Re: g_format_file_size_for_display()

2007-12-18 Thread JP Rosevear

On Tue, 2007-12-18 at 15:31 -0800, Dave Benson wrote:
 On Tue, Dec 18, 2007 at 05:14:31PM -0600, Federico Mena Quintero wrote:
  Big deal :)  When you see 61.2 MB you think, oh, file system overhead!
  ripples in the time-space continuum.  Nobody cares about that.
  
  Back to my original question:  should this function be called
  g_format_size_for_display()?  It's not for files only.
 
 I have no idea whether to use SI or computer units,
 but the fact that there's debate suggests holding
 off til there's more experience.

Actually, I see it as an argument for centralizing the call and
abstracting the decision away from apps.  Then whenever its tweaked you
fix it in one spot and it instantly propagates everywhere. 

 Once a few cut-n-pasted versions exist,
 then you can decide if they are all the same...

They are already a few like gnome-system-monitor, gnome-main-menu,
presumably nautilus.

-JP
-- 
JP Rosevear [EMAIL PROTECTED]
Novell, Inc.

___
gtk-devel-list mailing list
gtk-devel-list@gnome.org
http://mail.gnome.org/mailman/listinfo/gtk-devel-list


Re: Announcing: Project Ridley

2005-09-14 Thread JP Rosevear
On Fri, 2005-08-26 at 07:53 +0200, Murray Cumming wrote:
  But being as part of the official library pack makes them be official,
  and avoid people using different solutions for the same problem.
 
 That's the idea of being in the GNOME Development Platform. I don't see
 how putting the whole platform in one tarball makes it much more official.
 
 That's just an invitation to JDKesque quality problems due to lack of
 modularity - meaning:
 - you have to wait to release a bugfix because of a bug in another part of
 the library that surfaced in the meantime, and
 - you force people to suffer bugs in another part of the library when
 releasing a bugfix for one part of the library.

Its a fine line, but I can't see how network connections don't fit in as
an acceptable low level operation when we have the following in glib:

1) lexical scanner
2) xml subset parser
3) IO Channels

Even if not in glib we should be creating an official solution that
hooks nicely in to the mainloop rather than neon/curl/soup.

-JP
-- 
JP Rosevear [EMAIL PROTECTED]
Novell, Inc.

___
gtk-devel-list mailing list
gtk-devel-list@gnome.org
http://mail.gnome.org/mailman/listinfo/gtk-devel-list


Re: Announcing: Project Ridley

2005-08-24 Thread JP Rosevear
On Mon, 2005-08-22 at 13:37 +0200, Rodrigo Moya wrote:
 what about libsoup/network library? Wouldn't it also make sense to move
 it to a libgnet in glib?

I'm also for this, right now we are using multiple networking libraries
and we fix the same bugs in multiple places.  I think its odd as a
platform we have no official way to great an http/network connection
(yes libsoup is in the platform for evolution, but for instance
gnome-vfs uses neon).

-JP
-- 
JP Rosevear [EMAIL PROTECTED]
Novell, Inc.

___
gtk-devel-list mailing list
gtk-devel-list@gnome.org
http://mail.gnome.org/mailman/listinfo/gtk-devel-list


Re: library dependency reduction (gtkhtml)

2005-03-30 Thread JP Rosevear
On Wed, 2005-03-23 at 18:58 -0500, JP Rosevear wrote:
 On Mon, 2005-03-21 at 01:31 +0100, Philippe De Swert wrote:
  Hi,
  
  On Fri, 2005-03-18 at 08:04 -0500, JP Rosevear wrote:
  
   How did you remove bonobo exactly? the components/ subdir needs those.
  
  Well I found out that the components part is for html-editing. It says
  actually html-editor in the components dir. As I did not need this
  functionality at all, as I was looking into rendering html I removed
  this part. So indeed my statement is not entirely correct. So my hack
  only applies to rendering, and you are very correct to point this out.
  However my point stands for other parts of the library (src, art, a11y).
  Declare the one small _() macro and a lot of deps disappear. I believe
  this is useful for reducing installed size and memory usage across
  several libraries.
 
 Are you just pulling in that definition for glib?

err, from glib

-JP
-- 
JP Rosevear [EMAIL PROTECTED]
Novell, Inc.

___
gtk-devel-list mailing list
gtk-devel-list@gnome.org
http://mail.gnome.org/mailman/listinfo/gtk-devel-list


Re: Themeable colors

2005-03-29 Thread JP Rosevear
On Tue, 2005-03-29 at 12:29 +0100, Bill Haneman wrote:
 typedef enum {
 
GTK_PALETTE_DARK,
GTK_PALETTE_SEMI_DARK,
GTK_PALETTE_MID,
GTK_PALETTE_SEMI_LIGHT,
GTK_PALETTE_LIGHT,
 
 maybe, but not appropriate to 'inverse' themes, for instance...
 
GTK_PALETTE_RED
GTK_PALETTE_PURPLE
GTK_PALETTE_BLUE...
 
 No, definitely bad ;-)
 
 because this would encourage app writers to do something that basically is 
 hard-coding colors.  What we need are more semantically specified colors.  
 
 So to extend the current notion of SELECTED, PRELIGHT, FG, BG, BASE, TEXT, we 
 need other style names that indicate the context in which a color is to be 
 used, but not its hue or value.

Something like HIGHLIGHT might be nice, useful in Evolution for things
like outlining the current day in the mini calendar, draw Marcus-Baines
lines, highlight overdue tasks.

-JP
-- 
JP Rosevear [EMAIL PROTECTED]
Novell, Inc.

___
gtk-devel-list mailing list
gtk-devel-list@gnome.org
http://mail.gnome.org/mailman/listinfo/gtk-devel-list


Re: library dependency reduction (gtkhtml)

2005-03-23 Thread JP Rosevear
On Mon, 2005-03-21 at 01:31 +0100, Philippe De Swert wrote:
 Hi,
 
 On Fri, 2005-03-18 at 08:04 -0500, JP Rosevear wrote:
 
  How did you remove bonobo exactly? the components/ subdir needs those.
 
 Well I found out that the components part is for html-editing. It says
 actually html-editor in the components dir. As I did not need this
 functionality at all, as I was looking into rendering html I removed
 this part. So indeed my statement is not entirely correct. So my hack
 only applies to rendering, and you are very correct to point this out.
 However my point stands for other parts of the library (src, art, a11y).
 Declare the one small _() macro and a lot of deps disappear. I believe
 this is useful for reducing installed size and memory usage across
 several libraries.

Are you just pulling in that definition for glib?

-JP
-- 
JP Rosevear [EMAIL PROTECTED]
Novell, Inc.

___
gtk-devel-list mailing list
gtk-devel-list@gnome.org
http://mail.gnome.org/mailman/listinfo/gtk-devel-list


Re: library dependency reduction (gtkhtml)

2005-03-18 Thread JP Rosevear
On Fri, 2005-03-11 at 13:30 +0100, Philippe De Swert wrote: 
 Hello all,
 
 While doing some embedded work I decided to use gtkhtml to make a basic html
 helpviewer. This work fine, so it is not about that. But I neglected the
 several dependencies, which turned out to be quite high. So as the application
 was already working fine I decided to take a look at trimming down gtkhml. I
 started out with removing all includes that included bononbo includes and
 discovered that everything compiled fine. After solving a little problem
 related to () and N_() being defined in bonobo-i18n.h included from
 gnome-i18n.h by copying in the relevant stuff to gnome-i18n.h everything
 worked. So this relatively simple task removed three dependencies (libbonobo,
 libonoboui and libbonobo-activation) for this lib. Because of my problem
 described earlier this may actually happen more (and not only for libbonobo
 etc). I am going to try to find out about other inter-library dependencies
 that might exist (in my particular case) and maybe I can find some stranger
 relations. If these can be removed this could make a substantial memory
 reduction for gnome/gtk desktop/applications. As I do not have very much
 experience with all this I would like to have feedback from more experienced
 people (and I hope this post might inspire them to look at the number of
 useless includes that might actually add bloat when using gnome libs)

How did you remove bonobo exactly? the components/ subdir needs those.

-JP
-- 
JP Rosevear [EMAIL PROTECTED]
Novell, Inc.

___
gtk-devel-list mailing list
gtk-devel-list@gnome.org
http://mail.gnome.org/mailman/listinfo/gtk-devel-list