Found https://github.com/golang/go/issues/23796
On Tue, Jan 12, 2021 at 10:05 PM Henry wrote:
> I remember reading some proposals about it somewhere, either here at
> golang-nuts or at github. The general consensus was no, because interface
> is all about behavior. Struct is for data and interfa
I remember reading some proposals about it somewhere, either here at
golang-nuts or at github. The general consensus was no, because interface
is all about behavior. Struct is for data and interface is for behavior, or
something like that if I remember it correctly.
In your specific example, s
I'm doing some due-diligence and wondering if there's any old topics about
extending interfaces to support field sets.
ie.
type WithID interface {
ID string
}
type X struct { ID string }
type Y struct { ID string }
func PrintID(item WithID) {
fmt.Println(item.ID)
}
In the above the PrintID
>If you are using replace as part of your regular workflow, I do agree
>that
>you are probably misusing it. It's just hard to understand why, from
>what
>you said so far (barring the "I don't have an internet connection and
>want
>to use something from my module cache" case, which I find surprisi
With 'go/types' and the like, one can relatively easily automatically generate
the (de)serialization code.
See:
- go-hep.org/x/hep/brio
IIRC, specialization is off the table at the moment.
-s
Original Message
On Jan 12, 2021, 21:14, Keith Wipf < keithwi...@gmail.com> wrote:
On Tue, Jan 12, 2021 at 12:29 PM Keith Wipf wrote:
>
> Also, will Go generics support specialization? I thought that I saw
> somewhere that you would be able to switch on a type parameter at
> compile time. I'm thinking you'd need this to serialize ints and other
> primitive types because you can'
If you've ever worked with Rust, you've probably sceen the serde framework.
My question is, once Go has generics, would it be possible to write
something like this in Go? This seems like something that could really
benefit the language.
The only drawback I can see so far as that you'd have to manua
Jim, I used replace this way just yesterday. It works fine, but is not
easily discoverable as a process and it took me a minute or two to figure
it out. Also, one must be sure to remove the replace directive before
releasing the code, which is an easy detail to forget.
The same trick (and required
Ops. sorry, it is. Wonder why it wasn't working before (tried at least twice).
On Tue, Jan 12, 2021 at 3:12 PM Tong Sun wrote:
>
> That's a typo. Changing it to "E" you'd think the problem is solved?
>
> On Mon, Jan 11, 2021 at 10:14 PM Kurtis Rader wrote:
> >
> > You should start by asking your
That's a typo. Changing it to "E" you'd think the problem is solved?
On Mon, Jan 11, 2021 at 10:14 PM Kurtis Rader wrote:
>
> You should start by asking yourself why your program outputs "1", "2" and
> "S". Hint: after seeing "S" you don't break out of the inner loop until
> another "S" is seen
I have a third use case for replace:
3. Simultaneously developing a library and client for that library. For
example, if I have a library to access a network service or piece of
hardware and want to extend its API, I find that working on the client
along with the library helps me iterate to a good
Your `var db=initDB()` runs before init(), and it returns an error, so the
program aborts.
Here's a change which initializes db explicitly in the init function, and
it works as you expect:
https://play.golang.org/p/xsUU2hfnx-w
The init function is called after all variable initializations happen,
Hi all,
I'm looking into how contexts are used, and wondering: does anybody know of
an idiomatic use of context on a struct for the sake of a one-shot?
(a one-shot being some struct that gets used once, like an RPC or http
request)
For example, http.Request has a context on its struct. But, I th
Hello, please does anyone know why init() won't run?
Thank You very much!
https://play.golang.org/p/0PDUJlF8X1w
--
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
Er, lifetime of the struct *
On Tue, Jan 12, 2021 at 8:11 AM Jean de Klerk wrote:
> Hi all,
>
> I'm looking into how contexts are used, and wondering: does anybody know
> of an idiomatic use of context on a struct for the sake of a one-shot?
>
> (a one-shot being some struct that gets used once,
If you want to move the type info into genbox you can do
https://play.golang.org/p/_yhuvOhtLvl
On Tuesday, 12 January 2021 at 11:40:27 pm UTC+8 Alex wrote:
> > newItem = new(KdElem)
> > if itemfunc(arg, &newItem.item, newItem.size) != 0
>
> "newItem.item" is a "interface{}" with a nil value.
>
Looking at all goroutines with runnable state seem to be the solution but I
am not very sure about this.
*goroutine 15825281 [runnable]*
On Tuesday, January 12, 2021 at 9:16:47 PM UTC+5:30 varun...@gmail.com
wrote:
> Hi, I would like to know how to identify mapping between an OS thread and
>
Hi, I would like to know how to identify mapping between an OS thread and
go-routine from the stack trace i.e. which go-routine is scheduled for
execution on a OS thread.
*Background:* My application spawns very large number of go-routines (upto
2M) for heavy data processing. In some cases, I a
This is highly unlikely to be a compiler issue. Note that there are more
arguments in the stack trace than in the signature of your method so you can't
assume correspondence by placement. It looks like the first two arguments in
the stack trace are the type and value pointers for arg (which is i
> newItem = new(KdElem)
> if itemfunc(arg, &newItem.item, newItem.size) != 0
"newItem.item" is a "interface{}" with a nil value.
Doing "&newItem.item" then turns it into a "*interface{}"
That is why you cannot do "var realval *int64 = val.(*int64)"
To make it work as what I think you intend to d
Ian--
You missed the part about the arguments being "shifted left" above, which
kinda looks like a compiler issue. Maybe something I wrote contributed to
the problem, but I have no idea what.
If this is a known problem, perhaps I can hand-install aa newer version, or
something...?
murf
On Tue
Note the error says that val was nil, so your type assertion will panic. You
can read more about how to make type assertions that report success or failure
in the Go Tour: https://tour.golang.org/methods/15
Have you tried printing the type of what you receive before attempting the type
assertio
On Tue, Jan 12, 2021 at 2:49 PM Kevin Chadwick wrote:
> On 1/12/21 12:37 PM, Axel Wagner wrote:
> > Again, would be interesting how other languages can make this simpler,
> because I
> > genuinely can't imagine.
>
> The issue is that I believe go avoids files in order to speed up
> compilation.
>
On Tue, Jan 12, 2021 at 1:49 PM Kevin Chadwick wrote:
> Maybe I need to look into -mod vendor as I didn't know that existed.
> Otherwise
> disconnect your internet and then import a folder of local .go files that
> you
> use in multiple projects?
>
Ah, I see. I agree that this could be simpler.
On 1/12/21 12:37 PM, Axel Wagner wrote:
> Again, would be interesting how other languages can make this simpler,
> because I
> genuinely can't imagine.
The issue is that I believe go avoids files in order to speed up compilation.
Which I am thankful for. It's just I don't believe the replace dire
On 1/12/21 12:33 PM, Axel Wagner wrote:
> You also talk about onboarding - but I don't understand how the rest of your
> message relates to that. The replace-directive is an advanced feature that
> newcomers to the language don't need to be concerned with, as far as I can
> tell.
> It's really onl
On Tue, Jan 12, 2021 at 1:20 PM Kevin Chadwick wrote:
> I guess it's the
>
> Call code in an external package
> "https://golang.org/doc/tutorial/getting-started";
>
> That is a little hairy.
>
What is hairy about this? To me, that seems like pretty simple steps:
1. Find the library you need and
Hi,
I'm having trouble understanding and separating your core problems and what
you are suggesting to solve them.
The solution to describe, sounds to me a lot like what `-mod vendor`
achieves. I don't use that very often, so I don't really understand its
shortcomings - so it might be helpful if y
On 1/12/21 9:39 AM, Artur Vianna wrote:
> I don't think this tweet is very true, it took me many months of learning Go
> to
> start learning about go.mod. I think that's not the first thing one learns
> about
> a language, you can just 'vim main.go' and begin experimenting.
>
Fair enough. I had
On 1/12/21 9:43 AM, Axel Wagner wrote:
> git init
> go mod init
>
> I guess you *could* safe the `git init` part, but is that really worth the
> added
> complexity?
Sorry I meant replace keyword, not require in the original mail. Though
requiring a domain name is equally unintuitive.
I guess
On Tue, Jan 12, 2021 at 10:26 AM Kevin Chadwick wrote:
> Inspired by:
>
> "https://twitter.com/rakyll/status/1348723364894961666";
>
> "Yet another couple of hours spent on go.mod and go.sum. If I was giving
> #golang
> a first try today, I'd just give up and look for another option. What are
> w
I don't think this tweet is very true, it took me many months of learning
Go to start learning about go.mod. I think that's not the first thing one
learns about a language, you can just 'vim main.go' and begin experimenting.
How do you propose this tool will work?
On Tue, 12 Jan 2021, 06:25 Kevin
Inspired by:
"https://twitter.com/rakyll/status/1348723364894961666";
"Yet another couple of hours spent on go.mod and go.sum. If I was giving #golang
a first try today, I'd just give up and look for another option. What are we
doing as a community to address this?"
Couldn't go have a folder bas
33 matches
Mail list logo
