Pedro Furlanetto | 1 Apr 2011 01:00
Picon
Gravatar

Re: scaladoc: separating members

Approved!

On Thu, Mar 31, 2011 at 7:00 PM, Daniel Sobral <dcsobral <at> gmail.com> wrote:
> +1
>
> On Thu, Mar 31, 2011 at 17:50, Paul Phillips <paulp <at> improving.org> wrote:
>>
>> Attached is a screenshot showing the impact of a scaladoc change I just
>> checked in.  Like I said in the commit message, I can't even imagine not
>> wanting to do this, but since I figured out at some point that everyone
>> doesn't see the world quite the way I do, I submit it for your approval.
>
>
>
> --
> Daniel C. Sobral
>
> I travel to the future all the time.
>

Donna Malayeri | 1 Apr 2011 03:35
Picon
Gravatar

Re: scaladoc: separating members

I like it!

Archontophoenix Quar | 1 Apr 2011 05:07
Picon

Re: a spruced-up scaladoc

Ooh, pretty! I like the fonts in number 5.

However, I'd like it better if the icons for classes and traits were more visually distinct. Those little white-on-colored background t's and c's look awfully similar from a distance. Perhaps different colors for the backgrounds? And (as long as I'm on Santa's knee) maybe yet *another* color to distinguish concrete from abstract classes? When I'm exploring a library, often the first thing I do is try to figure out which types are concrete and which are abstract.

A

On Thu, Mar 31, 2011 at 11:09 AM, Donna Malayeri <lindydonna <at> gmail.com> wrote:
With the help of Rüdiger Keller, I've created some modifications to the colors and layout of Scaladoc. I have a few versions available:

CSS changes, but no color changes:
http://lamp.epfl.ch/~malayeri/scaladoc-versions/scaladoc2/

Minor additional color changes:
http://lamp.epfl.ch/~malayeri/scaladoc-versions/scaladoc3/

Lots more color changes:
http://lamp.epfl.ch/~malayeri/scaladoc-versions/scaladoc4/

Minor font change and sharper object/class/trait icons for the left pane
http://lamp.epfl.ch/~malayeri/scaladoc-versions/scaladoc5/

I personally prefer the last of these, but we all welcome your feedback!  What is good, what is bad, and why?

thanks,
Donna

Randall R Schulz | 1 Apr 2011 07:33
Favicon

Re: Re: a spruced-up scaladoc

On Thursday March 31 2011, Rüdiger Keller wrote:
> Hi Donna,
>
> nice job putting up all these versions of Scaladoc!
>
> Just for kicks, I attached an image to my mail that shows a variation
> of the title bar. The type name is underlined, because I though it
> could be used as a link to the companion object, if there is one.
>
> Everyone, please feel free to comment on this one, too.
>
> Regards,
> Rüdiger

Please tell me you're not forcing a serifed font. Let the user choose 
what their default font type is, please.

Randall Schulz

Donna Malayeri | 1 Apr 2011 08:48
Picon
Gravatar

Re: a spruced-up scaladoc

I am planning on doing different colors for traits and classes, so that's already on the to-do list. :)

Concrete vs. abstract classes is a bit more tricky, since that would require more changes to the HTML output. In Scala, you often see traits rather than abstract classes, so hopefully this isn't as important as compared to javadoc.

Donna

On Apr 1, 2011, at 5:07 AM, Archontophoenix Quar wrote:

Ooh, pretty! I like the fonts in number 5.

However, I'd like it better if the icons for classes and traits were more visually distinct. Those little white-on-colored background t's and c's look awfully similar from a distance. Perhaps different colors for the backgrounds? And (as long as I'm on Santa's knee) maybe yet *another* color to distinguish concrete from abstract classes? When I'm exploring a library, often the first thing I do is try to figure out which types are concrete and which are abstract.

A

On Thu, Mar 31, 2011 at 11:09 AM, Donna Malayeri <lindydonna <at> gmail.com> wrote:
With the help of Rüdiger Keller, I've created some modifications to the colors and layout of Scaladoc. I have a few versions available:

CSS changes, but no color changes:
http://lamp.epfl.ch/~malayeri/scaladoc-versions/scaladoc2/

Minor additional color changes:
http://lamp.epfl.ch/~malayeri/scaladoc-versions/scaladoc3/

Lots more color changes:
http://lamp.epfl.ch/~malayeri/scaladoc-versions/scaladoc4/

Minor font change and sharper object/class/trait icons for the left pane
http://lamp.epfl.ch/~malayeri/scaladoc-versions/scaladoc5/

I personally prefer the last of these, but we all welcome your feedback!  What is good, what is bad, and why?

thanks,
Donna


Donna Malayeri | 1 Apr 2011 08:58
Picon
Gravatar

Re: Re: a spruced-up scaladoc

That was just a screenshot based on one of my proposed CSS changes. My plan is to use the font style that people
like most, which seems to be the sans-serif font right now. It's not so easy to allow users to dynamically
change the font style, so someone who wants a different default would have to change their CSS file locally.

My thought was that perhaps people would prefer the serif font, as that is how it's done in Javadoc.

Donna

On Apr 1, 2011, at 7:33 AM, Randall R Schulz wrote:
> Please tell me you're not forcing a serifed font. Let the user choose 
> what their default font type is, please.

Nate Nystrom | 1 Apr 2011 09:20
Picon

Re: Re: a spruced-up scaladoc

Donna,

It would be nice to ensure that there's enough semantic information in the CSS classes and ids that someone
could easily write  a Greasemonkey script to change the default rendering.  For instance, abstract and
concrete types should have different CSS classes.

Nate

On Apr 1, 2011, at 08:58 , Donna Malayeri wrote:

> That was just a screenshot based on one of my proposed CSS changes. My plan is to use the font style that
people like most, which seems to be the sans-serif font right now. It's not so easy to allow users to
dynamically change the font style, so someone who wants a different default would have to change their CSS
file locally.
> 
> My thought was that perhaps people would prefer the serif font, as that is how it's done in Javadoc.
> 
> Donna
> 
> On Apr 1, 2011, at 7:33 AM, Randall R Schulz wrote:
>> Please tell me you're not forcing a serifed font. Let the user choose 
>> what their default font type is, please.
> 

Donna Malayeri | 1 Apr 2011 09:32
Picon
Gravatar

Re: Re: a spruced-up scaladoc

Agreed. For traits vs. classes, this is quite easy, but I can file an enhancement ticket for abstract types
in general. 

Donna

On Apr 1, 2011, at 9:20 AM, Nate Nystrom wrote:
> It would be nice to ensure that there's enough semantic information in the CSS classes and ids that someone
could easily write  a Greasemonkey script to change the default rendering.  For instance, abstract and
concrete types should have different CSS classes.

Philippe Lhoste | 1 Apr 2011 11:34
Picon

Re: a spruced-up scaladoc

On 01/04/2011 08:58, Donna Malayeri wrote:
> My thought was that perhaps people would prefer the serif font, as that is how it's done in Javadoc.

Where do you see that? In http://download.oracle.com/javase/6/docs/api/ pages, there is no 
font specifications except on some navigation items, and these are the classical 
Helvetica, Arial, sans-serif trio.

> On Apr 1, 2011, at 7:33 AM, Randall R Schulz wrote:
>> Please tell me you're not forcing a serifed font. Let the user choose
>> what their default font type is, please.

You wrote:
 > It's not so easy to allow users to dynamically change the font style

I don't think that's what Randall wrote. If you don't specify a font, the browser will 
pick either the default, or the fonts the user chose in its preferences / options.

Of course, Stylish, Greasemonly, user stylesheets and other facilities (depending on 
browser) can allow to customize these pages, but I doubt many users will make such change.

--

-- 
Philippe Lhoste
--  (near) Paris -- France
--  http://Phi.Lho.free.fr
--  --  --  --  --  --  --  --  --  --  --  --  --  --

Luc Duponcheel | 1 Apr 2011 12:05
Picon
Gravatar

I must be blind

Hello All,

why does this REPL session complain

scala> class Test {
     |  type P[X]
     |  trait Seg[W, Y, Z]
     |  sealed trait Seq[Seg[_, _, _], Y]
     |  case class PushPrompt[Y](p: Prompt[Y], seq: Seq[Seg, Y]) extends Seq[Seg, Y]
     |  case class SubSeq[Seg[_, _, _], Y, Z](val open: Seq[Seg, Z] => Seq[Seg, Y]) {
     |   def splitSeq: Prompt[Z] => Seq[Seg, Y] => (SubSeq[Seg, Y, Z], Seq[Seg, Z]) =
     |    p => {
     |     case PushPrompt(q, sep) => error("tbd")
     |    }
     |   }
     |  }
<console>:18: error: constructor cannot be instantiated to expected type;
 found   : Test.this.PushPrompt[Y(in class PushPrompt)]
 required: Test.this.Seq[Seg,Y(in class SubSeq)]
           case PushPrompt(q, sep) => error("tbd")


while this one does not

scala> class Test {
     |  type P[X]
     |  sealed trait Seq[Y]
     |  case class PushPrompt[Y](p: Prompt[Y], seq: Seq[Y]) extends Seq[Y]
     |  case class SubSeq[Y, Z](open: Seq[Z] => Seq[Y]) {       
     |   def splitSeq: Prompt[Z] => Seq[Y] => (SubSeq[Y, Z], Seq[Z]) =         
     |    p => {
     |     case PushPrompt(q, sep) => error("tbd")
     |    }
     |   }
     |  }
defined class Test


(the only difference is the trait Seg)

Luc

--
   __~O
  -\ <,
(*)/ (*)

reality goes far beyond imagination


Gmane