fields.
>
>
> Le mardi 22 novembre 2016 23:03:54 UTC+1, Tong Sun a écrit :
>>
>>
>>
>> On Tue, Nov 22, 2016 at 4:29 PM, Jesse McNelis wrote:
>>
>>> On Wed, Nov 23, 2016 at 8:16 AM, Tong Sun wrote:
>>> > Hi,
>>> >
>>> > How to architect the OO's virtual function in Go?
>>> >
>>> > Please take a look at this (not working) Go program
>>> > https://play.golang.org/p/qrBX6ScABp
>>> >
>>> > Please think of the "func Output()" as a very complicated function
>>> that I
>>> > only want to define once at the base level, not to duplicate into each
>>> sub
>>> > classes.
>>> >
>>> > How can I make it works so that the last output statement, instead of
>>> being,
>>> >
>>> > fmt.Printf("[%v] %s: [%0.2f]\n", k, v.Name(), v.Area())
>>> >
>>> >
>>> > will be this instead:
>>> >
>>> > fmt.Printf("[%v] %s\n", k, v.Output())
>>> >
>>>
>>> You define a function:
>>>
>>> func Output(s Shape) string {
>>>return s.Name() + "'s area size is " + s.Area()
>>> }
>>>
>>> Go uses interfaces for polymorphism.
>>> Other OOP languages can use inheritance for polymorphism too, but Go
>>> doesn't have inheritance.
>>>
>>
>> Thanks Jesse. That works.
>>
>> However I just realized that it is only part of my problem.
>>
>> I have a huge list of common variables that I defined in my "base class",
>> changing it from a member function to a pure function causes almost every
>> single variable now undefined.
>>
>> I can demonstrate the problem with this code
>> https://play.golang.org/p/QjCtD9rGpa
>>
>> So, once again, thinking in OO, I'll define all of the common variables
>> in base class, and common functionalities in virtual functions. How to make
>> that idea work in Go?
>>
>> For the above specific code, how to easily make "func Output" works?
>>
>> Thanks
>>
>> --
> You received this message because you are subscribed to a topic in the
> Google Groups "golang-nuts" group.
> To unsubscribe from this topic, visit https://groups.google.com/d/
> topic/golang-nuts/f_62HEOIBV4/unsubscribe.
> To unsubscribe from this group and all its topics, send an email to
> golang-nuts+unsubscr...@googlegroups.com.
> For more options, visit https://groups.google.com/d/optout.
>
--
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.com/d/optout.
