Well-connected applications

In a brief exchange with Steve Hodgson (@BestofTimes) on Twitter, I recommended TextSoap — a brilliant utility for munging and cleaning up text in a multitude of different ways. It's not a very glamorous application, and it seems pricey until you actually use it intensively. I've had it a while now, and on most days I use it once or twice, but it really pays its keep when I have big tasks that require a lot of text manipulation.

Yesterday I had to copy a lot of information from PDF files and emails and enter it into forms on a website, many (many) times over. Copying from PDFs resulted in odd problems with the text like line breaks instead of spaces between words, orphaned hyphenation and so forth. Since the forms required strictly ASCII text, I also had to convert curly quotes to straight quotes and accented characters to LaTeX format accents.

That would all have been a huge chore, but for TextSoap. It has a variety of built-in 'cleaners' (for example, straighten or smarten quotes), but you can also build your own cleaning components using regular expressions, and you can chain together the existing cleaners to form custom workflows. So cleaning the text was only a case of a making a few clicks.

However, TextSoap doesn't really do anything you couldn't do with a text editor and good knowledge of regular expressions and scripting languages (it has a 'convert to Markdown' cleaner, for example), albeit that it packages those functions up in a user-friendly way. The real beauty and utility lies in how easy TextSoap makes it to access those functions. You can use the Services menu (which has incidentally become a lot more powerful with Snow Leopard), which also means that you can assign global keyboard shortcuts to your most-used cleaners. You can use the contextual menu or even Automator workflows, and there are specific plugins for text editors like BBEdit. So whichever application you are in (provided it supports Services and/or Applescript), you have easy access to the cleaners.

That — it seems to me — is one of the most powerful things about Mac OS X as a platform, and the thing I miss most when I use other platforms. There is standardisation of a lot of important features¹, and a lot of different technologies that allow you to connect your applications together in useful ways that save time and effort.

¹ For example, I love Linux dearly, but it drives me absolutely crazy that there's no standard keyboard shortcut to quit an application. ↑

QuicKeys

I've had a bit of an on-and-off relationship with QuicKeys, but it has certainly been a long one. When I have stopped using it for periods, it has usually been because I have adopted other ways for accomplishing the kinds of tasks it deals with, and it seemed like overkill to have a separate application running to deal with those things. However, with version 4, QuicKeys has become even more powerful, versatile and easy to use, and I am using it in earnest again.

For those of you who haven't encountered QuicKeys, it could be described as a macro utility for your whole system. You can create 'shortcuts' from a series of steps which automate actions that you would otherwise perform manually. There are a very wide range of possible actions in steps, from executing applescripts or shell scripts, to selecting from menus, manipulating windows, typing text or dozens of other things. You have quite a lot of control over the timing of these events and whether you need to wait for a particular window before moving on, which helps a lot in making the shortcuts reliable.

Once you have constructed your shortcut, you can trigger it in many different ways, from the obvious hotkey or mouse click to running if it is a certain date and/or time, if a certain volume has just mounted, or if an event occurs in another application. What makes this even more powerful is that all of the shortcuts can be limited to certain scopes (i.e. active applications). In practice, this means that you can reuse triggers in multiple applications without worrying that the wrong thing will happen. So you can — for example — launch a particular web page when you press F1 in Safari, and check for new email when you press F1 in Mail.

There are now also abbreviations (text replacements which happen automatically when you type a trigger) which replaces the need for TextExpander or similar utilities. In fact, the scopes make it very easy for me to type two dashes and a space and have them replaced with the HTML entity for an em dash in MarsEdit, and a unicode em dash in a rich text editor (which I've already done a couple of times in writing this article!).

{Read more...}

Alpine

A few weeks ago, I got an EeePC 1000HE netbook for work. I needed a dedicated, cheap machine to run some Linux-only stuff on, and I thought it would be nice to have a lightweight portable to take on trips. I'll write more about the EeePC later, but I wanted to mention my new-found love affair with the console email client, Alpine. I installed the ArchLinux distribution with the Awesome window manager, both of which are very lightweight and speedy, and I wanted an email client to match.

Over the years, I must have used a pretty large number of email clients. From Eudora, Mailsmith, a brief and regrettable affair with Microsoft Entourage (about which we will never speak again) to settling for a number of years on Apple's Mail, it actually all started with Alpine's predecessor: Pine. Back in those far-off days, I used Pine by connecting to the University's VAX machine, and I hated it. To be fair, it has come a long way since then, and I have learned to love the command line. After a few days of playing with Alpine on the EeePC, I decided that I wanted to switch to it on my Macs too.

{Read more...}

New photoblog

I've been irritated for a while by the setup I had for my photoblog, Wings Open Wide. It looked nice enough, but I had to upload my photos manually, title them, describe them and set the location and so on. That was annoying when I was also uploading to Flickr, and thus duplicating effort, so I started to post fewer photos to my photoblog. I could have just swtiched to Flickr and directed people over there, but it feels a bit impersonal. Flickr is great for the social aspects, but it doesn't feel like my space. I also upload more stuff to Flickr than I want to display as my photographic portfolio.

I considered a variety of different approaches, which varied in complexity and functionality, before stumbling on Flogr. It's a fairly simple but rich PHP photoblog, which pulls your Flickr photos into a gallery that you can style to your liking. By default, it gets all of your photostream, but you can also restrict it to certain tags or photosets. I've set it up to pull in photos with the tag 'photoblog' which makes it easy for me to be selective. Optionally, you can cache the photos and information in a MySQL database, speeding up display somewhat. If you click the 'Details and Comments' link, you can see — well — details and comments, and you can click the comment link to go to Flickr to leave a comment. There's also a nice Lightbox view for the recent photos, allowing you to navigate between them, or view them full size by clicking the title at the bottom left.

I like it a lot, so I've set it up here. I didn't want to break any existing links, so I've left a link to the archived Wings Open Wide on the new page. I might do a bit more restyling of the page, but I'm pretty happy with it as it is. One thing I haven't quite sorted out yet is the display of thumbnails on this blog. I've used the Flickr badge, so it shows the right images, but links directly to Flickr rather than Wings Open Wide. That's not a big issue, but I might see if I can fix it at some point.

Fever

I've used RSS feeds to keep up with the blogs and other websites I wanted to read for a long time. I used to use NetNewsWire, which was (and still is) a great bit of software, but when Google Reader came out, I switched to that. I liked the quick keyboard shortcuts to navigate around, and the fact that using an online reader meant that I could read feeds on any browser and not have to deal with items that I'd already read elsewhere.

However, recently I've been finding that I'm overwhelmed with information, particularly if I've been too busy to check Google Reader for a few days. I subscribe to a few high volume feeds for the occasional useful item that they throw up, but that means an awful lot of stuff I'm not interested in to wade through. That's why I was intrigued by Fever, a new feed reader by Shaun Inman, who also designed the lovely web site analytics program, Mint.

There must be hundreds of feed readers out there, but Fever distinguishes itself in two ways. First it is a self-hosted service, meaning that you have to install it on a web host. This would be a problem for those without a web host, but it does mean that it is available on any computer you choose to access it from and also under your direct control, unlike Google Reader. Second, it offers a way to pick out interesting items from your feed, without overwhelming you with information. To do this, you assign your high volume, high noise feeds to a special 'Sparks' group. Any items that share links or topics in common with your other 'Kindling' feeds will promote those topics to the 'Hot' list. This shows you feeds grouped by topics, so you can skim them or read the individual feeds as you wish. It's a great way to keep up with the things everyone is talking about, without being submerged by your unread feeds count. Other people's Delicious feeds are perfect for Sparks, and you can add feeds to the group with impunity knowing that it's going to make your Hot list more interesting, rather than bogging you down.

It's always a bit of a gamble paying for something ($30 in this case) without being able to try it out, but the screencast was reassuring and showed most of the fetaures. It's really easy to install (unusual for a self-hosted service) and importing my feeds and groups from Google Reader was quick and accurate. The UI, as you can see in the screenshot above, is very clean and rather pretty. There are Google Reader-like keyboard shortcuts for most things, and it's easy to move around the feeds. There's also quite a lot of customisability at the level of groups or individual feeds, and by default, feeds with no unread items are hidden, making it easy to see what you've got. Shawn Blanc has some good tips for using Fever efficiently, and he's right that it works very nicely as a standalone Fluid app. But it's also good as a tab in Safari, particularly if you use this tip to make Safari open links with the target="_blank" attribute in a new tab, not a new window.

Fever also has a truly lovely iPhone interface — one of the best I've seen — for reading feeds on the iPhone. Of course, because it's getting data directly from your Fever installation on the server, there's no syncing issue, but that also means no offline reading, which iPhone apps like Byline offer, and which can be useful at times. If you have long articles to scroll through, the interface on the iPhone can be a little bit longwinded, so a few more navigational aids would be a benefit, I think, but other than that, it's really slick.

Fever is only at Version 1.0, but is already a very competent and useful web application, which distinguishes itself well from the competition. The only thing I really miss from Google Reader is some way to share articles I come across. It would be nice to have a command to post items to popular services like Delicious or Twitter, or to share items publicly. I'd also love a command to save an article to Instapaper, particularly on the iPhone, where you'd like to save long articles to read at your leisure. Alternatively, if there was an API available of some kind, people could find ways to pipe their saved articles to other services. I'm sure that these kinds of improvements will come with time. Shaun has done a great job, and it's obvious that he's thought quite deeply about what a good feed reader needs.

Tweetie for Mac

Tweetie -- a Twitter client for Mac OS X (the 'desktop' version, as I suppose we must refer to it now) -- was released today. I use Tweetie for the iPhone, and like it so much that -- until now -- I tended to access Twitter on the iPhone rather than the Mac, even if I was sitting in front of my computer. It's difficult to define exactly what is so good about it, but the smooth behaviour, beautifully designed user interface and carefully thought-out features all conspire to make Twitter much more fun to use than it would otherwise be.

Twitter for Mac does not use quite the same design, but brings the same design sensibility, and is equally fantastic. When I started to use Twitter, I used Twitterific, but while I liked it, it niggled at me in certain ways. I then switched between a variety of Twitter clients, most of which I stopped using fairly quickly because they also didn't sit right with me in one way or another.

Twitter is such a simple and well-defined thing that you'd think it would be easy to design a good Twitter client, but it really isn't. In my opinion, a good Twitter app needs to be there instantly when you want to read tweets or post something, but stay out of your way the rest of the time. I'm with John Gruber: I think Tweetie does that wonderfully. I thought I would miss Growl notifications or beeps when new tweets arrived, but I've found that I don't. The menu bar item turns blue when something new is waiting (you can tailor what triggers the notification), which is just enough to be a notification, without distracting you. The hotkey to trigger the main window is great for quickly summoning or dismissing the window to read tweets, and I love having a separate window and hotkey for writing them. In short, I really like Tweetie, and I'm happy to have the 'Twitter Experience' I love (I can't quite believe I've just typed that) on my Mac as well as my iPhone.

iPhone version of Papers

I'm a very keen user of Papers, the Mac software for collecting and organising a collection of journal articles and their associated PDF files. It never fails to impress colleagues when I pull it up and do a quick search to find some paper we've just been discussing. Now I won't even need to be in front of my computer to impress them, because Mekentosj have just released a version of Papers for the iPhone.

I've been playing with it for a few days now, and it's very well done. You can choose what you sync with your desktop version of Papers, so if you've got a huge collection, you don't have to fill your iPhone with it. I have a Smart Collection for my 'to be read' articles (which seems to grow by the day), so I've been syncing that. However, I might add a Smart Collection of my own papers too -- more about that later.

Most of the features available on the desktop version are available on the iPhone -- certainly all of those that make any sense on the platform are present. You can even use the search engines to find new articles and download the papers to read, which could be very handy. The PDF viewer works quite well, though there seems to be a little bug in the PDF viewer for third parties that means that some files are rendered with slightly blurry text, but I'm sure this will be fixed in time. I haven't found it a big problem with most files, and the convenience of being able to reduce my huge 'to be read' virtual pile in spare moments away from my computer overwhelms any minor issues with the display. You can also make notes on the paper, which are then synced back to the desktop.

I mentioned that I might keep a collection of my own papers on the iPhone. Why should I want to do that? Well, when you are viewing an article entry, you can choose to share the paper: either by email, or by sending it directly to nearby iPhone Papers users. I haven't had a chance to try out the latter, but I think that would be incredibly cool. Still, just being able to chat to someone at a conference or meeting about your work, then pull out your iPhone and send them a copy of the paper you're discussing, there and then, would be genuinely useful.

BubbleTimer

Although I started tracking my time using a home-brewed, Tinderbox-based solution, I still look out for new and 'frictionless' ways to record the time I spend at work on various projects. One project that I came across recently has really impressed me: BubbleTimer.

Based on David Seah's Emergent Task Timer PDFs, BubbleTimer offers a very simple, but very powerful interface. As you can see above, you set up tasks in the left-hand column, then click the bubbles corresponding to the 15 minute time slots that you spent on that task throughout the day. At the end of each row, you get a total for the time spent on each task, and at the bottom of the page, the total for the day, with a pie chart of the percentage of time on each task in a popup.

You can also set goals (more than or less than x hours on task y), and the Daily Goal line is coloured red or green depending on whether you have reached your goal or not. You can also print the daily summary and even export ranges of data to CSV format for further analysis in your spreadsheet or statistics package of choice.

It's a very impressive piece of work. If you set up broad categories of tasks at the start, you don't need to keep entering a description of the task, as the tasks carry forward to subsequent days. So it's very quick just to click a few buttons in an open browser window as you work. You can track tasks as you do them, or fill in the bubbles retrospectively (after time away from your computer, for example). Visually, it's easy to see when you've been having a focussed day (long strings of bubbles) and when your day has been fragmented by switching between lots of different tasks.

There are also some lovely touches: the 'pencil scribbles' that fill in bubbles when you click them are not identical, so that it does actually look like a hand-entered sheet. If you hover over the total for a row, you get a little popup showing sparkline graphs for that activity for the previous 7 days and the current month.

There's a 14 day free trial, and it costs a very reasonable $20 per year to use. If you want an easy way to see where all your time goes at work, or you have a New Years resolution to keep (more than 30 minutes exercise per day or 4 hours per week learning Italian), it's an excellent tool to help you keep on course.

Fastmail

You might wonder -- in this age of free email accounts with gigabytes of storage -- why anyone would pay money for an email account. Well, I've just done exactly that.

For a while, I've been consolidating numerous email accounts into one account using Gmail, so I can receive email at my usual range of email addresses, but check, search and send my email within one account. It worked fairly well, but Gmail's quirky implementation of the IMAP protocol has its own irritations, and I experienced a few reliability problems. The reliability issue is especially difficult when you're forwarding all your email to an account. Also, while Gmail allows you to send email from your Gmail account as if it comes from another account that you control, it sets one of the headers (I can't remember which one offhand) as "From [my own address] on behalf of [gmail address]", which isn't what I wanted at all. That meant I had to set up send-only accounts in Apple's Mail, which was an unwanted complication.

So I began looking around for another email provider, and came across Fastmail.fm. It had some very good reviews from happy users and seemed to have all the features I needed and more, so I bit the bullet and paid for an account.

I'm really happy with Fastmail. It certainly lives up to its name, and while some of the features can be replicated (with some effort) using Gmail, the level of polish and sophistication is wonderful. It's ideal for consolidating accounts, because you can set up "Personalities" which specify not just which "From" address to use, but which signature to use, an address to "BCC" to and a sent folder to store the mail in, among other things. So you have a great deal of control of how you send mail. The magic thing is that when you send email from a desktop client with a particular from address, it sets up the rest of the rules for that Personality automatically.

The spam handling is also very sophisticated (with much more control than Gmail gives you), and if you want them, there are flexible rules for filtering incoming mail. I mostly use my desktop email client (Apple's Mail), but the web interface is also quite powerful and very fast. It doesn't have all the visual polish of Gmail, but it's very functional.

I'm not abandoning Gmail entirely: I still use my incoming Gmail address (forwarded to Fastmail automatically, of course), and then I automatically BCC all my incoming and outgoing mail to another secret Gmail account for backup and emergency access if Fastmail should ever go down. I may have to pay for Fastmail, but if feels as if it's worth every penny. My one and only minor quibble is that it only checks external POP accounts every hour at a maximum. That's not a problem for most of my accounts (which I just forward directly to the account), but I have frustratingly little control over my work email account and can only pull it into Fastmail using the POP checker. That's not really Fastmail's fault though, as it would be unreasonable to check POP mail much more frequently than that. In practice, if I'm expecting some urgent work email, I can log into the web interface and trigger the checking manually. If there was a way to do that with a script, I'd be a very content bunny indeed.

Dropbox

Like many people who work on more than one computer at more than one location, I've had a perennial problem with making sure that all the files I need are on all the computers I use. My first attempt used a home-brewed set of rsync scripts to sync files up and down from Joyent's Strongspace file server. For various reasons, I then switched to using a portable hard drive with ChronoSync. I used that successfully for quite a while, but it meant that I had to carry a fragile hard drive around, and it took a while to sync things up and down at the beginning and end of each day.

Recently, there have been a rash of services which sync your data between computers, using an online store as an intermediary. I had a trial of SugarSync and liked it a lot, but when I tried Dropbox, I was smitten. Ironically, it has fewer features than SugarSync (though many of the missing features are in development at the moment), but there's something about the transparency of the way that it works that appealed to me. I liked it even better when -- after signing up for a beta with 2 GB of free space, I was entered into a competition to win 50 GB of space for a year and won! 50 GB is more than enough to hold everything I need to sync in my home folder, and gave me the opportunity to really try it out properly.

The way it works is very simple: you install the Dropbox client, which just shows itself as a menu bar item in Mac OS X. This creates a "Dropbox" folder in your home folder, though you can relocate it if you want. Everything you put into the Dropbox is automatically synced to all other computers running Dropbox. If you put stuff in ~/Dropbox/Public it's accessible to anyone to whom you give the URL, and a quick right-click on the file copies that URL to the clipboard.

At the moment, you can't choose to sync existing folders outside your Dropbox, but a clever trick with symlinks allows you to make it work. In the instructions, you're told to create a symlink in the original location to the original file in the Dropbox, but I do it the other way around (so that my files remain in their original locations), and it works perfectly. So in my Dropbox, I have a collection of symlinks to other folders, which means that everything I want to sync between computers (all of Documents, Music, Movies, Pictures, a few folders in Library and a few in Library/Application Support) get synced up.

It keeps revisions of files, so if I want to go back to a previous revision, I can, and it creates a copy if it can't reconcile changes made concurrently. In practice, I've never had a problem, and it has all worked transparently. Because it's constantly syncing changes, it takes very little time to sync the latest changes at the end of the day, so it's much faster than my previous methods. Likewise, if I lose the network connection for a while, it's not a serious problem, because syncing will just catch up as soon as I get back online. Also, if Dropbox went away tomorrow, all my files would be exactly where they've always been: on my hard drive. I'd just lose the syncing part and have to go back to one of my previous methods. I've also found the public box very useful when collaborating on documents with colleagues, rather than emailing attachments.

There are some interesting new features planned, the most useful of which will be the ability to specify folders to sync, without having to use symlinks, but it works so well right now, that I'm very happy with it as it is.

iPeng

I really like Apple's Remote application on the iPhone, which lets you use the phone like a remote control for iTunes, complete with library browsing and display of cover art. However, in the living room, I tend to listen to my music collection using my SliMP3 player, via SlimServer running on the iMac upstairs. So I was delighted to find a plugin for SlimServer, called iPeng which does the same job as Remote for my SliMP3.

It's not actually a native iPhone application: you access it via MobileSafari using a special URL, and it shows you a very nice interface for your library, with full control over playlists and so forth. However, it's easy to put an icon for the URL on the main screen, so in practice, you hardly notice that it's not a native app.

I do have a standard remote control for the SliMP3, but it's much easier searching and browsing my music collection from the iPhone, rather than fiddling about with the limited buttons on the physical remote. I love the way I keep finding additional uses for my iPhone -- it's certainly the most versatile gadget I own. Somehow, it manages to be a Jack and Master of All Trades.

Using Tinderbox as a task timer

I mentioned a little while back that I'm using Tinderbox as a kind of daybook to record thoughts, ideas, notes and activities throughout the day. I'm really enjoying using it, and the experience of gradually adapting the structure of my Tinderbox document has been interesting and fun. I also posted a request a while ago for suggestions for simple software that would allow me to log my time at work. People responded with some great suggestions, and I've since found a few more. For example, I love RedBook: it's a commandline task logger written in Ruby, which has a simple method of storing the data (as YAML-format files) and can export and display the data in a number of formats.

However, in the end -- good though they were -- these applications felt like adding another layer of complexity on to something that I wanted to be transparent and effortless. Then it hit me that I was already jotting down what I was doing in Tinderbox, so if I could store start and stop times as attributes of those notes, I'd be half way to logging my time. What follows is probably only interesting for fellow Tinderbox nerds or serious geeks. You have been warned

So I created a new prototype called 'timecard' which has key attributes of startTime, stopTime, intDuration (calculated in seconds from the start and stop times) and duration (which is nicely formatted in text as hh:mm:ss). I create a note as usual, using the title of the note to describe the activity, then drag it to an adornment, which has some OnAdd rules to set the tag attribute to 'timed', and set startTime to the time now. When I'm finished with that activity, I drag the note to another adornment which sets the end time to now and calculates the duration in seconds by passing the start and end time to a little Ruby script I wrote, helped a lot by a thread on the Tinderbox forums. The notes can sit on that adornment until I've got time to take stock, or the end of the day, whichever comes sooner! Then I drag them to another adornment, which just shoots them back into a Timecards container. I've set the adornments up to be a kind of electronic version of one of those time card punch machines you see in factories, and that's the way I think of it.

I've also set up a variety of ways to view the information that this system collects. In the first screenshot, you can see that I've set up a DisplayExpression for the agent which collects today's timecards to show me the total amount of time I've worked so far today. In the second, I've used the spiffy new bargraph() function to set the Pattern attribute of an agent collecting the past week's worth of timecards to show the total time for each day in a kind of sparkline graph. When I've accumulated enough data, I'll probably do the same for the past few months as well. Finally, I've created an HTML export template for the timecards, which formats the task descriptions and times for an agent (usually today's times) as a pretty HTML table, so I have a summary record for each day.

Since I don't have to collect precise hours for invoicing or anything like that, the absolute values are not particularly important. I only time tasks when I'm properly focussed, not when I'm being interrupted by calls, conversations with colleagues and so on, so it's quite a good relative measure of how much solid, productive work I'm able to accomplish. It has been really interesting to track it from day to day. I'm actually using it as a sort of game to motivate myself to devote uninterrupted time to important tasks, just so that I can watch the bars climb for each day!

Mobile Safari

Reading web pages on the iPhone is a very different experience than reading them the desktop version of Safari. Some things (like entering text and so on) are inevitably slightly more awkward because of the onscreen keyboard. But there's one thing in particular that I'd like to see in the full version of Safari: zooming in on page elements. For those who haven't used Mobile Safari, each page is initially displayed full width, which for most pages results in tiny text, but means that you can see at a glance the layout of content on the page. The magic comes when you double tap a page element like a column of text: it expands to fill the entire width of the page, which for most pages I've tried makes it perfectly readable. It also -- and this is why I want it in desktop Safari -- removes any of the other distractions on the page. This is perfect when you're reading typical sites with sidebars and other content at the side of the page, and is particularly good with online news sites where you have flashing adverts and other distracting elements to either side of the content.

It's the web browser equivalent of 'full screen mode' on a text editor, and I'm rather addicted to it. So much so that I tried double clicking a column in desktop Safari earlier today to concentrate on what I was reading.

Tinderbox daybook

When I started using Tinderbox again for planning various work projects, I noticed that a lot of people were using it for a simple daybook or journal. I've tinkered with various ways of keeping a record of the various things I do, people I talk to or ideas I have throughout the day, including a simple little plugin I wrote for Textmate to keep a journal in a plain text file. That worked quite well, but it wasn't as easy as it might have been to find things again when I needed them because it was one big flat file.

So I started playing around with keeping my journal in Tinderbox, and I've been using it for a couple of months now. It's deliberately very simple: I have a container called Daybook, in which all my snippets of text are kept. That container has an OnAdd action which sets the prototype as daybook (setting the colour and a few other attributes), and also sets the title with a datetime stamp of the creation date. That means I can just hit return to create a new note, hit return again to dismiss the dialogue setting the title (because it will be set automatically), then start typing in the window which appears. These daybook entries are sorted in reverse chronological order, and automatically collected by 'Today' and 'Past 7 days' agents, which do just what you'd expect. I also have another container for completed tasks, where I have one note for each day (again, auto-titled on creation with the date and 'tasks') which contains all the completed tasks for that day.

One of the things I really like about Tinderbox is that the DIY ethos of it means that you can make something as simple or complex as you like, and -- even more importantly -- you don't have to decide exactly how something should be set up from the start. Once I'd been using the setup I described above for a few days, I realised that it would be nice to collect my notes on articles I'd read in a separate place so that I could find them more easily. The infinitely flexible structure of notes meant that I didn't have to create a different kind of note to do this, or even go back and edit my previous notes on reading. When I make notes on a paper I've read, I tend to first paste in the reference and the link to the entry from the Papers application, so that I can find the original article easily from my notes. So all I had to do was create another agent called 'Reading' which searched for notes with the string 'papers://' in them, which is the start of the Papers URI format. It would then assign 'reading' to the attribute 'tags' for that note. Also, by setting the tag manually to 'reading' I could get the Reading agent to find the note.

I set up something very similar for bookmarks, so that if I dragged a URL on to a note to remind myself of some online resource, it would be collected by the 'Bookmarks' agent. It's important to note that these agents just store aliases of the original notes, so that all the originals are either in the Daybook container or the Completed tasks container. At any point, I could set up another agent or alter the existing ones, and view my notes in a different way.

Tinderbox has great text and HTML export capabilities, so I can export my journal for the day, week, month or whatever period I want, and it's easy to view in other forms or reformat for other uses. And if I ever want to use something else, the file itself is XML, so I could still get my data out.

blippr

I've just signed up to a new service (currently in beta, natch) called blippr. You've probably already guessed from the missing 'e' that blippr is a social web service. It offers 'Radically Short Ratings and Reviews' for books, films, games and music. The idea is that you 'blip' items, rating them on a four-point scale and writing a short review if you like. It's a little like Twitter in that you only get 160 characters for your reviews, which is both a good and bad thing. If I really like something, I want more space than that to write about it, but on the other hand, I often put off writing about stuff I listened to, watched or read in my media section, because it will take too long to write a full review.

The real benefit is that -- when a few more people have signed up -- it should provide a great way for people with tastes you admire to recommend things you would probably enjoy. I've also started using it as a wishlist to note down things I'd like to get see, read or listen to. I tend to read reviews and get interested in something, then forget to note it down anywhere.

I've only just started using it, but it's interesting so far. My profile is here, and if you'd like to try blippr for yourself, I have 3 invitations to give away.

Page 1 of 3 pages  1 2 3 >