Difference between revisions of "User:Mirek2"

Revision as of 00:18, 5 July 2009

Hi. Um, I'm behind the Iced Coffee proposal. I love the open-source movement and hope to contribute more to this and other projects. Feel free to criticize my ideas -- it will only help solve the problems.

By the way, here are some impressive pieces of software which we could take hints from:

• Apple iWork -- Numbers, notably
• Google Chrome, notable for its simplicity and intuitiveness
• Macromedia Fireworks MX. I truly love the interface of Fireworks, at least that of the old Macromedia version; the tools and the panels create a very nice workflow (the properties panel is super), and vector and bitmap tools blend together seamlessly.
• Pixelmator. It's funny how the interface can fit in so well under Mac OS X without actually sporting its native look. Pixelmator draws the user in with design and simplicity alone. That's a feat we should strive to accomplish -- eye candy is what attaches the user to the software and makes working with it enjoyable.
• Apple software, in general. Not to sound too Apple-centric, but, generally, Apple does make software that is, for the most part, logical, intuitive, uniform (with other software), and visually pleasant. What bugs then are the smaller things, which Apple rarely takes effort to perfect.
• Adobe Buzzword. To be clear, I really hate Adobe and think of it on the same level as Microsoft. But I do want to point out its word processor, which, design-wise, is very nice. I haven't used it very much, so I can't make judgements on the interface, but by a cursory look it seems very simple and organized, for the most part. It can also help us determine whether we want to use an accordion or not (personally, I'm still against it; the interface, which constantly moves items from one extreme end of the app to the other, annoys me)

Thoughts

Future Vision

Perhaps this isn't really reflected in the proposal, but here's what I think software in general should move towards: minimalistic, simple, powerful, easy, flexible, and streamlined. That means having a lot of things contextual, right there, just a click away when you need them, never there when you don't. It means merging things that belong together. It means having the application work almost anywhere, with any resolution, on any operating system, with any hardware. And it means removing stuff that doesn't belong (but maybe putting it into extensions). But most importantly it means having the most logical interface, so the placement of features is blatantly obvious even to a novice.

In terms of OOo, I think we need to sort all the commands, from scratch, into categories that actually make sense and accommodate the number of commands nicely and evenly. We can remove stuff like blinking text, which just encourages bad, annoying behavior, and merge features ("underline," "strikethrough," and "overline," templates with "New Document," "Recent" with "Open," "Customize..." with "Options," etc.). I am a strong proponent of a web-based OOo, which would bring OOo under virtually any platform, force it to be lightweight (without extensions, which would probably work only with the desktop version, which could be based on technologies like Mozilla Prism or Google Gears), and significantly improve collaboration and on-the-go edits.

Sadly, this ambitious vision is not even really reflected in this proposal. But I do hope that it inspires some good, creative ideas.

On Vertical UIs

Hm. I've been called on to make the central UI element in OOo vertical by default. Here's how a vertical UI would work:

• After the OO.o install, the user would be met with a choice of common UI options (as well as a link to the options dialog). I'm thinking these could be "Horizontal UI," "Vertical UI," and "Classic UI" (menus and toolbars). These would always be available from the Options dialog.
• The sidebar, which, like the fluid bar, provides quick access to styles, ways to edit them, and other proper tools, is almost as powerful as the fluid bar, so it could be used instead of a vertical interface and keep a horizontal one.

On Quick Access

In Iced Coffee, you can customize a quick access toolbar, but there is no such thing by default. Instead, quick navigation with temporary toolbars on hover over Fluid Bar categories should suffice.

On consistency

A forewarning: I'm a programming dummy, and I know nothing about how the OO.o code is set up. However, I have a hunch it's not set up in the way I'm describing below. (I know OO.o already ties together [that's why you don't download individual apps], but loosely.)

Think of OO.o as one extensive toolbox, and the different parts of the suite (Writer, Calc, Impress, etc.) as different types of paper used for different purposes, but all requiring pretty much the same tools for editing. That's what OO.o should really be. By keeping everything uniform across the applications, save the "paper" and the tools unique to individual applications, we can significantly trim down the code. Essentially, we'd be coding one application instead of the four/five/six we have (I'm not really sure how Base and Math fit into the picture). The tools unique for one application could be put under the app's default "options" file (as options would now be saveable), where command categorization would normally go (e.g. these "unique" tools would exist in all applications, but the options file would hide them). Thus, the kind of "paper" (basically blank pages for Draw, pages consisting mainly of a frame/frames for Writer, tabs of infinite tables for Calc, and animated pages with a fixed width for Impress) would form the essence of the individual applications.

What still needs to be done

While this is a nice step forward, a more organized UI is just a small step in battling to compete with the proprietary world. From the User Survey, it seems like we need to focus on the following issues:

OO.o incompatibility, especially regarding Impress

This is the biggest problem users seem to have in moving to OO.o. We need to make the suite completely compatible, and if we can't do that, then at least build in an indicator telling what might not work the same in MS Office. Sure we have the incompatibility warning when saving to MS files, but those don't tell the user anything about if and what exactly will be different.

We can start with a collection of compatibility test files (like the acid tests for browsers) containing a variety of things (the more commonly-used, at least; we don't need to focus as much on the advanced-user incompatibilities right now) currently incompatible in OOo.

Bug fixes, streamlining, getting rid of JRE dependences

OpenOffice.org is a pig. That's pretty much a wide-known fact nowadays. It also crashes often on certain systems (I've experienced a few crashes myself, and I consider my computer technologically up-to-date). And it has annoying Java dependences.

To get all of this fixed, we need to encourage more people to contribute. We need an easy-to-use, simple, clearly-worded, easily accessible website for OOo. There needs to be a set of tasks for each project available on the project's homepage, or linked to from one. There needs to be a clear, concise description on how to contribute, also on the homepage, above the task list. The task list would also indicate the progress of individual tasks/projects, and link to what has been done so far. Tasks would be sorted by urgency (what needs to be done most). And then there needs to be a link to all the resources one can use (the OOo logo for the artwork project, for example) and another link to the mailing list's past discussions (or we could integrate with Google Wave and run discussions right from the website). Lastly, there needs to be leadership and organization within each project.

So that's the deal with project sites. However, for those discouraged by the contribution process, we need a simple feedback/bug reporting site, linked to blatantly right from the OOo homepage, and, based on the category/ies specified, delivered to the correct project(s).

Design

Our artwork team needs to get organized. There needs to be ONE simple, clearly-worded, easily accessible website with a list of stuff to work on and a brief tutorial on how to submit proposals. Actually, this needs to happen with all the community projects.

We need to work on just about all graphical aspects in OO.o: the templates/themes, the gallery, the GUI, and styles. And we can also borrow from others. Although I never used it, the Open Clip Art Library seems like a good place to take clip art from. Lotus Symphony's Gallery is also a nice source for templates and clip art, but it doesn't seem like it's open source...

On Making Money

I'm not completely aware of how OpenOffice.org makes money, or how much money it needs and if those demands are being fulfilled (the depth of my knowledge is contained within these articles, on the same topic: [1] [2] [3]). However, I'm not sure if the direction it's taking (again, I'm basing everything only on the above topic) is a good one. Personally, I hate when software comes with crapware. A Google toolbar... okay, fine, I can take that. That won't actually affect my user experience with OpenOffice.org. And I definitely won't install it, since I think toolbars are dumb. An MS toolbar... a bit strange, advertising your biggest competitor, not to mention the software giant that got to its position unfairly and everybody want to see dead... borderline.

But I would much rather see text-based ads on OpenOffice.org. Those don't really affect my experience -- they're relevant and undistracting, don't take up space on my computer or affect its performance. I've lived with them for years in GMail and don't even notice them anymore (I even had to revisit GMail in writing this, to make sure they're really there). And they're the cash cow that feeds Google. Or maybe something else... The bottom line is, I think we should strive to keep advertising "optional" and not force it on the user, whether through installation or right through the office suite. With Firefox, things are different -- their customizable search box would be there no matter what, and putting Google as the default engine has no effect on the user's computer (it doesn't take up space and it doesn't have a toll on memory usage). If OpenOffice.org were to include online collaboration features, and it made Google Docs the default client (but kept clients customizable, like Firefox does), I'd have no problem with it.

On "Autos" and "Smarts"

Automatic actions, mostly AutoCorrect, have been, in my experience, a pain for the user in some cases. I have to admit that it also caused me pain a few times, notably automatic capitalization (I hate it so much; it's not like I can't capitalize things myself) or automatic spelling corrections (when I'm typing in Czech and forget to set the langauge to "Czech"), not only in OOo, but in competing suites as well. Then I have to use OOo's awful options labyrinth, look through a thousand options, and finally deactivate the things I don't want. Some of my less computer-savvy friends were victims of these automatic actions, with a number of incorrect "corrections" on their printed copy with the excuse of not knowing how to make it normal. So here's what I propose we should do:

• Make AutoCorrect less powerful by default (for example, uncheck automatic capitalization by default). Advanced users could still easily turn these options on through (a revamped) Options.
• Show an infobar when an AutoCorrect correction is made (the InfoBar), containing a button to undo the one correction, disable the particular AutoCorrect correction (e.g. "Stop capitalizing for me"), a link to the AutoCorrect Options section, and a "Don't show this again" button.
• The label "Smart," initially used as a synonym of "automatic" or "contextual," has been abused by a number of applications (such as "SmartArt" in MS Word). OO.o is better off not using it.