After a whole night of database churning (we need a faster server), alpha 8 is here. Send your bug reports and/or compliments to frassle@rura.org.

From yesterday:

Josh and I have been working hard on the next version of frassle. The scope of this release has sprawled; it started with a couple of features, ballooned into many more features and bugfixes, and then into bugfixes on the bugfixes. There's also the impending open source release, now with a somewhat barren SourceForge project page.

But tonight, features will be frozen. Bugs (OK, noncritical bugs) will be pushed aside. For this will be the evening of our deployment.

New features in alpha 8:

• major usability improvements to the publisher, including a wizard to shield you from the horrid complexity of noteset expressions
• an all-new aggregator, styled like Bloglines — and it actually works!
• an all-new full text search system — and it actually works fast!
• OPML import, so you can copy subscriptions from your favorite aggregator (so far) into frassle
• a smattering of icons, thanks to David
• pagination of the recent posts page, so you can go back in history
• many more small improvements and bugfixes

link

This is really more of a quick howto on the software development process focused around bug-fixing and feature implementation. A good read for aspiring testers.

link

A bounty of links to French learning resources.

link

The BBC site has an amazing collection of French learning materials online. I can listen while in the office, but I'll probably have to save speaking the thirty different example responses to comment ca va until I get home.

link

Transcript of Alexander's OOPSLA speech.

link

Rachelle, and anyone else: You're always allowed to talk about frassle! That's what being an open source project is all about: there aren't lots of secrets and restrictions and secret restrictions.

Speaking of open source, the alpha 8 release continues to be held up by my excessively social schedule. I may even skip today's Berkman blog meeting (although Rebecca MacKinnon and Bill Ives are likely to attend) in order to catch up on things. This weekend, Josh and I are going to focus on our upcoming OSCOM presentation.

link

There has been a very interesting thread among some frasslers recently. It started out with a question by a blogging newbie—Jennifer—about the basic motivations for blogging, but escalated into a discussion between Josh and MySQL-Wikipedia-dude (hereinafter MWd) on the relative merits of Wikis and blogs.

I like both Wikis and blogs, and I don't think they are distinct islands. I agree with MWd that many of the features common in Wikis—like photo and file uploading, version tracking, and multi-user editing—would be important for collaborative work (and often lone work, too). These kinds of features don't seem tied to either Wikis or blogs. They are all on the long list of desiderata for frassle.

So, if all these little features aren't intrisically tied to blogs or wikis, what are the differences between the tools? I want to look at three points:

1. The prominence of order in time as an organizing principle.
2. The ease with which a user can directly edit a piece of content she did not create.
3. The ease with which new relevant content can be found based on the context of an ongoing discussion or long-lived page.

1. The prominence of order in time as an organizing principle. Blogs push it; Wikis don't. On one hand, this makes blogs seem more fleeting—there is lots of stuff scribbled in that drifts off the screen after a little while. Frassle tries to straddle this line by offering another organizational system—categories. But reverse-chronological is still the primary view.

I would argue that, for all its downsides, time-ordering has two huge upsides: it's simple, and it works. For most people, it is too much work to decide upon an organizing principle for a growing body of work. On the other hand, without an organizing principle, one experiences what I'll call the existential quandary of wikis. What came first, the Wiki or its organizing principles? It's hard to add content when you don't know where it might fit in, so many people are reluctant to contribute to a Wiki. Although Wiki designers have included revision history in order to counteract that reluctance, revision control is way harder to understand than "it's a stack of posts ordered by time". With reverse-chron ordering, you might lose some stuff in the pile, but at least you know how the pile works. And hey, it's not so hard to find stuff either, what with all the linking and great search engines. So blogs bypass the existential quandary and yet retain many of the same benefits.

2. The ease with which a user can directly edit a piece of content she did not create. Wikis offer a new user more potential power. But not without a cost: despite impressive Wikipedia vandalism-recovery experiments, people still feel protective of the work they create. Additionally, when people who view the work you've made can easily associate it with you, it's possible to develop a reputation. The permanence and presence of this reputation is what incents people to sometimes compromise their immediate self-interest because of longer-term engagements and opportunities. Humans are socially evolved to put their stamp on the things they create, and the interactions they participate in. Few can contribute to the common good without wishing for at least some recognition. Blogs, on the other hand, are built around people; they even encourage some self-indulgence.

3. The ease with which new relevant content can be found based on the context of an ongoing discussion or long-lived page. The chief way of organizing a wiki is by linking. When there is a small collection of highly-interlinked content (e.g. a wiki for a single project), or a large collection of lightly-interlinked content (wikipedia), linking is either easy and necessary or hard but unneccessary. However, for many medium-sized Wikis, I find linking to be both hard and neccessary. Hard because there is a lot of new content, but neccessary because when something new comes in, if its useful period is to last beyond its time in the recent changes list, it must be linked from more established, related pages.

For this (large) class of challenges, I think Wikis offer only the illusion of simple, flexible organization. In fact, they tend to fall back to the reverse-chronological views typical of blogs, and the few relevant pages are drowned in a sea of mixed-topic recent-updaters.

By contrast, weblogs are really stupid. Nobody who has anything serious to organize or preserve will ever be satisfied with it appearing on her homepage for a day or two, then getting stuck in some calendar archive. This is why blogs offer tools like categories, search engines, technorati inbound links, subscriptions, subscribable egosearches, and so forth. While I think the most competently maintained Wiki can probably beat the best blog, in general the combination of organizing tools available for blogs earns them the win. It's a lot easier to find information relevant to the average weblog post than the average Wiki page.

Finally, although some of what I've written here seems to pit blogs against wikis, I don't think they are so opposite. I think that mainly, the tools we've seen so far have specialized on one end of these scales or the other, but that is a question of the tool designs, not conceptual constraints. Ultimately, I hope we can have tools that incorporate the range of features from blogs through wikis, and allow writers to select the right blend of features at every turn.

link

Thanks for these design hints. Right now, I'm fairly certain frassle won't appeal to the average AOL/Compuserve/Walmart user. As an alpha-stage application, its expected audience is the blog-loving early adopter. This person is likely to have a nice computer and a fast internet connection. That doesn't mean we should ignore the needs of such users, but at this point in frassle's development I think it is more important to make the basics of the interface usable. It's crucial we get the user experience right, because the powerful features of this system won't do anyone (but me) any good if there's no way to figure out how to use them. This is a challenge though, because frassle is premised on some ideas for which there isn't much user interface history.

So, if anyone has any interface ideas, I would love to see mockups, page flow diagrams, prototypes, etc. My personal preference is to start with pencil sketches in a notebook of a sequence of pages, then move onto the screen.

link

Andy, sorry about the bug where, if you attempt to post a note without a title, your post body is erased. This is fixed in the upcoming version.

link

Amid much ranting, this person (who I think I know) has a few good points about what in frassle is hard to understand. Some of these are addressed in the upcoming version, but others aren't. A lot of his points refer to categorization. Are you listening, Josh?