Joel Spolsky – Keynote

Joel’s keynote was on the theme of simplicity vs power, or rather, the third way.

I’m afraid that this is my interpretation entirely from memory, but essentially:

• A couple of fundamental assumptions are that users do not like to make decisions, and that just about every design decision should lead to the user being more likely to get laid.
• There has been a great upswing in simple interfaces, at least partially lead by 37signals. (Perhaps another contributory factor is the rise of internet applications? Services that can do one thing well.)
• However, some types of commercial applications (e.g. bug trackers) will need to cover the 80% of features that aren’t covered by these ‘simple’ applications. Alhtough 80% of people use 20% of the features, they are often different features. Otherwise you are cutting off too many people.
• Creating elegant code often takes more time and thought that simply typing out standard code. NB: Elegant was defined as doing things well, and making it look effortless.
• Creating elegant interfaces also takes more work, both in creating the UI and often more so in the coding behind it.
  A good example is the Amazon one-click. When you move form the (standard?) method of click, then confirmation screen (then rest of checkout process), you suddenly have more work to do. For example, queuing the order (in case they order more things), allowing undo within a certain time, making sure there’s a default address etc.
• Rather than not adding features to keep an interface simple, keep it simple by not forcing the user to make decisions. The interface should certainly start simple, but allow it to be extended, perhaps on a per client/site basis.

Joel’s assertions certainly match my experience with Content Management Systems, which are probably the hardest and least solved programming / interface problem. Whilst 37Signals have been very successful, I don’t see them tackling CMSs any time soon…)

Michael Sparks – Python

Michael essentially took us through the finer points of Python by demonstrating how you can create the Google-style spelling match function, in 21 lines of code (plus a few books worth of content in text files).

It was really good to see an expert use the Python interpreter well (including running into a couple of problems, and solving them). I’d never really understood how you use it for complex functions before, for example,  indentation matters in Python, and you put it in a command line?? Anyway, that explains the spaces being preferred to tabs thing, and it will certainly help in my dealings with Django. Hopefully Michael can post something about it soon, in the meantime the code is available.

Joel Spolsky – Fogbugz

We had plenty of Joel time today, this was his demo of Fogbugz, where it’s a disservice to describe it as bug-tracking software. It’s obviously modelled on Joel’s copious experience in developing software, and includes things like Bayesian driven email direction, and predicting launch times better than developers. I’m not sure it would work for us (juggling 15 web projects at the same time), but it will certainly make me wish for more next time I look at Trac or Bugzilla.

Reto Meir – Android

I have to admit, the technical side of mobile application development (rather than mobile web development) is still pretty intimidating for me, and doesn’t particularly float my boat. Reto also needs to either increase the font size in Eclispe, or use the ‘only when the cursor reaches the edge’ settings in OS X’s accessibility preferences.

From doing previous presentations I’ve learned the hard way to:

• practice it at 800×600 resolution,
• check it in 1024, and
• even if the projector supports more, keep it to 1024!

On the positive side, it was good to see some use of Google moderator. Not only does it make it easy to ask questions (and for people to vote up the best questions), Reto answered the top ones on the spot, and has since answered them all.

Remy Sharp – jQuery

Remy gave good talk on jQuery (presentation), something I’ve used quite a bit. It was great to include how to creat a plugin, and it explained a few things I’d seen around but didn’t understand. I liked Remy’s approach to the brief of using code,

The things that most interested me were:

• Having looking up “equality without type coersion” (the triple equals).
• jsbin.com looks like a very useful service.
• There’s a good IRC channel for getting help.
• How to add a plugin to jQuery.

Jeff Atwood – Stackoverflow

Jeff’s talk was less on code, and more the experiences around Stackoverflow, I can’t remember of it much specifically, but after listening to lots of podcasts it’s good to put a face to the voice!

Pekka Kosonen - Qt

Pekka’s presentation on Qt was intriguing, partly for the upfront honesty (tackling people’s scepticism about platforms from Nokia), and partly because of the possibilities of Qt.

Although Nokia was presenting about Qt, the platform itself is not originally from Nokia, in fact, it is designed to build cross-OS applications, desktop and mobile. Applications such as Skype and Google Earth have been built with it, so it seems that it can produce native looking applications.

It looks like you have to go whole-hog into Qt development, it’s an SDK, IDE and they may even have server-side compiling of code to for platforms you don’t have. The demo wasn’t incredibly convincing, but it does seem like a an option to investigate if you need to develop desktop/mobile applications across several platforms. See also getjar.

Phil Nash – iPhone

Phil’s presentation on iPhone development wasn’t quite what I expected, it took us through a short history of the language that iPhone development uses (Objective-C), before creating a little app from scratch. Phil’s dry sense of humour and occasionally cheesy slides helped make the first half more entertaining, especially as I wasn’t really target audience and haven’t quite got the different versions of C straight yet.

The outline of Objective C drew a few gasps: garbage collection? It seemed most of the audience are used to higher level languages, and Objective C seems like a step backwards.

However, the impression was improved a great deal once Phil opened up Xcode and put together a quick application. Phil was the only mobile developer who managed this. As well as the support of the IDE, emulation is obviously easier for the iPhone than Android or Qt, because you’re targeting a much more narrow range of devices.

Jon Skeet – Humanity: Epic Fail

The ‘phenomenal’ Jon Skeet presented a very funny look at why programming is generally hard, based on a rant he had recently about time zones. There isn’t much for me to say on this, but it’s worth checking out the presentation and audio on the Stackoverflow podcast soon.

Paul Biggar – How not to design a scripting language

I hadn’t really expected much of this talk, essentially what’s wrong with modern compilers, however, Paul’s enthusiasm really carried it.

Christian Heilmann – Yahoo! Developer Tools

For me the best was saved until last. I haven’t seen Christian talk on this before, and despite reading about Yahoo! pipes and YQL, I just hadn’t really groked it before.

It doesn’t really hit home until you see someone put an example together in front of you, but the upshot is that there are about a 100 APIs (everything from Amazon to Weather) that you don’t need to read.

You select a data source from the right of the console, adjust the ‘selector’, the YQL statement top left, and preview the results in the middle.

If the result is good, then copy the URL from the top right box, and use that with your backend (or ajax/ajaj), and hey-presto.
Example uses are:

• Showing all your disparate content on other sites in one place, for which Christian’s site is a prime example. It’s basically built with YQL, virtually all the content is from other sites! (Presentations, videos, books, blog articles etc.)
• YQLAutoTagger – automatic tag generation with a single line of JavaScript.
• Translate tweets.
• Create a set of Flickr photos from a particular geo-location.
• You can even scrape HTML for redisplay, narrowing the area using an xpath statement.

It uses a SQL like syntax to allow a great deal of granularity, both in terms of filtering things to get just the information you want, as well as using several input sources (URLs) at the same time.

I am not closing the YQL tab I have in Firefox until I’ve used it somewhere! I think this is the closest video on YQL I’ve seen that covers the same content.

Windows IE supports click-through such that if you happen to click on a link in a background window, it will follow that link.

I fall for this a lot, especially with a large monitor and multiple windows side by side or overlapping. And yet, why doesn’t the mouse wheel scroll for a background window on Windows? And on the Mac, where you have to click on a background application to activate it, why does mouse-wheel scrolling work?

Now, I do very much prefer the Mac way of allowing the scrolling of background windows by just moving the mouse over them and scrolling. In windows, you can allow this with an add-on: Katmouse (as recommended by Steve Gibson).

Is it inconsistent to generally disallow interactions with background applications, but allow scrolling? John Concludes:

in most cases, buttons are not good candidates for click-through. But draggable objects, like file and folder icons in the Finder (and similar apps, such as FTP programs), are.

I would add scrolling to that list, and continue using something to enable it on Windows as well.

At the end of its third week of availability, Google’s Chrome accounted for 0.77% per cent of the browsers that visited the 40,000 sites tracked by Net Applications, down from a 0.85 per cent share the week before.

On sites that I have access to, it’s varied between 0.4% and 2.6%, not particularly climbing or dipping. I heard of others having 8% on the first week after launch, dropping back to 1 or 2% the next week.

However, these sort of stats are probably missing the point, what sites is it that people are most likely to use Chrome on?

I suspect it is Google’s own sites, or rather Google’s applications.

It is an express aim of the browser to work well on their apps, and for people to be able to create their own ‘short-cuts’ to applications.

We seem to be moving away from ‘the blue E is the internet’, to application icons for Google Mail, Reader etc. Perhaps BaseCamp and others as well. That is what I now have on Windows at least.

NB: Comparing Windows based screen readers with VoiceOver is very difficult, and probably a fairly pointless exercise. I’m not trying to do that.

DOM Navigation

The first thing to check is always the settings, as these can have a massive impact on how a screen reader reacts, setting like ‘verbosity’ will change what is read out. Under the "Web" section of the VoiceOver utility, there is a new setting to choose "DOM navigation" as well as previously available "Group navigation".

One of the unique aspects of VoiceOver compared to traditional Windows screen readers was that you navigated in a 2 dimensional manner, able to go up/down and left/right across a page. This had the effect of making some things quicker, but also make it possible to miss content. For example, sometimes indented text (e.g. bullet points) would be missed because it was treated as a different column.

DOM navigation should eliminate that possibility, as the user goes through the content in source order. However, that does mean that in many cases people will have to go through more items before getting to the content. Therefore within-page navigation mechanisms become more important (see links and headings below).

DOM navigation is the default setting in VoiceOver, although some people may change it back.

In more general terms, VoiceOver no longer wraps around pages or other interfaces by default. When you get to the beginning or end of a page or interface, you get a definite ‘thud’ sound effect, like a door closing.

Implications

Source order matters, so the linear order of the HTML should make sense and be consistent.

Structural Elements

Previously, VoiceOver basically read out text. That was it. Nothing structural was announced, let alone navigable.

There are now quite a few ‘navigate by structure‘ commands, such as jump to the next heading, or the next quote. Unfortunately it seems that only the heading and link commands work.

Tables

The lack of special features for structure is especially unfortunate for tables, as it means tables are treated as standard text, read out row by row. There are commands in the manual for skipping to tables, reading out rows and columns, and reading the column heading of a cell. These commands work very well in iTunes (in a brief check), but don’t seem to have been applied to HTML tables in Safari.

So should web developers not use data tables? Or course not, even apart from it being perfectly valid in the spec, it’s a bug that’s likely to be fixed, perhaps even overnight in a minor update (in the same way that iTunes just started working).

Headings

Headings are now read out, for example the main heading on this page would be read out as Heading level 1, VoiceOver 2, a WebDev’s guide (recording in mp3). This should help users understand the page structure much better, assuming it uses headings appropriately.

There are also good jump commands, with skipping between headings, and also skip to the next/pervious heading of the same level. I put together a quick HTML test page for the purpose of testing this (I was offline at the time), and skipping to the next heading of the same level (e.g. 2) was invaluable, as you could skip the subsections.

Links

The links functions have changed quite a bit, although you can still create a list of links (VO – u), you can now skip to visited links as well.

Also changed are how they are announced, VoiceOver has options to have the link announced with the word ‘link’, a ping, or a change in pitch. I found it difficult to discern the difference in pitch with the default voice, especially with the pause before and after the link. However, the ‘speak link’ and ‘play tone’ options are very good.

One minor bug (perhaps) seems to be that links are not announced directly when in headings. You can get to them, the heading would be read as Heading level 3 with two items, and you can ‘interact’ with (think go down a layer into) the heading, and get to the text and link as though it were a paragraph.

You can also read out the URL of a link, which is very useful as the status bar does not show the currently focused link without using the mouse cursor.

The major bug with links is that within page links don’t work, although the screen moves, the location of the VoiceOver cursor does not. There is a workaround where you put the mouse cursor within the window, use the link, and put the VoiceOver cursor where the mouse is. I don’t think this is something you can expect a regular user to know or do.

Inline elements

I’m not quite so enthused with how VoiceOver deals with inline elements such as b, abbr etc. It basically stops. So if you add a strong to a sentence, VoiceOver will read out the text up to it, stop, the user presses next, VoiceOver reads the text in the strong, and stops again. (Recording of the last sentence.)

It has the effect of completely breaking up a sentence, which is rarely what was intended by the writer.

Given the excellent new voice, it seems strange that a different intonation or other effect was not used to indicate inline elements.

Images

Fairly straightforward: The alt attribute is read out followed by "image".

Styles

You can read out the appearance of any text, e.g. default text is: “16 point helvetica, black on black”. I’m not sure when that would be useful for browsing, although it must be very useful for editing documents.

Block elements

There is a little sound effect for going between block elements. Each time you go right (which often means down through a page), it gives you a little ‘do-di’, and going backwards through the document reverses it to ‘di-do’. However, apart from headings, structural block elements aren’t announced.

Lists seem to have taken a step backwards. In VoiceOver 1, bullets and numbering were read out. But in version 2 a bullet or number is treated as a new line, with no indication that it is a list.

Implications

The biggest problem here is that VoiceOver doesn’t understand within-page links, they essentially seem to be broken to the user because nothing happens. Although not that many sites use larger pages with internal links, if you wanted to change something on the screen with AJAX and send the user to the right place to read it, you can’t with VoiceOver.

VoiceOver has no caching issues that affect the Windows based screen readers, however, there is also not automatic announcement of updates, or means for a developer to send the user to the updated part of the page.

Title attribute

Hoorah! VoiceOver can read title attributes on almost anything. Termed ‘help tags’, VO-h will read out the title on acronyms, abbreviations, and just about any inline element I tried.

Titles can also be read out from block elements like ps and divs. If and if there are titles on parent elements it did what I hoped: read out the one closest to the current element.

If only you could tell when something has a ‘help tag’, as there doesn’t seem to be any way to get them announced.

Implications

VoiceOver users aren’t going to notice abuses of the title attribute, but without other uses in the OS I can’t see it being used much.

Forms

Forms behaviour in VoiceOver definitely favours better marked up forms, as text inputs, radio buttons and check boxes all use a properly attached label, rather than assuming nearby text is probably the label. Other elements (textboxes and selects) do not announce the label, presumably because they generally have sufficient explanation either within or around the element.

Given that there is no ‘forms mode’ that the Windows based screen readers tend to use, not using labels on these elements will probably work reasonably in most situations, especially when the explanation is read out first. For other screen readers, using the forms mode means you can loose the context around the form elements, which isn’t the case here.

There is one fairly major bug that was identified on the MacVisionaries list (I think by Rich Caloggero) where a multiple select is simply ignored by VoiceOver, and just completely skipped over. This was introduced in the Safari 3 beta, and has remained since (hat-tip to David Poehlman).

Accessibility oriented attributes

tabindex and accesskeys are generally not needed or usable, but it’s worth knowing what different technologies do.

Tabindex in VoiceOver works as it does when using Safari in generally, anything with tabindex is first in the tabbing order, other links and form controls come afterwards in source order. Unfortunately from an ARIA point of view, values of 0 and -1 on non-links don’t do anything.

Accesskeys work with the ctrl key, and generally work except for clashes. For example, I started off testing with "1" like the UK government guidelines suggest, but that activated ‘spaces’ rather than the accesskey.

Implications

Again, if you are using AJAX type updates in the page it is difficult to help screen reader users. Use of tabindex is likely to be confusing if people do use it, and unnoticed if they don’t.

Accesskeys will work in general, but aren’t announced and are likely to be unused without going to extraordinary lengths to draw people’s attention to them.

Flash content

You wouldn’t know it existed.

Overall

The new version of VoiceOver has many improvements from a web browsing point of view, albeit with a few bugs and omissions remaining. The headings and DOM navigation are likely to have the biggest impact on people’s use and the general usability of pages when using VoiceOver.

The main aspects I would like to see improved are making within page links work as you would expect, and enable table functions to be as good in Safari as they are in iTunes. There will also be a lot of content locked away behind Flash, even accessible Flash.

These weren’t under the Safari bit of the VoiceOver manual, but are in the command list at the bottom. After lots of text-things that probably aren’t very useful (e.g. find the next italic text), you get:

Find the next item of the same type or the same text style

VO-Command-S

Find the next control

VO-Command-J

Find the next element that is different

VO-Command-N

Find the next table

VO-Command-T

Find the next heading of the same level

VO-Command-M

Find the next heading

VO-Command-H

Find the next blockquote

VO-Command-Q

Find the next visited link

VO-Command-V

For finding the previous instance instead of the next, just add shift to the combination, and there is still the type-ahead-find style link chooser.

I’ll have a more thorough play later, but interestingly there are two modes of navigation, the DOM mode:

You can move the VoiceOver cursor in a linear way, progressing across or down the website by moving from item to item on the page. This method provides excellent control but may slow you down if the page contains many items. This method uses the Document Object Model (DOM), which is the order specified by the developer of the website. Depending on the website, it can seem random.

And the ‘grouped’ mode:

You can navigate by grouped items. The VoiceOver cursor moves from one group of information to the next, and you interact with the group if you want to read its contents. This method speeds up navigation and allows you to quickly survey the contents of a page.

VoiceOver used to primarily use the second mode in version 1, and you used to get some strange odering based on the visual layout. The DOM mode (linear) should be much closer to Windows screen readers, and should work better for accessible sites.

The meeting was at an LBi office in London, and had a great start with free pizza’s & beer, which after my three hours of travel was a very welcome sight. The topics up were SharePoint 2007′s (SP) search, and creating an accessible web site.

The first part on search went mostly went over my head, not due to the presenter (Riaz Ahmed), but because I’m really not an MS developer. It was quite funny though when Riaz asked the audience how many people used MS’s Live search, and three people put their hands up (out of around 50 MS developers?).

The presentation is up now, and the main thing that struck me in this part was how many damn servers you need to use, the minimum in any of the diagrams was three (web site, search & database). The example used was for LBi’s intranet, where SharePoint 2007 enterprise search was used to index about 6 million documents on a file system, several thousand wiki pages, some blogs, and the ‘people’.

One relevant (to me) question at the end of the search presentation was whether you could filter the search terms to prevent SQL injections. In this example it wasn’t really an issue due to being an Intrant, but they had used JavaScript to filter the search terms. This lead onto ‘is sharepoint reliant’, to which the reply was Yes and no, it depends how you architect it.

Method for Creating an Accessible SharePoint Site

After outlining their typical development team, Riaz suggested that it should be the ‘Experience Architect’ (aka Information Architect, Usability person, Interface designer etc.) who after suitable training should own the:

• Site content types and Page layouts
• Site structure
• Navigation

These are all things in SharePoint with relatively easy interfaces to control, and this represents quite a good direction for things to go.

Their development environment includes:

• Visual Studio 2005
• VS2005 extension for WSS
• Source control (TFS), because the expected source control method isn’t really up to it.
• SharePoint designer (not to create templates, but to validate the SharePoint aspects.
• WSS/Sharepoint SDK.
• Reflector

Apparently you also need you own SP farm!?

Issues

The main issues they had found with the default output from SP were complete lack of valid code, custom attributes, and inline styles.

I have to say, I really got the impression that for them, accessibility = valid HTML. This might be unfair, I’ll wait until the site goes live before commenting further (about structural code, alternative text etc.). This is not the new site yet.

Solution

They looked at several ways of controlling the code, and after an unsuccessful attempt with ‘Web Parts’, settled on filtering the output. Andrew Lister was quite funny and approachable, despite the obviously depressing content of the topic.

So, for any non-authenticated user, the content goes through a filter that strips out the rubbish. Obviously this put some overhead on the system, so they instituted caching, and asked Microsoft about a better solution.

Apparently Microsoft said wait until the next version, to which an audience member said They’ve been saying that since version 1!. (Alun David has been asking for about 6/7 years.)

The implications of this approach (and I certainly don’t have a better approach, this isn’t a critisim) are:

• You can’t use any out-of-the-box functionality. Nothing. Or at least nothing that goes to the front-end site (they weren’t concerned with authenticated users).
• The output is only as good as your filter, so if a user puts in unusual code to the content, it may cause unexpected effects.

For functions like the survey, they wrote a module from scratch, which is what you would have to do for any function. The normal SP functions are used for back-end functionality, but Riaz estimated it was 80% .Net development, and 20% SP development.

They will be doing another session in the near future that will go into more of the technical details.

I also have to wonder about the interface of the content editor, does anyone know how configurable that is?

Q & A

In the questions and answers it turned out that some people were working on making accessible Intranets, and had been talking about the issues with Microsoft. Apparently there have been some rumblings going up the food chain at MS in the last few weeks, I wonder if Molly’s post had anything to do with that?

Most people were pretty hacked off that MS hasn’t done anything about this, and if people using SP are hacked off, perhaps MS will listen. From the comments the best way to register complaints is through SharePoint NVPs (whatever they are).

Most CMSs started working on making sure they could output decent code a couple of years ago (with various level of success), but since SharePoint 2007 is now Microsoft’s primary CMS product, the question is: when will they start trying to catch up?

Further reading

• Alun David posted his thoughts on the Sharepoint2007.com site.
• The LBi presentation.
• Discussion on accessify

The content ‘protection’ involves fairly draconian measures, but I hadn’t realised what the impact could be for other systems, not just Windows.

Change from an open platform

The Security Now podcast interviewed the author of the article (transcript). This was the section of Microsoft documentation that summed up the change in direction:

It is easy to write software for it [Windows] because all the interfaces are well defined and published. And there are many good software tools available. The PC buses are also well defined, and anyone can design cards to plug into these buses. The openness of the hardware platform is essential to a vibrant PC ecosystem. In the current world, however, the industry is also working to prevent hackers from using that openness to pirate copyrighted content. The goal is to make the Windows-based PC a safer place for premium content, so that content providers will be happy to allow Windows-based PCs to play their content.

The stated aim is to make the new system (Vista and associated hardware) less open. According to Gutmann, Microsoft want to demonstrate to Hollywood that they’re really, really committed towards content protection.

The type of things you may notice with a new Vista install are:

• For viewing High-Def (HD) content such as HD-DVD or Blu-ray, the graphics card needs to support HDCP, a proprietary DRM mechanism, which includes encryption keys.
• Using non-protected mechanisms like S/PDIF (fairly standard for CD and other audio) may mean your sound is disabled.
• Without an HDCP enabled graphics card and display, premium content would be degraded. Viewing on an analogue (i.e. VGA) output will be intentionally fuzzy.
• The graphical sub-system restarting if it detects something dodgy, (possibly including fluctuations due to cheap hardware).

Another couple of telling comments:

• you’ve got this huge layer of bloat sitting in there, intercepting all audio and video output and content output and processing.
• the current processes simply don’t have the horsepower to do both video processing and encryption of high-bandwidth content.

You’ll notice that these things primarily affect ‘premium content’, such as the new HD formats. The worry for me is how long it is before you can only get the HD versions of things (mainly films I assume), and have to use these DRM encumbered systems (and encumbered definitely sounds like the right term, with all the performance issues associated with it).

There isn’t any point in using the new formats unless you have a full HD chain of hardware/software.

Effect on graphics cards

Device manufacturers are also having to show commitment to preventing copying of ‘premium material’, including keeping certain parts of their systems secret:

the language in the spec is kind of wishy-washy. It doesn’t say you can’t publish any specs at all. It simply says that some aspects of the device will have to be kept proprietary. But that’s kind of nasty for anyone writing open source software because, if you’ve got a graphics card, you need to know exactly how it works in order to write the drivers for it. And if half the thing is undocumented, it becomes very hard to support it properly.

It’s one of the recurrent themes on LUG radio, there are bugger all open-source graphics drivers. The changes graphics cards makers are having to make could drive the nail into the coffin of that possibility. There are (relatively) open alternatives, but I don’t know how good they are for games.

A second podcast from Security Now on the topic explores more of the Microsoft documentation, and aspects that imply your drivers could be revoked if Windows suspected tampering.

I ask quite a lot from an operating system, the main thing being that I can use formats that are interoperable, and preferably open. After that, it is a tool to act on that content, be it emails, films, documents or pictures. The more effective and easy to use it is is, and the less you have to maintain it, the better. For me, OSX has the best balance at the moment, but if a Linux distro offered as good an interface and hardware integration, I would be swayed.

Presumably Apply & Linux will have to jump through similar hoops, I just hope they can do it with less impact on the rest of the system.

Combine the content and performance issues with the insane activation scheme, perhaps Vista isn’t going to stem the (admittedly small) flow from Windows.

• operating systems
• windows
• OSX
• linux
• DRM
• HD-DVD
• data formats

I did a review recently, which drew some fire after Joe Clark posted it to the very active Macvisionaries list.

Whilst the review is valid in what it attempted, I will be posting an update because:

it turns out that using the official documentation probably isn’t the best way to learn how to use Voiceover. Although the points above are still generally valid, with help from the list, I will post an update soon. One point that will be corrected is that some of the difficulty with keyboard controls can be overcome by locking the ‘Voiceover keys’, cntl-cmd.

In the mean time, if you are interested in learning how to use Voiceover, make sure that you start from the install process rather than reading the manual first, and check out the newly set up Macpedia wiki.

If you want to learn to use Voiceover, the best way would be to go through the installation tutorial, and then read the very concise notes I made that would have helped me with things like “interacting with content areas”, and locking the “Voiceover keys”. (Also posted on the wiki.)

The review is still valid because there are very few accessible applications with Voiceover (in comparison with JAWs/Windows), and browsing web sites has quite a way to go.

However, there are a few reasons to pay attention to Voiceover.

It’s on OSX

Stating the obvious, but it make a huge difference to how the screen reader works. JAWs, Windows Eyes et al. have to work around the MSAA layer, and in comparison, feel very flakey.

JAWs in particular is a long way ahead of Voiceover in terms of supported applications and to a lesser extent for functionality, but Voiceover is part of the OS. It isn’t trying to work around the OS, and this makes a big difference in (perceived?) stability and reliability.

Strong community

There may not be many Voiceover users, the the community is very active if Macvionaries is anything to go by.

Apple hasn’t been particularly forthcoming with help & information (as is their usual MO), but the community is filling the gaps with the wiki (after a little prod ).

It’s early days

It’s only a version 1 product. Most of the Windows screen readers are at version 6 or more, Voiceover is probably better than they were at version 3 or 4.

It is difficult to assess Apple’s commitment to accessibility and Voiceover. There are some negative views on this, but I would hope that the thought of extra sales due to fulfilling Section 508 criteria might help. (I’m assuming that having a screen reader and magnifier help, but I don’t know US law that well.)

Chris Hofstader commented that Apple wasn’t interested in making the iPod accessible:

from the very first iPod released a few years ago to the fanciest one out there today, all had more than enough compute power and storage (with zillions of bytes left over) to run a speech synthesizer.

I’m not so sure, it seems that Apple are investigating (in fact, patenting) methods of creating a text-to-speech interface.

Probably not for accessibility reasons, granted, but it’s an interesting approach (using the host computer to create the speech, which is simply read out by the iPod when on the move). That would indicate that the performance of a standard voice synthesizer wasn’t up to the job.

The crux

Apple isn’t usually the first mover, but it tries to be the most stylish & usable product, and won’t rush something out to satisfy one audience that might get in the way of the majority. Apple could be working furiously on Voiceover behind the scenes, but we’d never know.

If Voiceover is improved, and a few more crucial applications become usable with it (Office & iTunes for example), it suddenly becomes a very cheap and easy alternative to Windows and JAWs.

The test for Voiceover’s future will be to see if it is updated in the next release of OSX (Tiger). If it is, I’ll be one of the first in line to try it out.

• Accessibility
• Voiceover
• Screen readers
• JAWs
• Apple
• OSX

One of my favourite podcasts is LUGradio, one of only two that I actually subscribe to (although I have many one-off podcasts). This is primarily for the entertainment value, but it is also good to keep up with what’s happening in the Open Source Software (OSS) world. One of their main ‘beefs’ is that any software they use should be open source, primarily on philosophical grounds.

I can’t remember a specific quote from LUG radio, but in parallel, Mark Pilgrim recently caused a minor kerfuffle because he switched from Apple to Linux, after 22 years of using Apple computers.

Mark has been a firm believer in open source software for years, and wrote possibly the most enlightening article on why open source software is more valuable which concludes:

In the long run, the utility of all non-Free software approaches zero. All non-Free software is a dead end.

However, the main issue of his switch revolved around data, the final straw being the undocumented proprietary email format that OSX uses for it’s mail program:

It took a lot of forethought on my part, not to mention actual time and effort, to convert all my disparate mail archives from all those different mail programs. I finally got everything into a single archive in an open, stable format… and just 3 short years later, Apple found a way to screw me one last time.

There are also some good points on data fidelity, as repeated conversions will loose data.

I don’t care if the software I use is proprietary, but I will refuse to use something that locks my data in a format that only it can read/write satisfactorily.

This way of thinking is common for ‘Web 2.0′ applications, from Simon Phipps:

it takes full substitutability for me to have the confidence to stay as well as the freedom to leave. That’s why Stewart is spot-on with Flickr’s policy and paradoxically will keep my business by allowing me to leave at any time.

But this approach is far from common in desktop programs, even with the greater overlap you see now between desktop and online applications.

Backups and data longevity

Tim Bray made a fantastic post on doing backups well. My approach is similar, but I tend to simply use rsync and copy my home directory (or ‘my documents’) to an external hard drive. (Yes, you can use rsync on Windows to.)

But for this approach, in fact, any backup approach to work, the format you save things in will determine if you can use the data in future. For example, the only documents I had from before University are Wordstar files, which I can’t read anymore because I don’t have that program. That is on the same platform, what happens to all that music you have in Windows Media format if you want to switch to another OS?

For these and other reasons there are only a few formats that I am happy using. From a cursory glance at my files:

XHTML

A human-readable text-based format is always a good start, and (X)HTML will be usable as now for many years. Even if we stopped using browsers, it is built from a public spec that could be used to build a browser in future. I use this for web sites and presentations.

JPEG / GIF / PNG

These are all based on published standards, and although not necessarily ‘open’ they are common enough to be supported for the foreseeable future.

MP3

Actually Ogg Vorbis would be the correct choice here, however, player support is still limited, and due to most people thinking MP3 = digital music, support is pretty much universal. All of my music is in MP3, or quickly converted to MP3 after buying through the iTunes store.

Open Office Docs

Open Office provides a pretty good alternative to MS Office, and uses an open format that is essentially XML. The XML and associated files are saved in what is effectively a zip file, so even if there were no office applications in future, it is still XML. It also uses the OpenDocument Format, which Microsoft will reluctantly be supporting.

PDF

A slightly surprising choice perhaps, but as far as having documents that are intended to be printed, PDF is a perfectly valid choice, based on a published spec.

Choice of Operating System

So what does that mean for my choice of OS and applications?

These days the choice is fairly open, with the Mozilla organisation providing the main applications I use (Firefox & Thunderbird) across the three main platforms, these aren’t a point of contention.

Another important application for any web developer is the humble text editor, and my first choice is jedit, another cross platform choice. Open Office has always been available for Windows and Linux, and now the Intel Macs have been released, it is no longer the poor cousin on Macs.

For media players, I favour iTunes, Media Player Classic, and also keep VLC. I switched to iTunes well before getting an iPod because it was far superior to anything else for organsing your music. It helped me turn 3,000 poorly IDed tracks into a coherent collection. It was much quicker at listing large playlists than Windows Media Player was at the time.

Thanks to jHymn I can buy music from the iTunes store and convert it to a regular (unmarked) MP3. However, it doesn’t work with iTunes 6, so if I switch to a Mac I’ll have to go back to buying CDs to rip.

That covers the main applications. Obviously it is somewhat different at work, it is very difficult to avoid the Windows lock-in there.

Deciding factors

So what OS would I choose at the moment? It should have everything I use currently (under Windows), but I would like better terminal access, which points me towards either OSX or Linux.

I’d make a different choice to Mark Pilgrim, despite having similar concerns. The deciding factors are:

• Hardware – Software integration: There are certain things that are hard to do on Linux, or Windows for that matter, that require good hardware/software integration.
• I’m not prepared to spend much time configuring a Linux system. I will give Ubuntu a go again at some stage, but it seems a lot easier to get going quickly on a Mac.
• I’m not a programmer, and whilst Mark Pilgrim or Tim Bray can contribute to changing an Open Source OS, I would be way out of my depth there. I’m also not willing to wait for it to happen.
• Windows genuine disadvantage is really annoying, Apple may not be very open, but they aren’t actively snooping.
• Windows Vista appears to be getting watered down, and probably won’t be where OSX is now, regardless of any improvements from Leopard that will be released about the same time as Vista.

I’m in a different position to Mark, and have made different choices (e.g. using IMAP).

So there you have it, I’m saving up for one of these.

Tags:

• operating systems
• windows
• OSX
• linux
• backups
• LUGradio
• data formats

However, if you do suffer from spam-bounces and use Debian & Exim, this may be the answer you are looking for.

The problem

I run my own mail server, have my own domains, and generally have complete control over my email. Part of the reason I went down this route is because it’s nice to have your own domain, but you end up getting a lot of spam. If your domain has been around since 1999, there is no way of avoiding it.

Having got a great Debian Linux set up from Bytemark, it comes with Exim (Mail Transfer Agent), and although I am not a sys-admin, it is fairly easy to set up Spamassassin and ClamAV to protect your inbox.

However, when some bloody spammer decides to use your domain as the reply address for their spam, you will receive hundreds or even thousands or bounces to random addresses at your domain. Wonderfully meaningful addresses such as “ccumcrt@mydomain.com”.

Last year I got about ten thousand bounces in 3 weeks, and then it stopped. This year I’ve been getting about 20 a day – very annoying because they aren’t actually spam, they are replies to spam.

Where and who solved the problem

Well, it wasn’t me! Did I mention I’m not a sys-admin?

I joined the Exim users list, investigated and Googled first, checked the docs (I’ve not been told to RTFM yet), and then posted my question.

I’m very glad to say I got a good response, lots of help from an active community. Special thanks go to Jakob Hirsch who’s first reply contained the best answer.

There were quite a few replies, and if you are interested in Exim configuration and whether to have a catch-all address, it is quite revealing. (The second part of the thread is in the next month.)

Solution

Jakob’s suggestion uses the fact that bounces don’t have a ‘sender’ header (examining the ones I’ve received confirms this), and combining this with the random local addresses, it gives this configuration to add to the ACL check_rcpt config file (on Debian that’s the 30_exim4-config_check_rcpt file).

deny
  senders = :
  ! local_parts = known@email.address : another@known.address
  message = This address never sends out mail, so it cannot get bounces.

Another poster provided a page which is trying to do something quite similar, but didn’t seem to work for my setup.

Jacob’s suggestion didn’t work exactly, but with a few experimental tries and combining with other replies, this is what works for me:

  deny
    senders = :
    !local_parts = ac : (other 'local parts' before the @, in a colon separated list.)
    message = This address never sends outgoing mail. \
              You are responding to a forged sender address.
    log_message = bogus bounce for system user <$local_part@$domain>

This basically says “if there is no sender and it doesn’t match a local address, then reject it”.

For full domain references you would use:

!recipients = user1@example.com : user2@example.net

Later down the thread there is a method for using your alias file instead of hard coding the potential addresses. I couldn’t get that working with my multiple domain alias files, but if I do I’ll update this page.

Fingers crossed it seems to be working, although my tests are probably quite basic:

1. Can I send email?
2. Can I receive email?
3. Can I receive a legitimate bounce? (I emailed a non-existent address at work)
4. Does it reject a bounce to a non-local address?

List of bounce subject lines

Just in case it turned out to be useful (e.g. for filtering), I kept the bounces. In case anyone finds it useful, here are the key ones that cover most of the bounces I received:

• DELIVERY FAILURE: (subject)
• Delivery Notification: Delivery has failed
• Delivery failure: Returned to Sender
• **Message you sent blocked by our bulk email filter**
• Delivery Status Notification
• Delivery Status Notification (Failure)
• Mail System Error – Returned Mail
• Mail delivery failed: returning message to sender
• Returned mail: User unknown
• Returned mail: see transcript for details
• Undelivered Mail Returned to Sender
• failure notice

This was taken from a grep of the folder I filtered them into, I’ve got the full list if anyone wants it, but most of the rest just had the subject from the initial email.

Tags:

• spam
• Email bounces
• exim