Discussion of fields in Microcontent Items. These fields describe the internal structure and the meta-data.
Today a new Twitter client arrived on the Mac: Tweetie. So time to check out the state of Twitter clients.
Since a long time I am a user of Twitterific. It is a simple client that seemed a nice fit to Twitter. You get a small icon in the menu bar, which changes colour once a new tweet arrives. Clicking on that icon unhides a special window, with a sequential list of the latest tweets. Each tweet shows an icon, the name of the author and the tweet itself. Authors and url’s have a different colour and link to the Twitter-site. Each tweet can be acted upon, such as a reply tweet, etc. Twitterific also interacts with Growl, so that each new tweet is presented on the screen for a short time.
Twitterific is really a downsized MicroContent Client. The new Twitter clients however show that a real full-sized client for Twitter has its rightful place on the Mac.
The first Client is Tweetie. The first thing that stands out is the sidebar with four options. With this the user can select his timeline, reply, direct messages and search. Very nice if you want to zoom in. I think I will like the search option. C licking on a tag will create a new tab with a corresponding search. Thus the tabs will show your history. The same happens if you click on a @username. This client also supports multiple Twitter accounts. It is also very easy to post images.
And then there is Nambu, which brings Twitter in the real MicroContent Client age. This Client has a real sidebar (list view pane) with categories (Twitter, Links, Nambu, Search). The Links category shows all tweets with url’s that I have recieved and sent. The Nambu category seems to be related to a Nambu url shortener and picture uploader.
The search category has a standard list with the most used words in tweets (current/today/this week). I do not know how long ‘current’ is. But Susan Boyle is on top in all three trend periods. The most important category is naturally the Twitter category with standard subcategories: home, mentions, direct, sent, favorites, people and groups.
Tweets are shown in standard inverse time order with the user icon, twittername/full name, tweet time and the tweet itself. One can easily reply, retweet or private tweet. (and can I recognize other Nambu user tweets?). Nambu indicates which tweets are unread. As all tweets seemed too be stored locally, it is possible to delete all tweets, read tweets, set all as read, etc. It also easy to follow/unfollow. Very nice is the possibility to translate tweets easily. It did work well for japanes and korean tweets.
A tweet is parsed such that hashtags, links and users are found. Clicking on a hashtag creates a search list for that hashtag (sweet). Clicking on a username opens a new window with the tweets for that user. It is then easy to follow or block that user. Interestingly shortened urls are expanded, so that you can see where they point to.
The people category can show information of your social network (followers and following) on an image grid. Hovering over an icon shows the information on that person. Groups allows you to define a group of people from the ones you are following, so that you are no longer overwhelmed. This again shows feed. (I guess people should be moved to a separate category).
Nambu is a very interesting Twitter client. I will continue using it.
Nambu should work on its menu bar. That is now totally beyond logic. There is also a secondary tool bar. This tool bar is different for each list. I get however the impression that it could be part of the main toolbar.
A new application for tagging on MacOSX is introduced: Tags by Gravity Apps. This allows a user to add tags to files, music, web-pages etc. Any application that supports AppleScript in the right way is supported (OmniWeb: fail, MarsEdit: fail). One can use Spotlight are the tags application to find tagged files. Tags has a simple tags browser (could be improved).
This is a very interesting application. I do not understand why they use the weird interface, let me at least select another, standard one.
I have a feeling that the standard way of finding files, the Finder, is slowly morphing into something else with all these addon’s and applications. Time for Apple to step forward here.
I started playing with the Listphile service. This service allows a user to create and manage Lists of Items. One can create either some standard lists, such as text, people lists, video lists, images lists and an atlas. Or one can add fields at will to create one's own list. In order to experiment a bit, I create a list with sundials. I added some fields to the standard atlas list to suit my need.
This service offers many features of desktop MicroContent Clients, so I will discuss the service using that checklist.
The first check is whether the Items in Listphile are MicroContent. Indeed these Items support structure, are flexible (one can extend them) and offer some MicroWeb possibilities through links. There are standard fields, such as titles, descriptions and tags. And each Item has a permalink.
The interface consists of only 2 web-pages: a View-pane and a Items-pane. The View-pane shows a single Item, the location is shown on a map. The Items-pane can be set to maps, lists or thumbnails mode. There are limited sorting possibilities. A separate page is available for a tag cloud.
The service has standard Item functions for viewing, adding, editing and deleting Items. There are only limiting functions for a List itself (editing title, address, etc.).
There are only limited functions for importing an exporting, i.e. no importing and exporting only as RSS and embedded Lists.
The fact that this service is on Internet allows to add some news features. A List can be a social List. This means that multiple authors can work on a single List (still have to check this out). And to help here one has the possibility to see authors and what has been edited. There is some moderation and access control. It is also possible to comment on an Item or author.
I miss the import and export possibilities. I would like to exporting as KML and OPML. I am now locked in. The service is still in beta and I did see some minor errors, but overall it works great.
There has been some discussion recently on the state of tagging. I did not follow all the ins and outs, but I get the impression that the pundits agree on the fact that we are waiting for the next phase of growth and then mainly in the enterprise world.
Ian Beck gives some advice no tagging. Basically he says that you first have to identify the Items (MicroContent and Files) you are tagging and then about the tags themselves. And the crux seems to be to chose a consistent method.
The way he talks about tagging reminds me of a more formal way of tagging, whereas I would prefer a more informal way. The formal way (and then I am not talking about ontologies) is for me already the method one uses to gather Items is folders. And these can be file system folders or more flexible ones in MicroContent Clients. And this formal way also refers to the colour labels of MacOSX. The formal way of tagging implies a strict use of terminology, where changes are difficult to make afterwards.
The informal way is more flexible and basically you can do as you like. One can introduce errors in this method as well, but those seem to be easier to repair. This informal approach however requires the user to traverse the tag cloud in an easy way, so that ill tagged items can still be found. And traversing should include the relation between tags, so that synonyms, etc can be found. The application that does this well is WebnoteHappy.
I have been playing a bit with HoudahGeo. This application allows the user to add geographic coordinates to images. For this the user only has to select an image and move the pointer around on a globe (bases on Google Maps). The images can be taken from the iPhoto library or a plain image file can be imported.
Once an image has been geotagged, the location can be exported as Spotlight tags, as EXIF info. Also the images can be exported to Flickr and to Google Earth as kmz file. The app also has support for GPS tracks (can not try that out).
I have mixed feelings about this application. I rather have something that is integrated with iPhoto. I do the Flickr exporting already there. HoudahGeo creates new sets for images with geocodes, but I already have those sets in iPhoto. The export as kmz-file is very nice.
Well I am still grasping with my workflow around locations. I am really waiting for camera producers to integrate GPS.
A nice discussion on the question whether tagging is a disruptive innovation. If I look at myself, I do not find it disruptive, yet? I am still struggling in using it. I like to use it, but have not yet find the right workflow. I am investigating whether I can extend tagging to the files in my filesystem.
The question of disruption is interesting. Are there other themes in the MicroContent world worthy of disruption. If I think about audio, video and images, I do see a disruption. For digital images it is all clear, the old camera passed away already. Sharing, printing and taking images is now done in a totally different way. And I guess the same is now how happening for audio. And we are just waiting for it to happen in the video market.
Tagamac makes a nice comparison between file libraries and file system from the point of view of tags. One extend the arguments to any MicroContent Type. We are slowly moving towards libraries. These libraries hide the file system details from the user. These libraries allow users to organise their files in any way they want, i.e. through the idea of (virtual) Lists (hand-picked and smart). The use of the structure of MicroContent (internal and meta) helps the user find and organise his files even better. The File System is really passé.
My second post on Yep looks more closer at the tagging support in this application. The fundamental idea of this application is that it allows a user to more easily manage his PDF documents. In order to help this management the application uses tags.
These tags must be set by the user. There is an auto-tag feature, but that does not work well for me. It might change when I tagged more documents. However one can define a set of tags taken from the file location, which works nice. I see a folder structure as basic tagging, just as using categories.
The user can type his tags into the tag field. The application helps here through auto-completion and a showing a set of popular tags. I noticed that I should follow a common set of rules for setting tags: no uppercase, no plural. But this also applies to the naming of my folders. I wonder whether I can change that. Or will that trouble my backup procedure?
Using tags for discovery of Items works quite nice. In order to use this one has to set the List-pane into tags-mode. This will show all tags as a tag-clous, which is presented as a list. The size of a tag indicates the popularity of the tag. Hovering over a tag shows that tag also on relevant Items in the Items-pane, which is nice. Clicking on a tag, changes the the Items in the Items-pane to only the Items with that tag. Thus this works similar to a pre-programmed smart-list. However also the content of the list-pane changes to show only the relevant tags for the new set of documents. This is similar to the browser approach of WebnoteHappy, but I prefer the latter. Interestingly I found documents that I did not place yet in the right categories (i.e. folders).
I think I like the application. I would like to see this approach for all files, or better for all printable files. I say here printable files, as these can be presented as PDF's. So I have to use the application a bit longer.
I came across an application called Yep. This application allows a user to manage his PDF's. The application has many features similar to MicroContent Clients, so worthwhile to have a look. And also to look into the difference between PDF's and MicroContent Items.
To be clear PDF's are not MicroContent. PDF's lack the internal structure of MicroContent. Although the internals of PDF's are accessible, such as the text, it is still not really structure. However this is what makes a PDF different from an Image. However the fact that a PDF can consist of multiple pages, it is no longer a simple Image. And thus a PDF is not something small either.
What Yep makes clear is that the standard solution (Finder on the Mac) for the management of PDF's is insufficient. And I guess the same can be said for files. So what does Yep add?
Unlike a standard MicroContent Client Yep can not show the content of an Item, so it only shows the meta-content. This meta-content is taken from the meta-content of the file-system, such as the file-name, creation dat, file, size, etc. Yep seems to add a extra layer on top of that data, though. This layer contains the tags, the author and an URL for instance. The tags can be fed back into the comments of the file meta-data to allow Spotlight searching.
The main view of the application is however the Items-pane. This Items-pane can be set into table or image mode. The table-mode shows as the default the file-name, creation-date, author and number of pages of the PDF. One can add/remove columns and set ordering. In image mode one sees the first pages of the PDF's on a rectangular grid. And can zoom and out, and a loupe is available to take a close up view of an Item.
The List-pane is for creating collections of Items. There are three modes available: file location, collections (hand-picked and smart) and tags. The tags are a combination of words take from folders, documents and hand-set. The application has various ways to browse through the tag-cloud, which I will look at in more detail. And I will analyse the application in more depth.