[dart-misc] DEP meeting notes

['Bob Nystrom' via Dart Misc]

Sorry I've been behind on meeting notes. Here are the notes from last
week's meeting:

https://github.com/dart-lang/dart_enhancement_proposals/blob/master/Meetings/2015-11-11%20DEP%20Committee%20Meeting.md

The week before that, we met informally but didn't discuss any concrete
proposals. It was more process administrivia, so I didn't write up notes.
There was no DEP meeting this week.

Cheers!

– bob

[Jan Mostert]

Binary: 01001b
Octal: 74346o
Hex: 35535FA4x

This would also useful, 3E2 == 300

Also, what would the downside of doing 1,300,588 instead of 1_300_588 or
will this screw with type inference?

Post by 'Bob Nystrom' via Dart Misc
Sorry I've been behind on meeting notes. Here are the notes from last
https://github.com/dart-lang/dart_enhancement_proposals/blob/master/Meetings/2015-11-11%20DEP%20Committee%20Meeting.md
The week before that, we met informally but didn't discuss any concrete
proposals. It was more process administrivia, so I didn't write up notes.
There was no DEP meeting this week.
Cheers!
– bob
--
For other discussions, see https://groups.google.com/a/dartlang.org/
For HOWTO questions, visit http://stackoverflow.com/tags/dart
To file a bug report or feature request, go to http://www.dartbug.com/new
---
You received this message because you are subscribed to the Google Groups
"Dart Misc" group.
To unsubscribe from this group and stop receiving emails from it, send an

['Harry Terkelsen' via Dart Misc]

I think that would make the grammar ambiguous for lists of expressions, eg
in a function call. What would f(1,300,588) mean?

Post by Jan Mostert
Binary: 01001b
Octal: 74346o
Hex: 35535FA4x
This would also useful, 3E2 == 300
Also, what would the downside of doing 1,300,588 instead of 1_300_588 or
will this screw with type inference?

Post by 'Bob Nystrom' via Dart Misc
Sorry I've been behind on meeting notes. Here are the notes from last
https://github.com/dart-lang/dart_enhancement_proposals/blob/master/Meetings/2015-11-11%20DEP%20Committee%20Meeting.md
The week before that, we met informally but didn't discuss any concrete
proposals. It was more process administrivia, so I didn't write up notes.
There was no DEP meeting this week.
Cheers!
– bob
--
For other discussions, see https://groups.google.com/a/dartlang.org/
For HOWTO questions, visit http://stackoverflow.com/tags/dart
To file a bug report or feature request, go to http://www.dartbug.com/new
---
You received this message because you are subscribed to the Google Groups
"Dart Misc" group.
To unsubscribe from this group and stop receiving emails from it, send an

--
For other discussions, see https://groups.google.com/a/dartlang.org/
For HOWTO questions, visit http://stackoverflow.com/tags/dart
To file a bug report or feature request, go to http://www.dartbug.com/new
---
You received this message because you are subscribed to the Google Groups
"Dart Misc" group.
To unsubscribe from this group and stop receiving emails from it, send an

[Jan Mostert]

True, I agree with that and underscores are certainly better than spaces.



On Fri, 20 Nov 2015 at 07:31 'Harry Terkelsen' via Dart Misc <

Post by 'Harry Terkelsen' via Dart Misc
I think that would make the grammar ambiguous for lists of expressions, eg
in a function call. What would f(1,300,588) mean?

Post by Jan Mostert
Binary: 01001b
Octal: 74346o
Hex: 35535FA4x
This would also useful, 3E2 == 300
Also, what would the downside of doing 1,300,588 instead of 1_300_588 or
will this screw with type inference?

Post by 'Bob Nystrom' via Dart Misc
Sorry I've been behind on meeting notes. Here are the notes from last
https://github.com/dart-lang/dart_enhancement_proposals/blob/master/Meetings/2015-11-11%20DEP%20Committee%20Meeting.md
The week before that, we met informally but didn't discuss any concrete
proposals. It was more process administrivia, so I didn't write up notes.
There was no DEP meeting this week.
Cheers!
– bob
--
For other discussions, see https://groups.google.com/a/dartlang.org/
For HOWTO questions, visit http://stackoverflow.com/tags/dart
To file a bug report or feature request, go to
http://www.dartbug.com/new
---
You received this message because you are subscribed to the Google
Groups "Dart Misc" group.
To unsubscribe from this group and stop receiving emails from it, send

--
For other discussions, see https://groups.google.com/a/dartlang.org/
For HOWTO questions, visit http://stackoverflow.com/tags/dart
To file a bug report or feature request, go to http://www.dartbug.com/new
---
You received this message because you are subscribed to the Google Groups
"Dart Misc" group.
To unsubscribe from this group and stop receiving emails from it, send an

--
For other discussions, see https://groups.google.com/a/dartlang.org/
For HOWTO questions, visit http://stackoverflow.com/tags/dart
To file a bug report or feature request, go to http://www.dartbug.com/new
---
You received this message because you are subscribed to the Google Groups
"Dart Misc" group.
To unsubscribe from this group and stop receiving emails from it, send an

['Lasse R.H. Nielsen' via Dart Misc]

Post by Jan Mostert
Binary: 01001b
Octal: 74346o
Hex: 35535FA4x

That's not really the C-language style. 0b01001, 0o74346 and 0x35535fa4 is
more consistent with the existing 0x prefiox.

Post by Jan Mostert
This would also useful, 3E2 == 300

3e2 already means 300.0

An alternative is to introduce "p" as the exponent for integers:
3p2 => 300
0x3p2 => 0x300
0b1p5 => 0b100000
(no negative powers allowed).

Also, what would the downside of doing 1,300,588 instead of 1_300_588 or

Post by Jan Mostert
will this screw with type inference?

As Harry pointed out, it will screw with parsing.

/L
--
Lasse R.H. Nielsen - ***@google.com
'Faith without judgement merely degrades the spirit divine'
Google Denmark ApS - Frederiksborggade 20B, 1 sal - 1360 KÞbenhavn K
- Denmark - CVR nr. 28 86 69 84

[Jan Mostert]

"That's not really the C-language style. 0b01001, 0o74346 and 0x35535fa4 is
more consistent with the existing 0x prefiox."

Ah, misread the meeting notes, for a moment thought the suggestion was 0b
or 1b or 010101b
Also prefer the 0b10101 style.

"3e2 already means 300.0"

Awesome!

"An alternative is to introduce "p" as the exponent for integers"

Power and exponent are not the same things.
I'd rather let 3p2 => 9 (or alternatively 3^2) and 3e2 => 300






On Fri, 20 Nov 2015 at 07:55 'Lasse R.H. Nielsen' via Dart Misc <

Post by 'Lasse R.H. Nielsen' via Dart Misc

Post by Jan Mostert
Binary: 01001b
Octal: 74346o
Hex: 35535FA4x

That's not really the C-language style. 0b01001, 0o74346 and 0x35535fa4 is
more consistent with the existing 0x prefiox.

Post by Jan Mostert
This would also useful, 3E2 == 300

3e2 already means 300.0
3p2 => 300
0x3p2 => 0x300
0b1p5 => 0b100000
(no negative powers allowed).
Also, what would the downside of doing 1,300,588 instead of 1_300_588 or

Post by Jan Mostert
will this screw with type inference?

As Harry pointed out, it will screw with parsing.
/L
--
'Faith without judgement merely degrades the spirit divine'
Google Denmark ApS - Frederiksborggade 20B, 1 sal - 1360 KÞbenhavn K
- Denmark - CVR nr. 28 86 69 84
--
For other discussions, see https://groups.google.com/a/dartlang.org/
For HOWTO questions, visit http://stackoverflow.com/tags/dart
To file a bug report or feature request, go to http://www.dartbug.com/new
---
You received this message because you are subscribed to the Google Groups
"Dart Misc" group.
To unsubscribe from this group and stop receiving emails from it, send an

['Lasse R.H. Nielsen' via Dart Misc]

Post by Jan Mostert
"3e2 already means 300.0"
Awesome!
"An alternative is to introduce "p" as the exponent for integers"
Power and exponent are not the same things.
I'd rather let 3p2 => 9 (or alternatively 3^2) and 3e2 => 300

We can't use 3e2 as an integer since it's already defined syntax for a
double. We need another syntax.
Using "p" is an option, "e" is not. It means the same, it's just for
integers instead of doubles.

Also "3^2" already means something else, so we can't use that for power.
If we wanted to add a power operator, it would have to be a new operator,
most likely "**".

/L
--
Lasse R.H. Nielsen - ***@google.com
'Faith without judgement merely degrades the spirit divine'
Google Denmark ApS - Frederiksborggade 20B, 1 sal - 1360 KÞbenhavn K
- Denmark - CVR nr. 28 86 69 84

[Jan Mostert]

Just think it would be strange to have a different operator for a different
type, it's not like you use a + sign for integers and another symbol to add
up floats ... wouldn't this solve it?

float f1 = 300e3; // this would give 300000 and since the type is float,
casts it to 300000.0
float f2 = 300.0e3; // this would give 300000.0 and no casting needed.
or the Java equivalent of 300.0 would be 300f, but that would look weird
doing 300fe3

int i1 = 300e3; this would give 300000;
int i2 = 300.0e3; this would give 300000.0 which would fail with some sort
of loss of precision exception when trying to cast it;

So the type before the e would transferred to the result.

If you're doing 0x1A1e3, would that mean
1A1 (hexadecimal) x (10 x 10 x 10 dec)
or
1A1 (hexadecimal) x (F x F x F hex)

Or would you have to specify the type of both sides, eg 0x1A1e0x3 to mean
the former and 0x1A1e3 to mean the latter?





On Fri, 20 Nov 2015 at 08:52 'Lasse R.H. Nielsen' via Dart Misc <

Post by 'Lasse R.H. Nielsen' via Dart Misc

Post by Jan Mostert
"3e2 already means 300.0"
Awesome!
"An alternative is to introduce "p" as the exponent for integers"
Power and exponent are not the same things.
I'd rather let 3p2 => 9 (or alternatively 3^2) and 3e2 => 300

We can't use 3e2 as an integer since it's already defined syntax for a
double. We need another syntax.
Using "p" is an option, "e" is not. It means the same, it's just for
integers instead of doubles.
Also "3^2" already means something else, so we can't use that for power.
If we wanted to add a power operator, it would have to be a new operator,
most likely "**".
/L
--
'Faith without judgement merely degrades the spirit divine'
Google Denmark ApS - Frederiksborggade 20B, 1 sal - 1360 KÞbenhavn K
- Denmark - CVR nr. 28 86 69 84
--
For other discussions, see https://groups.google.com/a/dartlang.org/
For HOWTO questions, visit http://stackoverflow.com/tags/dart
To file a bug report or feature request, go to http://www.dartbug.com/new
---
You received this message because you are subscribed to the Google Groups
"Dart Misc" group.
To unsubscribe from this group and stop receiving emails from it, send an

['Lasse R.H. Nielsen' via Dart Misc]

Post by Jan Mostert
Just think it would be strange to have a different operator for a
different type, it's not like you use a + sign for integers and another
symbol to add up floats ... wouldn't this solve it?
float f1 = 300e3; // this would give 300000 and since the type is float,
casts it to 300000.0

Not in Dart. There is no automatic conversion anywhere, objects have the
type they have. Also, Dart does not have "float" as a type :)

In this case 300e3 is defined to be the double value 300000.0, changing
that is a breaking change at the language level. It's also different from
pretty much any other language. I don't know any languages where 300e3 is
not a double (some of them have automatic conversion - but the literal is a
double literal).

Post by Jan Mostert
float f2 = 300.0e3; // this would give 300000.0 and no casting needed.
or the Java equivalent of 300.0 would be 300f, but that would look weird
doing 300fe3
int i1 = 300e3; this would give 300000;

Yeah, it would be nice, but we can't do that without changing the language,
and I don't think that's likely to happen for something like this.

Post by Jan Mostert
int i2 = 300.0e3; this would give 300000.0 which would fail with some sort
of loss of precision exception when trying to cast it;

Again, not in Dart. It will fail with a type error in checked mode because
a double is not an integer.

Post by Jan Mostert
So the type before the e would transferred to the result.
If you're doing 0x1A1e3, would that mean
1A1 (hexadecimal) x (10 x 10 x 10 dec)
or
1A1 (hexadecimal) x (F x F x F hex)

1A1 (hexadecimal) x (10 x 10 x 10 hex) more likely.

I would let:
integerLiteral 'p' decimalLiteral
be (the value of integerLiteral) * (the base of integerLiteral) ** (the
value of decimalLiteral) so
0xABCDp4 is 0xABCD * (16 ** 4)
and
0b1101p4 is 0b1101 * (2 ** 4) // aka 0b1001 << 4

Maybe we only need this for decimal literals, because base 2 and 16 can
both be handled by shifts instead.

Post by Jan Mostert
Or would you have to specify the type of both sides, eg 0x1A1e0x3 to mean
the former and 0x1A1e3 to mean the latter?

Ick, no.
And this also brings up another problem with 'e' that I forgot to mention:
It doesn't work with hex numbers: 0x1A1e3 is already a valid hex-literal,
where the 'e' means something. That's another reason for using, for
example, 'p' instead for literals.

/L
--
Lasse R.H. Nielsen - ***@google.com
'Faith without judgement merely degrades the spirit divine'
Google Denmark ApS - Frederiksborggade 20B, 1 sal - 1360 KÞbenhavn K
- Denmark - CVR nr. 28 86 69 84

[Jan Mostert]

Exponent is probably not that useful for integers in any case, interesting
discussion though :-)



On Fri, 20 Nov 2015 at 12:06 'Lasse R.H. Nielsen' via Dart Misc <

Post by 'Lasse R.H. Nielsen' via Dart Misc

Post by Jan Mostert
Just think it would be strange to have a different operator for a
different type, it's not like you use a + sign for integers and another
symbol to add up floats ... wouldn't this solve it?
float f1 = 300e3; // this would give 300000 and since the type is float,
casts it to 300000.0

Not in Dart. There is no automatic conversion anywhere, objects have the
type they have. Also, Dart does not have "float" as a type :)
In this case 300e3 is defined to be the double value 300000.0, changing
that is a breaking change at the language level. It's also different from
pretty much any other language. I don't know any languages where 300e3 is
not a double (some of them have automatic conversion - but the literal is a
double literal).

Post by Jan Mostert
float f2 = 300.0e3; // this would give 300000.0 and no casting needed.
or the Java equivalent of 300.0 would be 300f, but that would look weird
doing 300fe3
int i1 = 300e3; this would give 300000;

Yeah, it would be nice, but we can't do that without changing the
language, and I don't think that's likely to happen for something like this.

Post by Jan Mostert
int i2 = 300.0e3; this would give 300000.0 which would fail with some
sort of loss of precision exception when trying to cast it;

Again, not in Dart. It will fail with a type error in checked mode because
a double is not an integer.

Post by Jan Mostert
So the type before the e would transferred to the result.
If you're doing 0x1A1e3, would that mean
1A1 (hexadecimal) x (10 x 10 x 10 dec)
or
1A1 (hexadecimal) x (F x F x F hex)

1A1 (hexadecimal) x (10 x 10 x 10 hex) more likely.
integerLiteral 'p' decimalLiteral
be (the value of integerLiteral) * (the base of integerLiteral) ** (the
value of decimalLiteral) so
0xABCDp4 is 0xABCD * (16 ** 4)
and
0b1101p4 is 0b1101 * (2 ** 4) // aka 0b1001 << 4
Maybe we only need this for decimal literals, because base 2 and 16 can
both be handled by shifts instead.

Post by Jan Mostert
Or would you have to specify the type of both sides, eg 0x1A1e0x3 to mean
the former and 0x1A1e3 to mean the latter?

Ick, no.
It doesn't work with hex numbers: 0x1A1e3 is already a valid hex-literal,
where the 'e' means something. That's another reason for using, for
example, 'p' instead for literals.
/L
--
'Faith without judgement merely degrades the spirit divine'
Google Denmark ApS - Frederiksborggade 20B, 1 sal - 1360 KÞbenhavn K
- Denmark - CVR nr. 28 86 69 84
--
For other discussions, see https://groups.google.com/a/dartlang.org/
For HOWTO questions, visit http://stackoverflow.com/tags/dart
To file a bug report or feature request, go to http://www.dartbug.com/new
---
You received this message because you are subscribed to the Google Groups
"Dart Misc" group.
To unsubscribe from this group and stop receiving emails from it, send an

[Anders Holmgren]

From talking to users, his impression is that users are reaching for

const because they really want

immutable objects. So maybe the real problem we should be tackling is

explicitly supporting objects that

are deeply immutable after construction.

Want muchos ^^

Sorry I've been behind on meeting notes. Here are the notes from last
https://github.com/dart-lang/dart_enhancement_proposals/blob/master/Meetings/2015-11-11%20DEP%20Committee%20Meeting.md
The week before that, we met informally but didn't discuss any concrete
proposals. It was more process administrivia, so I didn't write up notes.
There was no DEP meeting this week.
Cheers!
– bob

[Anders Holmgren]

Bob,

Do you have a rough roadmap for the ghost of DEPs past. Something like

1/ DEPs we plan to implement soon
2/ DEPs we plan to implement in the medium term
3/ DEPs we plan to implement in the long term
4/ DEPs we accepted but have no idea whether we will implement them.

I suspect many are like me and read these posts with interest and get
excited when their pet DEPs (pet features) are discussed and make their way
through this process, only to get disheartened as the DEPs disappear from
sight when they come out the other end.

There are many DEPs that I have gotten excited about such as NNBD and
generic methods, that I'd love to know where they fit into the rough scheme
of things

A

Post by Anders Holmgren

From talking to users, his impression is that users are reaching for

const because they really want

immutable objects. So maybe the real problem we should be tackling is

explicitly supporting objects that

are deeply immutable after construction.

Want muchos ^^

Sorry I've been behind on meeting notes. Here are the notes from last
https://github.com/dart-lang/dart_enhancement_proposals/blob/master/Meetings/2015-11-11%20DEP%20Committee%20Meeting.md
The week before that, we met informally but didn't discuss any concrete
proposals. It was more process administrivia, so I didn't write up notes.
There was no DEP meeting this week.
Cheers!
– bob

['Bob Nystrom' via Dart Misc]

+floitsch

Post by Anders Holmgren
Do you have a rough roadmap for the ghost of DEPs past. Something like
1/ DEPs we plan to implement soon
2/ DEPs we plan to implement in the medium term
3/ DEPs we plan to implement in the long term
4/ DEPs we accepted but have no idea whether we will implement them.

I don't, at least not anything very formal. One of the things we're still
figuring out is how to track proposals effectively without adding a lot of
overhead to the process.

Florian is probably the best person to ask about priority and stuff like
that.

Post by Anders Holmgren
I suspect many are like me and read these posts with interest and get
excited when their pet DEPs (pet features) are discussed and make their way
through this process, only to get disheartened as the DEPs disappear from
sight when they come out the other end.

I feel your pain.

There are many DEPs that I have gotten excited about such as NNBD and

Post by Anders Holmgren
generic methods, that I'd love to know where they fit into the rough scheme
of things

Generic methods are being very actively worked on by the DDC and analyzer
folks. If you read Vijay's status updates on the dev-compiler list, you'll
see they generally talk about progress on generic methods.

My understanding—and don't hold me to this—is that when/if that is done,
then they'll move on to NNBD.

Those two features are the highest priority for a lot of people on the team
(including me).

Cheers!

– bob

[Anders Holmgren]

Post by 'Bob Nystrom' via Dart Misc
+floitsch

Post by Anders Holmgren
Do you have a rough roadmap for the ghost of DEPs past. Something like
1/ DEPs we plan to implement soon
2/ DEPs we plan to implement in the medium term
3/ DEPs we plan to implement in the long term
4/ DEPs we accepted but have no idea whether we will implement them.

I don't, at least not anything very formal. One of the things we're still
figuring out is how to track proposals effectively without adding a lot of
overhead to the process.

Sure I don't want it to slow down getting the features out

Post by 'Bob Nystrom' via Dart Misc
Florian is probably the best person to ask about priority and stuff like
that.

Post by Anders Holmgren
I suspect many are like me and read these posts with interest and get
excited when their pet DEPs (pet features) are discussed and make their way
through this process, only to get disheartened as the DEPs disappear from
sight when they come out the other end.

I feel your pain.
There are many DEPs that I have gotten excited about such as NNBD and

Post by Anders Holmgren
generic methods, that I'd love to know where they fit into the rough scheme
of things

Generic methods are being very actively worked on by the DDC and analyzer
folks. If you read Vijay's status updates on the dev-compiler list, you'll
see they generally talk about progress on generic methods.

I heard that they were working on an annotation based mechanism (which I
understand is an interim solution). Are they also working on direct
language support already or is that still in the future?

Post by 'Bob Nystrom' via Dart Misc
My understanding—and don't hold me to this—is that when/if that is done,
then they'll move on to NNBD.

OK I'll try to contain my excitement and not get my hopes up but I have to
say that just made my day. Dart is gonna be so awesome once those two
features land IMO

Post by 'Bob Nystrom' via Dart Misc
Those two features are the highest priority for a lot of people on the
team (including me).
Cheers!
– bob

['Florian Loitsch' via Dart Misc]

Post by Anders Holmgren

Post by 'Bob Nystrom' via Dart Misc
+floitsch

Post by Anders Holmgren
Do you have a rough roadmap for the ghost of DEPs past. Something like
1/ DEPs we plan to implement soon
2/ DEPs we plan to implement in the medium term
3/ DEPs we plan to implement in the long term
4/ DEPs we accepted but have no idea whether we will implement them.

I don't, at least not anything very formal. One of the things we're still
figuring out is how to track proposals effectively without adding a lot of
overhead to the process.

Sure I don't want it to slow down getting the features out

Post by 'Bob Nystrom' via Dart Misc
Florian is probably the best person to ask about priority and stuff like
that.

It doesn't look that way, but my highest priority is

configuration-specific imports.
They have nice benefits, and are non-breaking.

My current plan is to bundle breaking chances (like generic methods and
NNBD) and release them in a bigger release.
The scope of that release isn't completely defined yet, but those two
features have a good chance of getting in. They are mostly gated by tooling
support and our experience when migrating old code to it.
If it's a painless experience (which requires good tools), then they have a
higher chance, than if it is a real pain.
A few people here are still busy with other tasks, so we haven't had a lot
of time to experiment with tools, user experience and prototypes.
Over time, more people will work on those.

Post by Anders Holmgren

Post by 'Bob Nystrom' via Dart Misc

Post by Anders Holmgren
I suspect many are like me and read these posts with interest and get
excited when their pet DEPs (pet features) are discussed and make their way
through this process, only to get disheartened as the DEPs disappear from
sight when they come out the other end.

I feel your pain.
There are many DEPs that I have gotten excited about such as NNBD and

Post by Anders Holmgren
generic methods, that I'd love to know where they fit into the rough scheme
of things

Generic methods are being very actively worked on by the DDC and analyzer
folks. If you read Vijay's status updates on the dev-compiler list, you'll
see they generally talk about progress on generic methods.

I heard that they were working on an annotation based mechanism (which I
understand is an interim solution). Are they also working on direct
language support already or is that still in the future?

Post by 'Bob Nystrom' via Dart Misc
My understanding—and don't hold me to this—is that when/if that is done,
then they'll move on to NNBD.

OK I'll try to contain my excitement and not get my hopes up but I have to
say that just made my day. Dart is gonna be so awesome once those two
features land IMO

Post by 'Bob Nystrom' via Dart Misc
Those two features are the highest priority for a lot of people on the
team (including me).
Cheers!
– bob

[Anders Holmgren]

Thanks a lot for the update Florian

Post by 'Florian Loitsch' via Dart Misc

Post by Anders Holmgren

Post by 'Bob Nystrom' via Dart Misc
+floitsch

Post by Anders Holmgren
Do you have a rough roadmap for the ghost of DEPs past. Something like
1/ DEPs we plan to implement soon
2/ DEPs we plan to implement in the medium term
3/ DEPs we plan to implement in the long term
4/ DEPs we accepted but have no idea whether we will implement them.

I don't, at least not anything very formal. One of the things we're
still figuring out is how to track proposals effectively without adding a
lot of overhead to the process.

Sure I don't want it to slow down getting the features out

Post by 'Bob Nystrom' via Dart Misc
Florian is probably the best person to ask about priority and stuff like
that.

It doesn't look that way, but my highest priority is

configuration-specific imports.
They have nice benefits, and are non-breaking.
My current plan is to bundle breaking chances (like generic methods and
NNBD) and release them in a bigger release.
The scope of that release isn't completely defined yet, but those two
features have a good chance of getting in. They are mostly gated by tooling
support and our experience when migrating old code to it.
If it's a painless experience (which requires good tools), then they have
a higher chance, than if it is a real pain.
A few people here are still busy with other tasks, so we haven't had a lot
of time to experiment with tools, user experience and prototypes.
Over time, more people will work on those.

Post by Anders Holmgren

Post by 'Bob Nystrom' via Dart Misc

Post by Anders Holmgren
I suspect many are like me and read these posts with interest and get
excited when their pet DEPs (pet features) are discussed and make their way
through this process, only to get disheartened as the DEPs disappear from
sight when they come out the other end.

I feel your pain.
There are many DEPs that I have gotten excited about such as NNBD and

Post by Anders Holmgren
generic methods, that I'd love to know where they fit into the rough scheme
of things

Generic methods are being very actively worked on by the DDC and
analyzer folks. If you read Vijay's status updates on the dev-compiler
list, you'll see they generally talk about progress on generic methods.

I heard that they were working on an annotation based mechanism (which I
understand is an interim solution). Are they also working on direct
language support already or is that still in the future?

Post by 'Bob Nystrom' via Dart Misc
My understanding—and don't hold me to this—is that when/if that is done,
then they'll move on to NNBD.

OK I'll try to contain my excitement and not get my hopes up but I have
to say that just made my day. Dart is gonna be so awesome once those two
features land IMO

Post by 'Bob Nystrom' via Dart Misc
Those two features are the highest priority for a lot of people on the
team (including me).
Cheers!
– bob

[Don Olmstead]

With the NNBD are union types in play? Its always sounded like those would
go well together.

Post by Anders Holmgren
Thanks a lot for the update Florian

On Fri, Nov 20, 2015 at 3:08 PM Anders Holmgren <

Post by Anders Holmgren

Post by 'Bob Nystrom' via Dart Misc
+floitsch

Post by Anders Holmgren
Do you have a rough roadmap for the ghost of DEPs past. Something like
1/ DEPs we plan to implement soon
2/ DEPs we plan to implement in the medium term
3/ DEPs we plan to implement in the long term
4/ DEPs we accepted but have no idea whether we will implement them.

I don't, at least not anything very formal. One of the things we're
still figuring out is how to track proposals effectively without adding a
lot of overhead to the process.

Sure I don't want it to slow down getting the features out

Post by 'Bob Nystrom' via Dart Misc
Florian is probably the best person to ask about priority and stuff
like that.

It doesn't look that way, but my highest priority is

configuration-specific imports.
They have nice benefits, and are non-breaking.
My current plan is to bundle breaking chances (like generic methods and
NNBD) and release them in a bigger release.
The scope of that release isn't completely defined yet, but those two
features have a good chance of getting in. They are mostly gated by tooling
support and our experience when migrating old code to it.
If it's a painless experience (which requires good tools), then they have
a higher chance, than if it is a real pain.
A few people here are still busy with other tasks, so we haven't had a
lot of time to experiment with tools, user experience and prototypes.
Over time, more people will work on those.

Post by Anders Holmgren

Post by 'Bob Nystrom' via Dart Misc

Post by Anders Holmgren
I suspect many are like me and read these posts with interest and get
excited when their pet DEPs (pet features) are discussed and make their way
through this process, only to get disheartened as the DEPs disappear from
sight when they come out the other end.

I feel your pain.
There are many DEPs that I have gotten excited about such as NNBD and

Post by Anders Holmgren
generic methods, that I'd love to know where they fit into the rough scheme
of things

Generic methods are being very actively worked on by the DDC and
analyzer folks. If you read Vijay's status updates on the dev-compiler
list, you'll see they generally talk about progress on generic methods.

I heard that they were working on an annotation based mechanism (which I
understand is an interim solution). Are they also working on direct
language support already or is that still in the future?

Post by 'Bob Nystrom' via Dart Misc
My understanding—and don't hold me to this—is that when/if that is
done, then they'll move on to NNBD.

OK I'll try to contain my excitement and not get my hopes up but I have
to say that just made my day. Dart is gonna be so awesome once those two
features land IMO

Post by 'Bob Nystrom' via Dart Misc
Those two features are the highest priority for a lot of people on the
team (including me).
Cheers!
– bob
--

For other discussions, see https://groups.google.com/a/dartlang.org/
For HOWTO questions, visit http://stackoverflow.com/tags/dart
To file a bug report or feature request, go to http://www.dartbug.com/new
---
You received this message because you are subscribed to the Google Groups
"Dart Misc" group.
To unsubscribe from this group and stop receiving emails from it, send an

[Filipe Morgado]

Unfortunately, I don't think so, even though it be would useful for JS/TS-interop.

They not only go well together, they're fundamentally the same feature.

[Anders Holmgren]

Yeah union types would be very nice too. I never thought of it that way but
I guess they can be viewed as essentially the same feature

Post by Filipe Morgado
Unfortunately, I don't think so, even though it be would useful for JS/TS-interop.
They not only go well together, they're fundamentally the same feature.
--
For other discussions, see https://groups.google.com/a/dartlang.org/
For HOWTO questions, visit http://stackoverflow.com/tags/dart
To file a bug report or feature request, go to http://www.dartbug.com/new
---
You received this message because you are subscribed to a topic in the
Google Groups "Dart Misc" group.
To unsubscribe from this topic, visit
https://groups.google.com/a/dartlang.org/d/topic/misc/aadjSmvmi8Y/unsubscribe
.
To unsubscribe from this group and all its topics, send an email to

[kc]

Post by Anders Holmgren
Bob,
Do you have a rough roadmap for the ghost of DEPs past. Something like
1/ DEPs we plan to implement soon
2/ DEPs we plan to implement in the medium term
3/ DEPs we plan to implement in the long term
4/ DEPs we accepted but have no idea whether we will implement them.

Something like this:
https://github.com/dotnet/roslyn/issues/2136

Post by Anders Holmgren
I suspect many are like me and read these posts with interest and get
excited when their pet DEPs (pet features) are discussed and make their way
through this process, only to get disheartened as the DEPs disappear from
sight when they come out the other end.
There are many DEPs that I have gotten excited about such as NNBD and
generic methods, that I'd love to know where they fit into the rough scheme
of things

Communication for Dart 1.x was poor.

K.

[Ladislav Thon]

Post by 'Bob Nystrom' via Dart Misc
https://github.com/dart-lang/dart_enhancement_proposals/blob/master/Meetings/2015-11-11%20DEP%20Committee%20Meeting.md

I had a working implementation of binary literals in the VM years ago, and
it took me just a few hours. And that was probably the biggest piece of C++
code I ever wrote, in a codebase I was never really familiar with.

What I'm saying is that 1. these small pieces really are quite small :-),
2. the community might help with them.

LT

P.S.: sorry for being terse, need to run.

[kc]

Would, for example, we rather have the team implementing for binary
literals—including parser support in the analyzer, dart2js, VM, syntax
highlighters, tests, specification, etc.—or should we spend that time doing
things like making analyzer faster or adding VM features the Flutter folks
want?

The latter is Darts opportunity to become relevant outside Google.

K.

Post by 'Bob Nystrom' via Dart Misc
Sorry I've been behind on meeting notes. Here are the notes from last
https://github.com/dart-lang/dart_enhancement_proposals/blob/master/Meetings/2015-11-11%20DEP%20Committee%20Meeting.md
The week before that, we met informally but didn't discuss any concrete
proposals. It was more process administrivia, so I didn't write up notes.
There was no DEP meeting this week.
Cheers!
– bob