After talking with Andy about the DOM API, we were kind of debating about convention of current versus cur. To be consistent with other commonly used abbreviations we decided to stick with cur.
To consistently use cur and val I'm proposing the following API changes:
sys::Locale.current => cur
sys::Locale.setCurrent => cur // make as static field? its really threadlocal
sys::TimeZone.current => cur
gfx::GfxEnv.current => cur
Enum.values => vals // this is generated by the compiler
sys::Map.values => vals
sys::Type.isValue => isVal
dom::Elem.value => val
flux::CachedOptions.value => val
json::JsonParser.value => val // not public
json::JsonWriter.writeValue => writeVal // not public
brianMon 7 Dec 2009
Promoted to ticket #855 and assigned to brian
DanielFathMon 7 Dec 2009
I'm tottally for val but cur feels weird IMHO. I think curr is the better alternative.
tacticsMon 7 Dec 2009
I'm +1 for cur. I use it all the time already.
brianSat 12 Dec 2009
Ticket resolved in 1.0.48
These API changes have been completed.
Type.isValue and Enum.values must be fixed immediately due to be intertwined with the compiler code.
The remainder of the API changes leave the old slots in place with deprecation warnings. I will officially remove these slots in 1.0.49 to give everybody a build for fixing their code.
brianFri 18 Dec 2009
Couple more things to think about as we try to standardize on when to use an abbreviation. I am thinking:
prefer msg to message
prefer rec to record
These are common terms that seem to get shortened almost as much as id for identifier.
If we do want to use msg and rec then I would propose:
I also made Err.msg non-nullable - doesn't seem like a good reason to leave potentially NPE issues dangling out there for that field. If you have your own constructors with a nullable msg, you should change them to:
class MyErr : Err
{
new make(Str msg := "", Err? cause := null) : super(msg, cause) {}
}
brianMon 21 Dec 2009
Couple more places where I think we have some inconsistency:
prefer "addr" to "address"
prefer "option" to "opt"
This effects the following APIs:
inet::IpAddress => IpAddr
inet - some of the various APIs
web::WebReq.remoteAddress
wisp::WispReq.remoteAddress
util::AbstractMain (everything uses opt)
EDIT: actually not sure about opt versus option for AbstractMain
brianMon 21 Dec 2009
Another one - we inconsistently use err and error; switched everything to use err:
brian Mon 7 Dec 2009
After talking with Andy about the DOM API, we were kind of debating about convention of
current
versuscur
. To be consistent with other commonly used abbreviations we decided to stick withcur
.To consistently use
cur
andval
I'm proposing the following API changes:brian Mon 7 Dec 2009
Promoted to ticket #855 and assigned to brian
DanielFath Mon 7 Dec 2009
I'm tottally for
val
butcur
feels weird IMHO. I thinkcurr
is the better alternative.tactics Mon 7 Dec 2009
I'm +1 for
cur
. I use it all the time already.brian Sat 12 Dec 2009
Ticket resolved in 1.0.48
These API changes have been completed.
Type.isValue
andEnum.values
must be fixed immediately due to be intertwined with the compiler code.The remainder of the API changes leave the old slots in place with deprecation warnings. I will officially remove these slots in 1.0.49 to give everybody a build for fixing their code.
brian Fri 18 Dec 2009
Couple more things to think about as we try to standardize on when to use an abbreviation. I am thinking:
These are common terms that seem to get shortened almost as much as id for identifier.
If we do want to use
msg
andrec
then I would propose:andy Fri 18 Dec 2009
I like it.
tactics Fri 18 Dec 2009
Sounds good.
brian Sun 20 Dec 2009
I made the
msg
andrec
changes.I also made
Err.msg
non-nullable - doesn't seem like a good reason to leave potentially NPE issues dangling out there for that field. If you have your own constructors with a nullable msg, you should change them to:brian Mon 21 Dec 2009
Couple more places where I think we have some inconsistency:
This effects the following APIs:
EDIT: actually not sure about opt versus option for AbstractMain
brian Mon 21 Dec 2009
Another one - we inconsistently use err and error; switched everything to use err:
tcolar Fri 9 Apr 2010
The Enum documentation (source code & this site) both still says "values" instead of "vals"
lel4866 Fri 9 Apr 2010
Not sure why you prefer option to opt. Seems opposite to the preferences for all the other abbreviations. I prefer opt.