Re: [go-nuts] Re: why is the "1" in the code demoed in Go spec deduced as a "float64" value instead of an "int" value?
On Thursday, 8 March 2018 00:02:58 UTC+1, Ian Lance Taylor wrote: > > > If we ignore the type context, then > > var v int64 = 1 << s > > fails on 32-bit systems if s > 31 (because 1 is assigned type int, > which is 32 bits, and the shifting a 32-bit value by more than 31 bits > gives you zero). > > If we don't ignore the type context, then in > > var v float32 = 1 << s > > fails because we can't shift a float32. > > Thanks! This explanation is very clear. The spec already contains an example of this rule: var k = uint64(1
Re: [go-nuts] Re: why is the "1" in the code demoed in Go spec deduced as a "float64" value instead of an "int" value?
On Wednesday, March 7, 2018 at 6:02:58 PM UTC-5, Ian Lance Taylor wrote: > > On Wed, Mar 7, 2018 at 1:36 PM, > > wrote: > > > > On Wednesday, March 7, 2018 at 4:26:19 PM UTC-5, di...@veryhaha.com > wrote: > >> > >> get it almost. > >> > >> But I feel > >> > >> var v = float32(1< >> is a little different to > >> var v float32 = 1< >> > >> > >> For the former one, we think "1" can be assumed as an "int". > >> But anyway, I get the main point of the design. > >> It is just a wording accuracy problem. > >> > > > > On the other hand, any bad if "1" is deduced as an "int" value? > > If we ignore the type context, then > > var v int64 = 1 << s > > fails on 32-bit systems if s > 31 (because 1 is assigned type int, > which is 32 bits, and the shifting a 32-bit value by more than 31 bits > gives you zero). > > If we don't ignore the type context, then in > > var v float32 = 1 << s > > fails because we can't shift a float32. > > We could ignore the type context for float types but not for integer > types, but then the rule is even more complicated. > > Ian > > > Good design. Thanks for the explanation. > > > >> On Wednesday, March 7, 2018 at 3:59:48 PM UTC-5, Ian Lance Taylor > wrote: > >>> > >>> On Wed, Mar 7, 2018 at 12:54 PM, Volker Dobler > >>> wrote: > >>> > Looks suspicious. Without crosschecking the Spec: Might be > >>> > a bug. File an issue? > >>> > >>> It's not a bug. See Andrey's reply. > >>> > >>> Ian > >>> > >>> > >>> > >>> > On Wednesday, 7 March 2018 21:39:32 UTC+1, di...@veryhaha.com > wrote: > >>> >> > >>> >> var s uint = 33 > >>> >> var u2 = float64(1>>s) // illegal: 1 has type float64, cannot > shift > >>> > > >>> > -- > >>> > 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...@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: why is the "1" in the code demoed in Go spec deduced as a "float64" value instead of an "int" value?
On Wed, Mar 7, 2018 at 1:36 PM, wrote: > > On Wednesday, March 7, 2018 at 4:26:19 PM UTC-5, di...@veryhaha.com wrote: >> >> get it almost. >> >> But I feel >> >> var v = float32(1<> is a little different to >> var v float32 = 1<> >> >> For the former one, we think "1" can be assumed as an "int". >> But anyway, I get the main point of the design. >> It is just a wording accuracy problem. >> > > On the other hand, any bad if "1" is deduced as an "int" value? If we ignore the type context, then var v int64 = 1 << s fails on 32-bit systems if s > 31 (because 1 is assigned type int, which is 32 bits, and the shifting a 32-bit value by more than 31 bits gives you zero). If we don't ignore the type context, then in var v float32 = 1 << s fails because we can't shift a float32. We could ignore the type context for float types but not for integer types, but then the rule is even more complicated. Ian >> On Wednesday, March 7, 2018 at 3:59:48 PM UTC-5, Ian Lance Taylor wrote: >>> >>> On Wed, Mar 7, 2018 at 12:54 PM, Volker Dobler >>> wrote: >>> > Looks suspicious. Without crosschecking the Spec: Might be >>> > a bug. File an issue? >>> >>> It's not a bug. See Andrey's reply. >>> >>> Ian >>> >>> >>> >>> > On Wednesday, 7 March 2018 21:39:32 UTC+1, di...@veryhaha.com wrote: >>> >> >>> >> var s uint = 33 >>> >> var u2 = float64(1>>s) // illegal: 1 has type float64, cannot shift >>> > >>> > -- >>> > 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. -- 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 is the "1" in the code demoed in Go spec deduced as a "float64" value instead of an "int" value?
of course you can force it: https://play.golang.org/p/QIWDeMDJlAq On Wed, Mar 7, 2018 at 1:36 PM, wrote: > > > On Wednesday, March 7, 2018 at 4:26:19 PM UTC-5, di...@veryhaha.com wrote: >> >> get it almost. >> >> But I feel >> >> var v = float32(1<> is a little different to >> var v float32 = 1<> >> >> For the former one, we think "1" can be assumed as an "int". >> But anyway, I get the main point of the design. >> It is just a wording accuracy problem. >> >> > On the other hand, any bad if "1" is deduced as an "int" value? > > >> >> >> On Wednesday, March 7, 2018 at 3:59:48 PM UTC-5, Ian Lance Taylor wrote: >>> >>> On Wed, Mar 7, 2018 at 12:54 PM, Volker Dobler >>> wrote: >>> > Looks suspicious. Without crosschecking the Spec: Might be >>> > a bug. File an issue? >>> >>> It's not a bug. See Andrey's reply. >>> >>> Ian >>> >>> >>> >>> > On Wednesday, 7 March 2018 21:39:32 UTC+1, di...@veryhaha.com wrote: >>> >> >>> >> var s uint = 33 >>> >> var u2 = float64(1>>s) // illegal: 1 has type float64, cannot shift >>> > >>> > -- >>> > 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. > -- Michael T. Jones michael.jo...@gmail.com -- 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 is the "1" in the code demoed in Go spec deduced as a "float64" value instead of an "int" value?
On Wednesday, March 7, 2018 at 4:26:19 PM UTC-5, di...@veryhaha.com wrote: > > get it almost. > > But I feel > > var v = float32(1< is a little different to > var v float32 = 1< > > For the former one, we think "1" can be assumed as an "int". > But anyway, I get the main point of the design. > It is just a wording accuracy problem. > > On the other hand, any bad if "1" is deduced as an "int" value? > > > On Wednesday, March 7, 2018 at 3:59:48 PM UTC-5, Ian Lance Taylor wrote: >> >> On Wed, Mar 7, 2018 at 12:54 PM, Volker Dobler >> wrote: >> > Looks suspicious. Without crosschecking the Spec: Might be >> > a bug. File an issue? >> >> It's not a bug. See Andrey's reply. >> >> Ian >> >> >> >> > On Wednesday, 7 March 2018 21:39:32 UTC+1, di...@veryhaha.com wrote: >> >> >> >> var s uint = 33 >> >> var u2 = float64(1>>s) // illegal: 1 has type float64, cannot shift >> > >> > -- >> > 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] Re: why is the "1" in the code demoed in Go spec deduced as a "float64" value instead of an "int" value?
get it almost. But I feel var v = float32(1< > On Wed, Mar 7, 2018 at 12:54 PM, Volker Dobler > > wrote: > > Looks suspicious. Without crosschecking the Spec: Might be > > a bug. File an issue? > > It's not a bug. See Andrey's reply. > > Ian > > > > > On Wednesday, 7 March 2018 21:39:32 UTC+1, di...@veryhaha.com wrote: > >> > >> var s uint = 33 > >> var u2 = float64(1>>s) // illegal: 1 has type float64, cannot shift > > > > -- > > 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] Re: why is the "1" in the code demoed in Go spec deduced as a "float64" value instead of an "int" value?
On Wed, Mar 7, 2018 at 12:54 PM, Volker Dobler wrote: > Looks suspicious. Without crosschecking the Spec: Might be > a bug. File an issue? It's not a bug. See Andrey's reply. Ian > On Wednesday, 7 March 2018 21:39:32 UTC+1, di...@veryhaha.com wrote: >> >> var s uint = 33 >> var u2 = float64(1>>s) // illegal: 1 has type float64, cannot shift > > -- > 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.
