The original code was in the style
   for(x <- lst) {..}
A lot of newer code looks like
   for (x <- lst) {...}
Some code also occasionally does
   for ( x <- lst ) {...}

This is not a big deal but because I am obsessive compulsive I must make
all whitespace in a given file consistent. This results in big patches with
mostly trivial changes. Can we standardize on one of the above and
opportunistically convert towards it?

Let's call the above A, B, and C respectively. I vote for A, but am pretty
cool with B too.

-Jay

Prefer A, but ok with B.

Thanks,
Neha

On Tue, Sep 18, 2012 at 2:35 PM, Jay Kreps <[EMAIL PROTECTED]> wrote:
> The original code was in the style
>    for(x <- lst) {..}
> A lot of newer code looks like
>    for (x <- lst) {...}
> Some code also occasionally does
>    for ( x <- lst ) {...}
>
> This is not a big deal but because I am obsessive compulsive I must make
> all whitespace in a given file consistent. This results in big patches with
> mostly trivial changes. Can we standardize on one of the above and
> opportunistically convert towards it?
>
> Let's call the above A, B, and C respectively. I vote for A, but am pretty
> cool with B too.
>
> -Jay

I prefer A, but go with C only when iterating over certain collections
(maps/tuples etc) to take advantage of Scala's destructuring capabilities;
in these case the spaced brackets make it easier to read.

- Prashanth

On Tue, Sep 18, 2012 at 5:42 PM, Neha Narkhede <[EMAIL PROTECTED]>wrote:

> Prefer A, but ok with B.
>
> Thanks,
> Neha
>
> On Tue, Sep 18, 2012 at 2:35 PM, Jay Kreps <[EMAIL PROTECTED]> wrote:
> > The original code was in the style
> >    for(x <- lst) {..}
> > A lot of newer code looks like
> >    for (x <- lst) {...}
> > Some code also occasionally does
> >    for ( x <- lst ) {...}
> >
> > This is not a big deal but because I am obsessive compulsive I must make
> > all whitespace in a given file consistent. This results in big patches
> with
> > mostly trivial changes. Can we standardize on one of the above and
> > opportunistically convert towards it?
> >
> > Let's call the above A, B, and C respectively. I vote for A, but am
> pretty
> > cool with B too.
> >
> > -Jay
>

I'm also obsessive compulsive and prefer B.

Joel

On Tue, Sep 18, 2012 at 3:01 PM, Prashanth Menon <[EMAIL PROTECTED]
> wrote:

> I prefer A, but go with C only when iterating over certain collections
> (maps/tuples etc) to take advantage of Scala's destructuring capabilities;
> in these case the spaced brackets make it easier to read.
>
> - Prashanth
>
> On Tue, Sep 18, 2012 at 5:42 PM, Neha Narkhede <[EMAIL PROTECTED]
> >wrote:
>
> > Prefer A, but ok with B.
> >
> > Thanks,
> > Neha
> >
> > On Tue, Sep 18, 2012 at 2:35 PM, Jay Kreps <[EMAIL PROTECTED]> wrote:
> > > The original code was in the style
> > >    for(x <- lst) {..}
> > > A lot of newer code looks like
> > >    for (x <- lst) {...}
> > > Some code also occasionally does
> > >    for ( x <- lst ) {...}
> > >
> > > This is not a big deal but because I am obsessive compulsive I must
> make
> > > all whitespace in a given file consistent. This results in big patches
> > with
> > > mostly trivial changes. Can we standardize on one of the above and
> > > opportunistically convert towards it?
> > >
> > > Let's call the above A, B, and C respectively. I vote for A, but am
> > pretty
> > > cool with B too.
> > >
> > > -Jay
> >
>

I like B because it helps me to easily reach the actual body when if or
for conditions span multiple lines.

Also there are other ambiguities in the code mainly with the use of
whitespaces and conciseness, it would be cool if we could have a consensus
on them. Please pardon me for   trying to make everybody's life worse. :\

Some examples that I observed:

1. Number of characters in a line. This seems to be a subjective decision
right now.

2. Some files have semicolons at the end of lines.

3. Newline or no newline after the try block.

4. Whitespace between a keyword/identifier and { or no whitespace, this is
different sometimes within the same method.

5. If (Š) one-linear() OR
   If (Š)
     One-liner()

6. Naming standard for class instance members vs method local variables.
Well, this is not an ambiguity but being a new guy, I would definitely
appreciate having different standards here for easier reading of code.

7. /*Š*/ or //...

8. Some instances of function parameter indentation.

9. Rare placement of { in the next line after if or for.

10. def method(): contents.get or def method: contents.get

11. This happens in unit tests.
Val config = new KafkaConfig(props) {
(Number of whitespaces?) override val logFileSize = 1000
}

Thanks,
Swapnil

On 9/18/12 3:01 PM, "Prashanth Menon" <[EMAIL PROTECTED]> wrote:

>I prefer A, but go with C only when iterating over certain collections
>(maps/tuples etc) to take advantage of Scala's destructuring capabilities;
>in these case the spaced brackets make it easier to read.
>
>- Prashanth
>
>On Tue, Sep 18, 2012 at 5:42 PM, Neha Narkhede
><[EMAIL PROTECTED]>wrote:
>
>> Prefer A, but ok with B.
>>
>> Thanks,
>> Neha
>>
>> On Tue, Sep 18, 2012 at 2:35 PM, Jay Kreps <[EMAIL PROTECTED]> wrote:
>> > The original code was in the style
>> >    for(x <- lst) {..}
>> > A lot of newer code looks like
>> >    for (x <- lst) {...}
>> > Some code also occasionally does
>> >    for ( x <- lst ) {...}
>> >
>> > This is not a big deal but because I am obsessive compulsive I must
>>make
>> > all whitespace in a given file consistent. This results in big patches
>> with
>> > mostly trivial changes. Can we standardize on one of the above and
>> > opportunistically convert towards it?
>> >
>> > Let's call the above A, B, and C respectively. I vote for A, but am
>> pretty
>> > cool with B too.
>> >
>> > -Jay
>>

I prefer B.

Thanks,

Jun

On Tue, Sep 18, 2012 at 2:35 PM, Jay Kreps <[EMAIL PROTECTED]> wrote:

> The original code was in the style
>    for(x <- lst) {..}
> A lot of newer code looks like
>    for (x <- lst) {...}
> Some code also occasionally does
>    for ( x <- lst ) {...}
>
> This is not a big deal but because I am obsessive compulsive I must make
> all whitespace in a given file consistent. This results in big patches with
> mostly trivial changes. Can we standardize on one of the above and
> opportunistically convert towards it?
>
> Let's call the above A, B, and C respectively. I vote for A, but am pretty
> cool with B too.
>
> -Jay
>

Not sure if we reached any consensus here. It seems like a tie: 3 votes for
A and 3 votes for B. Any other people want to chime in to break the tie :)?

Thanks,

Jun

On Tue, Sep 18, 2012 at 2:35 PM, Jay Kreps <[EMAIL PROTECTED]> wrote:

> The original code was in the style
>    for(x <- lst) {..}
> A lot of newer code looks like
>    for (x <- lst) {...}
> Some code also occasionally does
>    for ( x <- lst ) {...}
>
> This is not a big deal but because I am obsessive compulsive I must make
> all whitespace in a given file consistent. This results in big patches with
> mostly trivial changes. Can we standardize on one of the above and
> opportunistically convert towards it?
>
> Let's call the above A, B, and C respectively. I vote for A, but am pretty
> cool with B too.
>
> -Jay
>

I also liked Prashanth's suggestion of standardizing a few other aspects of
style.

One option would be to just adopt an existing style guide off the shelf.
There is one from twitter and a few other places. I think that might be an
easy way to go. We could use whatever whitespace style they have.

If we want to do a custom style guide then I guess we should just file a
JIRA to improve the existing style guidelines and try to get consensus on
various aspects.

-Jay

On Wed, Sep 26, 2012 at 10:57 AM, Jun Rao <[EMAIL PROTECTED]> wrote:

> Not sure if we reached any consensus here. It seems like a tie: 3 votes for
> A and 3 votes for B. Any other people want to chime in to break the tie :)?
>
> Thanks,
>
> Jun
>
> On Tue, Sep 18, 2012 at 2:35 PM, Jay Kreps <[EMAIL PROTECTED]> wrote:
>
> > The original code was in the style
> >    for(x <- lst) {..}
> > A lot of newer code looks like
> >    for (x <- lst) {...}
> > Some code also occasionally does
> >    for ( x <- lst ) {...}
> >
> > This is not a big deal but because I am obsessive compulsive I must make
> > all whitespace in a given file consistent. This results in big patches
> with
> > mostly trivial changes. Can we standardize on one of the above and
> > opportunistically convert towards it?
> >
> > Let's call the above A, B, and C respectively. I vote for A, but am
> pretty
> > cool with B too.
> >
> > -Jay
> >
>

I am a fan of consistency

http://twitter.github.com/effectivescala/ works for me though I a don't
have a strong opinion to using this or another but we need to have a
consensus!

On Wed, Sep 26, 2012 at 5:42 PM, Jay Kreps <[EMAIL PROTECTED]> wrote:

> I also liked Prashanth's suggestion of standardizing a few other aspects of
> style.
>
> One option would be to just adopt an existing style guide off the shelf.
> There is one from twitter and a few other places. I think that might be an
> easy way to go. We could use whatever whitespace style they have.
>
> If we want to do a custom style guide then I guess we should just file a
> JIRA to improve the existing style guidelines and try to get consensus on
> various aspects.
>
> -Jay
>
> On Wed, Sep 26, 2012 at 10:57 AM, Jun Rao <[EMAIL PROTECTED]> wrote:
>
> > Not sure if we reached any consensus here. It seems like a tie: 3 votes
> for
> > A and 3 votes for B. Any other people want to chime in to break the tie
> :)?
> >
> > Thanks,
> >
> > Jun
> >
> > On Tue, Sep 18, 2012 at 2:35 PM, Jay Kreps <[EMAIL PROTECTED]> wrote:
> >
> > > The original code was in the style
> > >    for(x <- lst) {..}
> > > A lot of newer code looks like
> > >    for (x <- lst) {...}
> > > Some code also occasionally does
> > >    for ( x <- lst ) {...}
> > >
> > > This is not a big deal but because I am obsessive compulsive I must
> make
> > > all whitespace in a given file consistent. This results in big patches
> > with
> > > mostly trivial changes. Can we standardize on one of the above and
> > > opportunistically convert towards it?
> > >
> > > Let's call the above A, B, and C respectively. I vote for A, but am
> > pretty
> > > cool with B too.
> > >
> > > -Jay
> > >
> >
>

--

/*
Joe Stein
http://www.linkedin.com/in/charmalloc
Twitter: @allthingshadoop <http://www.twitter.com/allthingshadoop>
*/