So what do I mean about opening files in Notational Velocity? It's right there in the open with menu, it obviously already does this. Well, no, not quite. The trick is that, if you use Notational Velocity to open a file, it's essentially an import. In other words, the file gets duplicated into NV. That's all very well, usually. The trick is that, if you've done a search in, say, spotlight, and you've found an NV note in the result, You can't open it in NV. Or rather you can, only it'll duplicate your note so that you have two copies of it. So you either have to compromise and open it in your default editor, or exit spotlight, go to NV and search for the note title.

Neither of these scenarios are exactly ideal. Enter our original search in NV script. With a couple of modifications and an application called Platypus, you can write your own open in Notational Velocity application.

For convenience sake, I'm going to attach my version of the application I'm talking about to this post, but I have no idea whether it will work for you. I don't have another Mac to try it on, so your milage will definitely vary. If it *doesn't* work for you, or if you just want to know how I did it, here're the basics.

The first thing to do is make this work from the command line. The script we wrote in the first part of this series will work, with a couple of modifications. It now looks like

#!/bin/bash
string=`basename -s ".txt" "$*"`
osascript <<-Script
tell application "Notational Velocity"
activate
search "$string"
end tell
Script

The change here is that, once you give it the search string, it uses the basename utility to remove the ".txt" extension from it.

Now obviously, if you're using html or rtf files in NV, you'll have to change this to reflect that. Save this version, and test it. You should still be able to do what we did before, search for a note from the command line. Since you're note title is unlikely to end in ".txt", it should still work. Now go into your NV notes directory and enter a existing filename, for example "test note.txt". Make sure it's a note that exists. What should happen is that NV jumps to the front with the note selected.

Now again, the same rules apply here. Particularly if your note title is short, you may not get the note title you're looking for, but it should be in the list of notes that appear.

Once you've gotten this to work, it's time to write something the finder can use. The trick here is that you can't just use this script. Finder wants a full fledged Mac Application. Since I'm no objective C coder, I don't really know how to do this. After a bit of research, the easiest way I found was to use a program called Platypus. This application takes a script in just about any language and wraps it in an application. So go ahead and download it. When you open it, you'll get a screen that lets you configure what you want your app to do. First, give it a name and the path to the script you want to wrap. In my case it was nv and ~/bin/nv. You can probably leave the output set to the progress bar if you like, although I set it to nothing. Click on "advanced options" and a whole lot of other stuff will appear. The first thing we need is "is droppable". Check that, so that it can accept files. Once you do, you'll be able to select what types of files it can accept. If you're saving your notes as text, this is obviously "txt", but change this if you're using html or RTF.

The only other thing to be sure of is that "remain running after completion" isn't checked. Now click on create, tell it where to save it, and you should have a working application.
Test it out by going into your NV directory in finder and control clicking a file. Choose "open with" and select your new application. If it isn't in the list, choose other and find it. You should be in NV with your note open.

Now obviously, this isn't the most sophisticated script in the world. It doesn't check that you're in the NV directory before it runs, so if you're not careful, you'll end up with an empty note with whatever your filename is as the title, probably not what you want. Admittedly, I could probably fix the script to do that, maybe importing the note if you're not in the NV directory and searching for it if you are, although I haven't needed to do that yet, but if you're careful with it, it can be a useful part of your workflow.

In the final part of this series, I'll show you how I created a way to do the same thing with a URL, making it possible to have basic note linking capabilities in NV, and to refer to NV notes from other applications.

Click here to download:

nv.zip (58 KB)

So, to the details. What's made all this possible is a recent update to the NV code base, search for a note with Applescript. Notice, I've said search. Looking at this feature, and thinking about it a little, makes me realise why requests for features like opening notes in NV and opening URLs in Nv aren't as straight forward as they seem. I'm willing to bet that there's no such thing as an "open" command. The closest we can get is "search" and feed it the note name.

Even doing this, depending on your note title, you might not get the exact note you're looking for. For example, say you had a note called test. But you also had a note titled "testing programs". Depending on how your Database is sorted, searching for test might bring up the "testing programs" note first, not test. Now obviously, you can go through the search results and access that note easily, but it might well be unexpected if you did this through "open with" or clicked on a URL. Just be sure to keep this in mind if you're planning on setting any of this up.

First, be sure to go tohttp://github.com/scrod/nv/
And grab the latest source. Either download it, or if you have git installed, clone it.

Once you've done this, change into the source folder and open the Notation.xcodeproj file. This should open in xCode. build the project, and once it's finished, quit xCode again. You'll find your newly compiled program in build/Development inside the source folder. Make sure you quit any running versions of NV. Back up your notes directory, and try to run it. If you have no problems, back up your old version of NV somewhere, and move the newly compiled one into your applications directory.

Yes, this is a bit risky, I'm the first one to admit that. But provided you back up your notes and your original version of NV, nothing you can't get yourself out of. And I, personally, have been running a development build of NV for quite a while now, with no problems at all.

Now for the script. Call it whatever you like, I called mine nv, and put it somewhere on your path.

 
#!/bin/bash 
osascript <<-Script 
tell application "Notational Velocity" 
activate 
search "$@" 
end tell 
Script 

Now try it out. Typingnv test note
should put you in Notational Velocity, with "test note" in the search field.

If you haven't got a note called that, it'll be blank, obviously, and you can hit enter to create it. Try searching for a note you have that does exist, as I said before, make your search terms pretty specific while you're experimenting.

So what are we doing here? This is essentially an Applescript with a bash script wrapped around it. All the bash script does is take everything after "nv" on the command line. The rest is using osascript - a way to run Applescript from the shell, and passing it to Notational Velocity's search command.

This post might well be all you need. I'm using this script in other scripts and services myself, and it's cleaned them up a lot. If you're a Quicksilver or Launchbar user, this will also let you do a NV search or even create new NV notes from there. If you have trouble getting this to work, drop me a note in the comments, and I'll try to help.

But if you want to go further, stay tuned.

In my quest to see what other people are doing with Notational Velocity, I came across this post on dougist.com Notational Velocity - Show in Finder

He's doing what I am, storing his NV data in separate text files and needed to be able to jump to them in finder. You should read the post, but basically he's using spotlight for this. I tried it a few times and for me, it's not all that nice. I don't really like Spotlights finder interface, since it's hard to customise and, at least in my case, gave me a lot of other things - web pages, emails etc - that I didn't want to see.

So I decided to break out automator again and write myself a service. What I wanted to do was jump to finder with the file selected. With a bit of judicious applescripting, it's not that hard,. What I came up with is a hack and kind of nasty. Anyone with more scripting skills than I have can feel free to improve it, but it pretty much works. If you want to try it yourself, here's how.

Create a new automator service. Set it to only work in NV, but otherwise leave it at the defaults. Drag in the run applescript action and paste in this script.

on run {input}

tell application "Finder"

activate

reveal "volume:Users:username:Documents:notational data:" & input & ".txt"

end tell

end run

Now obviously, you need to change the location NV's looking in for the folder with the one you're using, and it needs to be an oldstyle Mac path, not a new style unix one. Replace volume with the name of your boot volume, username with your username and the rest with wherever NV stores all its files. Also, if you're using rtf or html, change txt to the right extension.

Give it a shortcut, highlight a note title, hit the key and bingo. You should be in finder with your note file highlighted.

Now you *might* get errors on certain notes doing this. The usual cause is that the note title contains text that OSX doesn't understand in a file name. The only solution to this seems to be to rename the note without whichever character it's unhappy about. The ones I had trouble with included slashes, but there might be others. 

It's not a bullet proof solution, of course, but it does go to prove that with scripting and services, there are a lot of things you can do to add your own features to an application.

So we all need a good notes app, right? that place where we throw random musings, articles we come across on the net, that phone number someone's just read you over the phone that you just have to write down immediately, random shopping list items and so on.

The Mac seems to have a huge number of these things, and trying to find the one that suits your brain and, heaven help us in the case of Voiceover, is accessible ends up being a case of trial and error. Up until the last week or so I've been a die hard Journler user. It has enough features to choke a horse, a great Applescript dictionary which is useful if you're a tinkerer like me and was pretty reliable.

But all things must come to an end. Given the lack of development on Journler lately, I've been keeping a lookout for an alternative. And I think I've found it - Notational Velocity.

At first, looking at the features, it doesn't even look like it's in the same class. The feature set is tiny, but that's actually a plus. Finding things in Journler always felt like a bit of a chore, in notational velocity, it's the main point of the app. Journler always felt a bit sluggish - NV feels snappy. Journler was cluttered - NV has nothing but a search bar, a notes list and the note itself. Not to mention, NV lets you save everything as text, Journler used RTF files. Make a note of that text files thing. It'll be important later on.

So I exported my Journler notes as plain text, threw them into a directory and bingo, all my stuff is now in NV. But what about the features? I like minimalism, but some of them were nice.

The first problem I ran into was, well, NV's import selected text service is just a bit primitive. The applescript I'd cobbled together for Journler, at least it Safari and Mail, was pretty nice. Grab the subject or page title as the note title, put the rest in as text.

NV, though, never seemed to get the note titles right. Sounds petty, but it annoyed me. What to do? Easy. Use the applescript I'd already cobbled together for Journler. My skills with applescript are a little primitive, I have to admit, but some judicious searching and I found out how to write out text files using applescript. All I needed to do was copy my existing script in case I ever wanted to go back and, in all the spots where I was writing to Journler, write to a text file in NV's notes directory using the note title I wanted as the file name.

It worked like a charm. NV found the new note I'd thrown into its directory instantly, or at least as close to instantly as made no difference. It's even a slightly better system in some ways, since even if for some reason I'm not running NV, it works independently without needing to start it. 

This is only one example of how this could be useful, though. As a text file nut, I've got a lot of scripts that write to and create text files. Nv doesn't care if you throw new files into it, or even if you edit a file it's already got, which adds a lot to its flexibility.

That's all very well, but what about features in NV itself? I'm actually typing this blog post inside it, and one of the things I'd like to know is how many words I've typed. Journler used to have this in a prominent place, NV doesn't. There is, though, a wc command in unix. I could select my text, copy it to the clipboard, switch into terminal and type something like

pbpaste|wc -w

This basically means, take the contents of the clipboard and pass it to wc, the word counter app, which then tells me how many words there are.

Now there are a lot of ways to execute shell scripts with keystrokes in OS X. The one I've used, though, is a service in automator. Create a new service that takes selected text, drag the run shell script action into it. By default the selected text is passed into the script so no clipboard mucking around. My script looks like

wc -w|say

Assign it a keystroke in keyboard preferences and my workflow is down to select all wordcount.

And yes, those of you watching closely will notice I've piped it into say. I'm a Voiceover user, having the word count spoken to me is fine. If you need it printed to the screen, you might have to use an applescript with an alert dialog or something along those lines. this is more of a grab bag of ideas than a step by step, if you'd like me to write more details about how you might do this in another post, drop me a note in the comments and I'll see what I can do.

The next feature I wanted was Journlers email entry command. What I've come up with for this isn't quite as good as that, but without applescript in NV, it's the best that can be done. Just assign a keyboard shortcut to  the email selected text service. Select all, hit the hotkey and you're in mail.

The disadvantage is that, obviously, this service doesn't put the note title in the subject the way Journler used to. but other than that, it's pretty good, and I certainly wouldn't give up a notes app for a feature like that.

The last feature I'm going to talk about are file attachments. I never used this a lot, but it was occasionally nice to have a movie or a sound file attached to a journal entry.

the solution to this, again, isn't perfect, but given how I think I'll use the feature, is probably good enough. It's based on the fact that NV recognises all the URLs that OSX can throw at it. This includes links to files, emails etc as well as web pages. There are a lot of apps that take advantage of this, including, ironically enough, Journler. For all of those, copy and paste is your friend. If I want to refer to a file in NV, all I need to do is copy it and paste it in NV. I'll get a file link that jumps to it in finder. The same applies to emails, web pages, and any other application that uses URLs.

Admittedly, if I use NVs synch capabilities to look at notes elsewhere, these links won't work. At the moment, though, this isn't vital. If it *becomes* vital, I can probably move the files somewhere remote and make them accessible, but I'll cross that bridge when I get to it.

As I said before, this isn't very detailed, and given that I've only been using Notational Velocity for a little while, I'm bound to change how I'm working with it a bit before I settle down. I'm still working on how best to use tags or, for that matter *whether* to use tags, given the bare bones support for them at the moment. I've just started to realise how useful it is for keeping running logs, such as programmers logs, given that each note has its own undo history and cursor position, and how quick they are to jump to. but that's probably another post in another few weeks, once I've worked with the system for a while.

I hope, though, that this has given you some ideas for your own scripts and hacks. With a bit of imagination, taking advantage of OSX's built in features, and a bit of scripting knowledge, you can make what's already a great app even more useful.

Welcome to what I hope will be the first of several articles on Quicksilver, a
program I, personally, have found incredibly useful for doing things in OS X.

I won't be covering a lot of features in this first article. Quicksilver is a
complicated beast and explaining everything it does would take a manual which
has been written and which I've linked to at the end of this post. What I'll
cover here is a basic introduction to what it is and what it can do, and
recommended configuration options, both for VO and Quicksilver. If you're trying to do anything in my other
articles, and you're having trouble getting them to work, check back here and
make sure you've set everything up properly.

So, what is Quicksilver?

This isn't as easy a question to answer as you think. At the most basic level,
it's an application launcher. Press the hotkey, usually control-space, type
all or part of an application, hit enter. Instant launch. The same applies to
files and urls, only they get launched in their default application.

But so what? In leopard, at least, Spotlight can do a fine job of this, and to
be honest, if that's all you need, stick with it. But there's a lot more you
can do with Quicksilver.

Even as an application launcher, it has other features. The most interesting
of those is, it learns from you. I can launch most of my applications with a
single keystroke. m is mail, s is safari and so on. But what happens with
something like i? The first time you type it, it's likely to give you ichat.
But I, for one, don't use it very much and don't mind having to type a few
characters to get it. What I want i to give me is itunes. So type i, arrow
down until it says itunes and press enter.

The next time you type i, it will still be ichat, and you'll still have to
arrow down. But the *third* time, it won't. It'll be itunes, since that's what
you've told it you want.
The other thing that makes it different from straight Spotlight, is that you
can, once you've found something, *do* something to that thing, and that's
where the magic comes in. Find a document in your documents folder, press tab,
and you've got a bunch of actions you can perform on that document. Email it
to a friend, copy or move it to somewhere else, open it with another
application and a lot more. Still a bit unsure? I've linked a few more
introduction posts at the end, so have a look through those.

But isn't Quicksilver dead?

For those of you who've been around a bit longer, you might wonder whether
Quicksilver's still worth it. The original author has pretty much stopped
updating it, and even though he's released the source, no-one's really managed
to release a workable new version in a while.

For me, the answer's still yes. For you, it might not be. If one of the
quirks/bugs it still has make it unusable for you, if you're already using
applications that do a lot of what Quicksilver does or if you just plain avoid
unmaintained software you might want to avoid it. But I still think you should
give it a try. I feel absolutely lost on a computer without this app, and I
still think it's one of the best swiss army knives on the Mac.

A quick Snow Leopard update is probably in order here. There were quite a lot of problems with Quicksilver and Snow Leopard, and to be honest, I'm currently not using it and took the opportunity to play with OS X's new services system, which I'll talk about in another article, I think.

On the other hand, the fact that it *is* a bit flakey in Snow Leopard seems to have prompted some more development, and apparently the most recent version's looking pretty good and it looks like there's more to come. QS looks like it might get a lot more development than it has in a while, so things are looking up.

Also on this subject, if anything I talk about here *doesn't* work in the latest version of Quicksilver, Have a look through the google group to see if it's a known problem, and if it isn't, drop the Quicksilver folk on there a note.

Installation

So, you've decided to give it a look. If you're still running Leopard or Tiger, I still think you should try to get a copy of Quicksilver b54 from blacktree.com, although I'm not sure if you can still download it from there. If you can't, or if you're running Snow Leopard, go to
http://github.com/tiennou/blacktree-alchemy/downloads and get the latest version from there.

Once you've downloaded it, open the disc image and launch it. It will tell you
you're running it in a disc image, and ask you to choosesomewhere to copy it
to. Once you've done that, click continue. Have a look at the license
agreement and check the accept checkbox. It'll tell you that it's a beta, and
that it'll be deactivated in six months, but ignore it. I've been using it far
longer than that, and it hasn't been deactivated.

You'll then be taken to a page of recommended modules. Interact with the table
and check most if not all of these. The only exception might be the
development module. Unless you're a hardcore OS X programmer, selecting this
will add a lot of unnecessary programmer documentation to your search
results, which you probably don't need.

Once you continue from this screen, the modules will be downloaded from the
net, so it might take a while. If you think this is taking a ridiculously long
time, or it seems to freeze, just cancel out of it. You can install all these
modules later.

On the next screen, you can change what hotkey you use to activate QS, and how
often the catalog is updated. Feel free to change the hotkey, if you like,
although future articles assume you're using control space. As for the
catalog, you might need to increase the update time. Leave it for now, but if
you find that your computer freezes every ten minutes, I'd increase it, but
remembering that Quicksilver won't find anything new you put on your hard
drive until the update's been run, or until you do it manually yourself.

After this, there's a "getting started" screen and a support screen and you're
done. Well, not quite.

All this still applies to b57, except that you'll probably have download a tar.gz file. Just uncompress it, copy the quicksilver application to where you want it, and run it. Also, the plugin download thing's a lot more obvious in this version, and you should pretty much here it tell you each plugin as it downloads it.

If you didn't manage to install the plugins during the install, go into the
plugins section and choose "recommended". Click the checkbox next to each one
and install them. As soon as you check the box, it will try to install the
plugin, so check the checkbox and wait. Voiceover will say "quicksilver has
new window" when the install's done, even though you can't see it. Go down to
the next one and check it until you've installed all the ones you want.

While you're here, you might want to have a look at what other plugins are available. If it looks like there are only a handful, make sure you click on the "reload plugin list", button, which is the second one to the right of the plugins table. If you're looking at a plugin, and you want more information about it, there's an info button to the right of the search box. Activate that, and a drawer will appear next to the plugins table. If there's any information about the plugin you're looking at, it will appear there.
Now for the preferences.

This is a slightly non-standard pane, since it uses a table to select
sections, a bit like the VO preferences.

Applications
The only one I think is essential here is advanced options. A lot of the later
stuff I talk about won't work without this. The other thing is, if you don't
want QS to appear when you're cmd-tabbing between applications, uncheck "show
icon in dock."

appearance

As a VoiceOver user, you can probably turn almost all the options here off. Superfluous visual effects, load icons and load previews. The only other thing to check is that the command interface is set to primer. Make sure you do this, since it's the only way Quicksilver will be accessible with VoiceOver.

command

You can change your activation hotkey here if you didn't and need to.
There's a few things to play with here and they're mostly a matter of
preference. I'd leave them for now and change them once you've been using it
for a while. I would recommend, though, changing spacebar behaviour to "jump
to argument field" since it'll make your life easier later.

extras

This, again is a bit nonstandard. you'll have to VO-right past a few
random controls, and the preferences for this section are in another table
with checkboxes. The ones I'd recommend here are "capitalized keys modify
action in command window" and "pull selection from front application instead
of finder."

The rest . . .

Going back to the preference table here, there's a blank entry. I'm not
actually sure why, but ignore it.

From here on, you probably won't need to change anything. The next preference
option is actions, and here you can check and uncheck actions that quicksilver
gives you, and decide on the order they're going to appear. Have a look at
this, if you like, since it'll give you an idea of what Quicksilver can do for
you, but you probably won't need to go here a lot, unless an action you think
should appear doesn't, or unless there are actions you never want to see.

The rest of the preference panes depend on what plugins you've installed,
since they're used to modify settings for those plugins.

The Catalog

The final section to look at is the catalog. This is where you tell Quicksilver what to index on your computer. This is another good one to have a look through, particularly if you add a new plugin, since different plugins add different things here. For example, the Safari plugin let's you choose to index things like bookmarks and history. Adding to this catalog can be a little tricky with VoiceOver, and will probably get at least a quick tip of its own, but what we're here for right now is to check some checkboxes.

In the first table, select quicksilver as the catalog you want to change. Stop interacting with that table, and go to the next one, where you'll see entries such as "proxy objects" and "internal commands". On the right of these, there are check boxes. For now, check them all. You're almost sure to need some of these in your more advanced use of Quicksilver, and those that you don't use will just end up at the bottom of your search results anyway.

And after that, you're pretty much done with Quicksilver itself.
If you're using Leopard

If you're a Leopard user, there are a couple of VoiceOver settings you might want to have a look at. press VO-f8
to bring up the voiceover Utility menu. go to verbosity, and select
announcements. Make sure that "speak when status text changes" is checked. This
will make using Quicksilver a lot nicer, trust me.

The other one you might want to check, although it's not quite as important, is "allow cursor wrapping" under navigation. This makes navigating in Quicksilver itself a bit faster.

conclusion

So that's the basic Quicksilver setup guide. I know, I know, I haven't given you much, if any information on how to use it yet, and when you first activate it, you'll be pretty confused. But don't panic. This article is already quite long, so I've split it into two parts. So feel free to go on to part two, where you'll get a basic crash course on how to make sense of Quicksilver in VoiceOver, and some tasks you can use it for.

links:

Quicksilver User Manual warning. This is an 8MB pdf file.
http://vjarmy.com/archives/2004/03/quicksilver_a_b.php quicksilver: a better OSX in just ten minutes: ancient, but still a good overview.
http://www.43folders.com/topics/quicksilver All the Quicksilver posts on 43folders. A lot of useful tips here.

So, you're a Voiceover user on the endless hunt for a good text editor. You might be a programmer like me, or someone who does desktop publishing in Latex or you just want to write text and don't need all the bells and whistles of a word processor. But there's a problem. Most of the text editors everyone raves about in OS X are completely unusable. For whatever reason, the built in text editing interface in OS X just doesn't seem to want to do what text editor authors want it to do, and so heavy hitters like BBedit and Textmate and even aquamacs, the GUI version of Emacs, just don't work for us.

There are alternatives. Smultron, SubEthaEdit even textedit, but for me at least, they're not quite good enough. They either don't have enough programming features, or the ones they do have take a lot of scripting. And so, we come to MacVim.

As the name suggests, MacVim is a cocoa port of an existing editor, Vim, which has a pretty long history in the unix world and, in fact, is built into OS X without you downloading anything as a terminal application. MacVim is a lot better than this, since, being a GUI OS X application, you can open files in it from finder, a lot of the familiar OS X keystrokes work, you have access to the clipboard and so on.

Vim, however, has a pretty steep learning curve. It's not really like any other editor you've ever used, and so if you don't find, say, textedit lacking in features, you should probably go no further. But if you're up for some adventure, and don't mind doing a bit of fiddling to get your environment the way you want it, keep reading.

The first thing to do, after you've downloaded this, is create a file in your home directory called ".vimrc". You'll probably want to use another text editor to do this, and it might warn you that this is a hidden file when you try to save it, but to get the most out of MacVim with VO, this is essential.

In the file, cut and paste these settings
set winheight=10
set lines=50
set virtualedit+=onemore
set scrolljump=10
set errorbells

No, these aren't the only settings you'll ever need, but they'll make using it with VO a lot more comfortable.

Now that you've created this file, you can actually run MacVim. What you'll see, is a fairly ordinary application: edit area, toolbar, menu. The first odd thing will be, it looks like your edit area's full of tilde characters. You can safely ignore these. They're not actually *in* your file, they're what Vim displays when it has nothing to display. In other words, it's telling you you're in an empty file, and you'll probably never think about them again for the rest of the time you use the editor.

I'm presuming at this point that you know nothing about Vim. If you do, you'll have to put up with a bit of boring commentary first. When you first run vim, you're in what's called "normal mode". You can only enter commands here, not type text.

To start typing, you have to hit i for insert.

To stop typing text, you hit the escape key and you're back in normal mode. Modes are one of the huge things you have to get your head around when you first start using Vim. There are a bunch of getting started with Vim tutorials on the net, and I'll put some of those at the end of this, but the user manual, which you can get to from the help menu in macvim is a pretty good place to start.

The usual things will work in macvim, cmd-q to quit, cmd-w to close a window that kind of thing. Have a quick look at the menu.

So if I'm not going to tell you how to use the editor, what am I going to talk about for the rest of this article? Well, the inevitable, I'm afraid. MacVim is completely usable without question, I'm using it, and have been for a year now, but it's more than a bit quirky with Voiceover.
firstly, VO will speak all characters typed in insert mode, regardless of what you have your verbosity settings set to. Trust me, as a longtime no echo person, I feel your pain.

Secondly, when you arrow around in normal mode, you'll quickly realize that a blank line does not just say newline. What you'll get is "space, space, space, space . . ." etc. I have no idea why this is, as with most of these things, but I haven't found it as irritating as I expected to. Just hit down, or whatever you wanted to do next, and it'll probably get cut off by whatever VO wanted to say next, probably the next line of text. You very quickly realize that space means a newline and hit down, trust me.

A side effect of this is that you really can't practically use any of the VO commands that read large chunks of text in Vim.
You'll just have to highlight what you want to read, and use one of the commands to read highlighted text. If you don't know about these, I've explained them in my VO performance article.

things are a little odd when scrolling through text with your arrow keys. It seems to have something to do with the way Vim itself works, but when it scrolls down a screen of text, the line won't be spoken completely, and you'll have to arrow down and back up again to hear it completely.

It's kind of hard to explain, but you'll see what I mean if you play with it for a bit. It's the reason for setting the window height, the screen height and that scroll jump thing. To be honest, though, again, I've kind of gotten used to it.

All errors and messages are displayed on the bottom line of the MacVim window. Unfortunately, since Leopard doesn't let you interact with individual lines of text anymore in most circumstances, you can't set a hotspot to read it automatically when it changes. This is where error bells comes in. Whenever it pops up some error like not letting you switch buffers without saving or that the editor command doesn't exist, it'll beep. Then at least you can go down to the message line and work out what the error is. You spend a lot less time going "huh?" with it turned on, or at least I found that. Also, you may occasionally find that Voiceover won't read this line. If you find it jumping around or not fully reading it, turn off cursor tracking, and you shouldn't have a problem. Again, I have no idea why this is. Do you sense a trend, here?

And the last one I'll mention here is about arrows. In vim, traditionally, you don't use arrow keys to move through text. The keys are j: down, k: up, h: left and l: right.

The catch for us in Voiceover is with j and k. Only the word you arrow to will be spoken when you use them. In other words, if you're at the beginning of the line and you press j, only the first word at the beginning of the next line will be spoken.

The easy answer to this, given that the normal arrow keys also seem to work, is just don't use those commands. The slightly complicated reason you *do* need to know those commands is that, when marking text to cut and paste or perform other commands, the arrow keys don't seem to work to select text. This can occasionally mean doing things like going into text selection mode, doing a search for the text at the end of the selection and going out again. This will make a lot more sense once you start working with the tutorials.

Anyway, that's the really brief overview. There's a whole lot more that could be said about vim and MacVim, and probably about VO and MacVim, and I'm sure if any of the rest of you decide to give it a go and find it useful, there'll be tips to exchange. Maybe even, since this is open source, more work could be done with the accessibility layer to make this easier to use.

resources:
http://www.vim.org/
Use vim like a pro: http://tottinge.blogsome.com/use-vim-like-a-pro/
Efficient editing with vim: http://jmcpherson.org/editing.html

 What can you expect to find here? I'm a blind computer programmer who uses a Mac

 So tutorials on Voiceover, the OS X screen reader, articles and links about programming, and whatever other random stuff happens along.