Gnome 2.16 has been released

Posted on 2006-09-08 at 06:47

Gnome 2.16 has been released. Read a review of the new Linux Desktop. Read the release notes for this version to see what's new. Love the gnome. Hug the Gnome.

Make A Comment

Eye Candy for Gnome

Posted on 2006-06-17 at 13:28

The workplace switcher for Gnome 2.16 is gonna be a bit blingier:

The new Gnome Dockbar bling is kinda neat too (watch it in action by clicking below):

And don't forget the wonderful new(ish) Gnome Bling Manager!:

That's among the many reasons why I say Vista is a waste of money!

Make A Comment

More mock ups of a next generation Gnome interface

Posted on 2006-02-07 at 08:37

I found more mock ups of what people are wanted/expecting out of the next version of Gnome. I've added them to my existing photo gallery of mockups so that you can go to one spot and look at them all.

Make A Comment

Mock ups of a next generation Gnome interface

Posted on 2006-01-18 at 20:47

These are not my own, but I was quite impressed with these ideas for a next gen interface look and feel for Gnome. I found them on gnome-look and art.gnome.

Make A Comment

Mono and Gnome APIs

Posted on 2006-01-13 at 14:45

I've said it before. I'll say it again. Linux will not ever become mainstream until it can make itself more approachable by the average corporate programmer.

I like Mono and what they have done for Gnome. But let's step back and look at the big picture. We need our own API, not a reworked version of one Microsoft has handed us. Take what's good about .NET and Java and come up with a new API that is specific to the Gnome platform. Moving away from a reliance on Sun or Microsoft's propriatary works, rightly or wrongly, will help to bring the community together. Additionally, making that new standard approachable in the way that Java and .NET are will help bring new developers to the Gnome platform.

Remember that Windows has users because it has applications. It has applications because it courts the average corporate programmer successfully. Gnome could learn a bit from that. It means slaughtering a few sacred cows of the Gnome world. An example? How about drag-n-drop form design? This is not an afterthought. If you want to be taken seriously by the corporate programmer community (the people you NEED to be taken seriously by!) then this is an absolute requirement of every IDE. Period. No debates. And I don't mean some sort of decoupled Glade interface that you can reference in your code, but an integrated drag-n-drop form designer where you can doubleclick a control and write the code associated with the default action of that control. I can design a gui in code just like you, but frankly I'm not gonna. Not when Visual Studio (hands down the best IDE on the planet right now...and I am an ardent Linux supporter) offers me so more simpler a way to do the same thing. Do you hear me Monodevelop? Do you hear me Eclipse? I'm talking to you!

Make A Comment

New and Improved List-O-Things That Gnome Needs

Posted on 2005-12-29 at 10:44

A while back, I wrote down a list of things I wanted to see Linux get. It's been a year and a half since I wrote that list and I wanted to revisit it to see what has been done and what needs doing now.

My first list in review:

So what about now? What do I think Linux and Gnome need now? Let me tell you:

Make A Comment

Gnome 3.0 Needs To Clarify and Unify the API

Posted on 2005-08-18 at 08:05

It's time to invite in the outside developers. It's time for Gnome to design a clear and hierarchical Application Programming Interface modelled after Java or Mono. Something that can be bound to many languages, but which is, itself, simple and powerful. Said API should be heavily based around the First Class Objects I described yesterday. A progmrammer that hasn't ever worked in Gnome should be able to sit down and figure it out. Something that looks sane like this:

Person John = New Person("John Doe")
John.SendMessage("Hi John!")

or maybe something like this:

Event BirthdayParty = New Event
BirthdayParty.AddPerson "Bob"
BirthdayParty.AddPerson "Alice"
BirthdayParty.AddPerson "Krusty the Clown"
BirthdayParty.SendInvitations(viaEmail, withRVSP)

Of course, those are extremely simplified, but I don't see why the final API couldn't be pretty close to that easy. A simple dotHierarchy like that found in Java and .NET or Mono is the way to go. No reason why programmers couldn't choose to write their apps in C or even Assembly, but the API bindings should expose the objects in a manner similar to that which I've described briefly here. This mess about each different subsystem having it's own API that may or may not bear any resemlbance to other APIs and which does not relate to the system as a whole sucks. It's ain't working. Gnome is not sucking in new developers at any serious rate. Take a play from Microsoft's playbook: Attract the developers, let them write interesting apps, then watch the users follow the interesting apps like breadcrumbs to your platform. It works.

Make A Comment

Gnome 3.0 Needs To Make Eye Candy Useful

Posted on 2005-08-18 at 08:04

Use Eye Candy. Use it liberally. Let there be transparent windows and morphing movies and glowing buttons. But make each piece of eye candy be justified. Make each piece prove that it should be included. The criteria should be as simple as Gnome itself. Does this piece of eye candy make the user's experience easier? If the answer is yes, then include it. If not, then discard it. Don't make it a circus of pulsating pixels, bothersome banners of information, and wobbly windows. Make each thing matter. And make it degrade gracefully. Test the video card. If the card can take it, turn it all on. If the card can't, only turn on that which ican be handled. What do I mean when I say "useful" eye candy? I mean a task bar that shows a miniaturized version of the task. I mean buttons that brithen when you hover over them, or glow when the user's attention is needed. I mean sliding sidebar windows that come out when the user must be informed of something. I don't mean transparent windows just for the sake of being able to see the desktop underneath. Transparent windows should be used only to let the user know something underneath requires their attention. that's what I mean when I say "useful".

Make A Comment

Gnome 3.0 Needs Focal Computing

Posted on 2005-08-18 at 08:03

Show only one thing at a time. Make applications take the screen. Conform the entire screen to the task at hand. If the user is typing a letter, let the whole screen be about typing that letter. This goes deeper than making all windows run maximized. I mean to suggest a screen that is intensely focused on the task at hand. If the user is copying files to a disk, let the whole screen be about those files and that disk. That is not to say the user can't move that task to the background and move on to another, but that while the user is focused on that task, make it the computer's whole world. Offer inline help related to the task, allow the user easy access to all related activities. If the user is playing music, let the screen be about the artist, the album, the song. Let the screen offer to find more music from that artist. Let the screen offer to queue up the rest of the album from which the song is playing. Let the screen offer the user an immersive musical experience. If the user is typing a letter, let the screen offer to look up words. Let hte screen offer to find related correspondence on the system. Let the screen offer help with formatting. You get the idea.

Make A Comment

Gnome 3.0 Needs To Own The Corners

Posted on 2005-08-18 at 08:02

SymphonyOS does this to great effect. Gnome needs to own all the corners. This is prime screen real estate. There are no easier spots to click than the corners and the four most common functions should be tied to them.

Make A Comment

Methods of Project First Class Objects

Posted on 2005-08-17 at 08:10

A Project FCO is a different sort of FCO in that it is really just a grouping of other FCOs. A user may choose to create a new Project, and drag in 3 or 4 Person FCOs, 2 Event FCOs, and maybe a Multimedia FCO (a presentation). In this way, he can track a group of objects at a glance without having to resort to a keyword search. A Project object would allow a user to:

Note that amongst the options is the ability to add or remove keywords. This would be a way for a user to add whole groups of objects dynamically, wherein any FCO that has a given keyword would automatically show up as a member of the Project. Also, note that the slideshow ideaa is a useful one for things like Pictures, sequential video footage, or similar content.

Make A Comment

Methods of Event First Class Objects

Posted on 2005-08-17 at 08:09

Events are appointments or tasks that are to be tracked, attended to or otherwise addressed. An event should alow the user to:

You may ask why I merge appointments and tasks into the Event category. this is becuase a task is little more than an appointment with a deadline rather than a specific time, at least by my way of thinking. It'd be easy enough to split these up if others disagreed. It's worth noting that this presents one way to set up a meeting: create an event, then assign it to other people, who in turn might recieve Event objects via email or IM. For an alternative way to set up a meeting, see the Project FCO below.

Make A Comment

Methods of Conversation First Class Objects

Posted on 2005-08-17 at 08:08

Technically, conversations are a sort of subset of the Person FCO, but there interest to the user warrants the promotion for FCO. The user should be able to click on a conversation (like an email or IM) and choose whether he wants to:

For the most part, this FCO is self-explanatory, so I'll not waste your time stating the benefits.

Make A Comment

Methods of Person First Class Objects

Posted on 2005-08-17 at 08:07

People ought to be considered FCOs. Much of what we do on the computer nowadays revolves around working with and speaking with other people. To that end, there should be People objects that have the following actions associated:

The user should be able to click on a person and choose whether he'd like to speak with them (via email, IM, VOIP, or by whatever other means are available) or visit that person's web page or blog. Additionally, if viewing People objects in aggregate, it might be nice to be able to view the blogs in an aggregated manner similar to blog planets or to stream a message to all of them at once ("Hey, just thought you'd wanna know that we're home with the new baby now and taking visitors!").

Make A Comment

Methods of Data First Class Objects

Posted on 2005-08-17 at 08:06

Almost as much as pictures, users like to be able to store random data. We are not as organizaed as we'd like to be, but computers should facilitate organization. To that end, Gnome should make it trivial for users to start a new table of data to be stored just as you would a letter to mom or a picture of your dog. That data should be easy to edit (something that looks like a spreadsheet would suffice) and it should be easy to find data when you search. People could keep lists, names, ideas, or whaever else interests them it these data files. Look to applications like Glom for inspiration here for easy user access to data tables. With such a table, users would be able to:

This is a slightly unusual idea, so it might require a use case example: Bob wants to help Alice plan their wedding. He starts by creating a new table and calling it "Wedding Guests". A step-through guide pops up and asks bob what he wants to store. Bob tells it that he wants a column for Names and a column for Where They Live. The table "Wedding Guests" is created with two colums in it. Those columns are, by default varchar(255) or something large and nondescript like that. This insulates Bob from the underlying details of designing a real data table. When he wants to start making his list, he clicks on the table and selects "Edit". This brings up a window with a grid in it that shows the list as it exists. Bob clicks in a line and types in the name he wants to add, then moves to the next cell to type in the state from which the guest will be traveling. Later, Alice decides to use Bob's table, since he's compiled the list already but she needs to be able to see whether they have RSVP'd yet, so she clicks on the table and selects "Modify the table", and she is offered a screen that lets her add another column called "RSVP'd?" Bob and Alice live happily ever after all because of Gnome 3.0.
This FCO is probably the most likely to cause controversy, as it is the largest departure from the traditional user experience, but I feel strongly that it warrants some consideration.

Make A Comment

Methods of Picture First Class Objects

Posted on 2005-08-17 at 08:05

Users like pictures. They like to store them, look at them, print them, and show them off. Graphics, in Gnome 3.0 should be associated with the following actions:

View and print are obvious, but I argue that a user should be able to take a group of pictures and convert the lot of them into a presentation (a multimedia FCO) so that the lot of them can be sent off to another person (a People FCO) for easy sharing of memories and ideas.

Make A Comment

Methods of Document First Class Objects

Posted on 2005-08-17 at 08:04

A Document FCO is any text-based file; be it an OpenOffice letter to Mom, a PDF file downloaded from the web, or a simple text-based reminder note (like those created by Tomboy or similar apps). With a Document, the user can:

For viewing, I'd argue that there should be a universal viewer for such documents. Evince is working toward being that universal viewer and I support that direction (for all that's worth).

Make A Comment

Methods of Multimedia First Class Objects

Posted on 2005-08-17 at 08:03

Multimedia FCOs would include movies, music, presentations and any other object that requires or allows VCR-like control. The gnome desktop would recognize the following actions with respect to Multimedia FCOs:

The salient technical points of merit here are that playback, recording, and editing of these files becomes decoupled. Though one application can cover all three modes of use, they need not. The user would just click on the file and get the cotext menu that asks what action should be performed. Perhaps a default action is stipulated so that the user only gets asked if he right-clicks the object.

Make A Comment

Methods of All First Class Objects

Posted on 2005-08-17 at 08:02

All FCOs will have the following actions associated with them:

Make A Comment

Gnome 3.0 First Class Objects

Posted on 2005-08-17 at 08:01

For version 3.0 of Gnome (currently codenamed Topaz), there has been some talk about centering the desktop experience around a set of what are being called "First Class Objects" (FCOs).

These FCOs will define and fence the user experience. For example, rather than firing up a Word Processor to write a letter to Mom, you might instead just "Create a new document" then "Edit the document". Though it seems like an obvious evolution, it is nonetheless a major break from the previous paradigm.

The user would no longer be concerned with applications, per se, but with the products of applications. The user experience would begin to focus on the end results of computing rather than the computing itself. Like any self-respecting linux nut, I have my own ideas as to how this should play out.

Firstly, I see 8 basic FCOs: Multimedia, Documents, Pictures, Data, People, Conversations, Events, and Projects. I describe how I understand each in the blog entries that follow this one.

I see an interface that strips away all the garbage that doesn't push the user toward these FCOs. Applications would be hyper-specialized but follow a design that blends them all together in terms of look and feel. It's the natural evolution of the Gnome HIG. Why open a whole mail program is all you want to do is read a particlar email thread? Aggregate as much as possible and open the APIs to allow developers to tie into the desktop aggregation. You can view some example screenshots of this sort of aggregation. You can read about the current developer thoughts on FCOs. My thoughts here deviate from those ideas slightly, but are heavily based on the same base concepts. The entries below that detail properties of those FCOs are an expansion of what is found on the linked official Gnome page.

And since I'm making sweeping statements about how others should do stuff that benefits me, let me add that gnome should start right away stealing liberally from SymphonyOS, which implements some really cool ideas.

Make A Comment

Multimedia in Ubuntu

Posted on 2005-07-18 at 08:02

To fix many of the problems I had with Ubuntu's multimedia, I did the following:

I still have to resolve the issue with quicktime where many of the movies (not all?!?) don't seem to have any sound, though they otherwise play fine. Weird issue.

Make A Comment

Gnome 2.10 feature preview

Posted on 2005-01-19 at 08:02

Someone has posted a draft of the Gnome 2.10 release feature preview. Looking good.

Make A Comment

Why is no one fixing the problem?

Posted on 2004-10-18 at 08:01

It's been a while since I blogged about personal finance options in Gnome. Things have not changed even a little bit. Frankly, I expected some movement on that front. Everyone knows its one of Linux's primary weaknesses (along with smaller hardware and miscellaneous issues).

Make A Comment

And speaking of .NET development in Linux

Posted on 2004-08-13 at 08:02

I've drooled and waited, and yet still after all this time still no Fedora install packages for MonoDevelop. Who do I gotta smack to get this done?!?

Make A Comment

My First Linux App

Posted on 2004-08-13 at 08:01

Having now written an app for Linux, I am left wondering what I should use my newfound great powers for. Forthwith, I present to the world, the "Click the Damn Button!" application written in C#, using GTK#, and compiled to run in mono on Linux:

using System;
using Gtk;

public class ButtonClicker
{
    public static void Main()
    {
        Application.Init();

        Window window = new Window("Click The Damn Button!");

        Label label = new Label("Name");
        Entry entry = new Entry();
        Button button = new Button("Click Me!");

        window.DeleteEvent +=
            new DeleteEventHandler(window_DeleteEvent);
        button.Clicked += new EventHandler(button_Clicked);

        VBox vbox = new VBox();

        HBox hbox = new HBox();
        hbox.PackStart(label, false, false, 12);
        hbox.PackStart(entry, false, false, 12);
        vbox.PackStart(hbox);
        vbox.PackStart(button, false, false, 12);

        window.Add(vbox);
        window.SetDefaultSize(200, 100);
        window.ShowAll();

        Application.Run();
    }

    static void window_DeleteEvent(object o,
                DeleteEventArgs args)
    {
        Application.Quit();
        args.RetVal = true;
    }

    static void button_Clicked(object sender, EventArgs e)
    {
        Console.WriteLine("You clicked the button!");
    }
}

Where is the proof that it works, you ask? Well, suck on this screenshot, bizzotch:

Make A Comment

gTweakUI: Just what the doctor ordered

Posted on 2004-06-10 at 08:01

gTweakUI is an app that offers an easy interface for changing settings that power users have been asking for for a while. If you are like and use Gnome, but want to change more than the standard UI offers without having to go into gConf, then this might be what you need.

Make A Comment

Lists are all the rage

Posted on 2004-05-31 at 08:01

And so I, too, will join in the fray with my List of Things Linux Needs Now (patent pending):

Make A Comment

Personal Finance Ideas in Gnome

Posted on 2004-03-22 at 08:02

What they need to do is create one that integrates fully in the new Gnome environment. Recurring or otherwise expected payments should show up as a "Personal Finance" calendar in Evolution. Among the list of valid payees should be the members of your Gnome address book. This should work in reverse as well. If I schedule a lunch meeting with someone, when I open the Personal Finance app, it should see that I had a lunch meeting and ask me if I paid, how much it cost, and whether it was a tax deduction or not. These are the sorts of things that will make Gnome a totally integrated, desktop leader. And why not make the Personal Finance software act as a system service like the Evolution-Data-Server? Other apps could query for financial info. OpenOffice Calc could offer more advanced financial analysis tools by grabbing data from the service. Evolution could add a "Business" address book that consists of people for whom payments have been made/received.

Make A Comment

Personal Finance Options in Gnome

Posted on 2004-03-22 at 08:01

I'm still waiting for a real personal finance solution for Gnome that rivals what Microsoft put out with Microsoft Money. There are a few out there (Kurush, GnuCash, CBB) but they either aren't complete, suck, or just plain look like ass.

Make A Comment

Easter Eggs

Posted on 2004-03-17 at 08:01

In Evolution, create a new mail message and in the "To:" field type "why?" or "easter-egg?" Cool.

Make A Comment

Monodevelop

Posted on 2004-03-04 at 08:04

Monodevelop gets its own web site. Additionally, it has put out its first release!

Make A Comment

Straw

Posted on 2004-03-02 at 08:03

Just emerged Straw. Doesn't work. Fails because of a "gnome" dependency. I guess I didn't compile Gnome with Python support or something lame like that. Irritating.

Make A Comment

F-Spot and gThumb

Posted on 2004-03-02 at 08:02

I'm all about Mono and I wouldn't mind seeing more of Gnome built on it, but I really like the gThumb application and it looks like F-Spot is starting to take hold in the Gnome community.

Make A Comment

Fuzzy Clock

Posted on 2004-03-02 at 08:01

Gnome gets a fuzzy clock. Neat.

Make A Comment

GTKFileSelector Widget

Posted on 2004-02-24 at 08:01

The new GTKFileSelector Widget is starting to look good. It's long overdue.

Make A Comment

Monodevelop

Posted on 2004-02-23 at 08:01

I'm really starting to drool over the upcoming release of monodevelop. Much goodness will be had.

Make A Comment