Microsoft Office

Out of the two announcements, Microsoft's is probably the least exciting, but, on the other hand, there is a good chance that it will see a lot more actual use than JaJah's voice translation. The Microsoft Research Machine Translation team has just released this update to MS Office 2003 and 2007 to the Office team for integration, but they already offer instructions on their blog for setting this up yourself without having to wait for the official update.

The integration with Windows Live Translator allows you to translate English texts into Chinese, Japanese, Korean, French, German, Italian, Arabic, Dutch, Portuguese, and Spanish, as well as vice versa. We have tested the Windows Live Translator and the translations were generally about as accurate as you would expect from machine translations. There are various mistakes and words it doesn't recognize, but overall, the translation is relatively readable and gives you at least some impression of the original text.

JaJah Babel

JaJah Babel is clearly the sexier product of the two. You can call access numbers in the U.S., England, or Australia, and after a voice prompt, you simply speak the text you want to be translated into Mandarin. The service will then replay your message, you acknowledge the accuracy of the input, and after a short delay, you will hear the translation. Given our general lack of knowledge when it comes to Mandarin here, we can't vouch for the accuracy of the translation, but the service itself worked very well and seemed to understand at least our initial input accurately.

If you are in China, of course, the fact that you have to call an international number to get this to work is a bit of a limitation.

Other Translation Services

There seems to be quite an interest in working on consumer oriented translation services right now. Just yesterday, we wrote about Mloovi, which translates RSS feeds trough Google Translate, and earlier last month, we wrote about the collaborative dictionary and translation service Lingro.

Babel Fish

JaJah's product is especially interesting here because it takes speech as its input and it will get even more interesting once it works for other languages beyond Mandarin as well. JaJah is offering this service based on IBM's technology, and given IBM's expertise in doing voice-to-voice translation, it will probably only be a matter of time before we see support for more languages. Besides other projects, IBM already supports the U.S. Army with an English to Iraqi Arabic translation service.

There has always been a lot of hype around the possibilities of instant voice translations, but very few products were ever good enough to make it in the consumer/business market. JaJah represents a major step forward here, even if its voice prompts make the service a bit less frictionless than the science-fiction ideal if autmated, instant translation.

There's also a problem of congestion. In densely populated areas, as more and more people set up home networks, throughput is dropping as the airwaves get cluttered. David Heinemeier Hansson posted today about noise on his urban wifi connection, and judging from the comments his post received, it's a common problem. The solution? Perhaps mobile broadband.

Moconews pointed to Ericsson’s marketing chief Johan Bergendahl, who said mobile broadband will supplant wifi hotspots as the preferred method of on-the-go web access. "Hotspots at places like Starbucks are becoming the telephone boxes of the broadband era," he said. Moconews also pointed to a report that mobile broadband uptake is on the rise, and another report that city-wide wifi deployments are also growing -- perhaps indicating that wifi isn't ready to go away yet.

But let's just suppose that mobile broadband does become the dominant method of connecting to the Internet. Let's pretend that high cost, limited service territories, and speed barriers are overcome. What would the result be?

Perhaps the most interesting result would be that cellular voice networks would be forced to give way to VoIP, and services like Skype or JAJAH might be well positioned for a mobile broadband dominated world. With fast, reliable, and ubiquitous mobile broadband access, cellular voice plans would be irrelevant given cheaper VoIP options. We've already seen some voice-over-IP services target mobile users, such as JAJAH's iPhone optimized version.

That would be mean less costs for consumers -- no more free public wifi (presumably), but a slimmer cell phone bill and a single broadband connection that could cover mobile phone, web access, and landline telephone. Throw in a VoIP television service like Joost (which is starting to test live streaming this month) or Livestation, and the future might be one connection that covers all of your media and communication needs.

That simple future is a ways off, though. Mobile broadband isn't cheap, it isn't as fast as wired broadband, and it doesn't have the coverage necessary to be a truly reliable alternative to wifi hotspots (let alone voice networks). But is that future coming? Probably. HSPA+, which may arrive late this year, will offer speeds of up to 42 Mbps down and 11 Mbps up, which is significantly faster than my 20/2 Mbps cable line, and astronomically faster than the iPhone's current EDGE network (which is around 240 kbps down).

There are a lot of barriers to overcome -- not least of which, how to handle billing consumers when roaming on other networks and how to make the various competing standards play nice (or get networks to conform to a single standard). But once those issues are solved, fast, cheap, ubiquitous mobile broadband may be on the horizon.

When do you think mobile broadband will be fast enough, cheap enough, and ubiquitous enough to supplant wifi? Will that ever happen? If it does, what other side effects will we see? Let us know your thoughts in the comments below.

While Microsoft’s revenues are dispersed into many areas such as home software, enterprise software, entertainment, and Internet advertising, Google’s revenues currently depend solely on online advertising. While this may sound like a weakness, Google is currently well ahead of the competition and the barriers to entry in online ads are pretty high. Microsoft knows that, otherwise they wouldn’t be so ambitious about acquiring Yahoo!

But what Microsoft can do here is to change the rules of the online advertising game by making it more open in general and less profitable for Google. There have been many companies who have claimed to make advertising more open; OpenAds and RightMedia (which is now owned by Yahoo!) are just a couple of them. But what these companies actually did was make the advertising process more transparent. Advertising is still under the control of networks which manage literally everything; publishers, advertisers, parameters, matching algorithms, etc.

However if one can split up this "network atom" and divide it into more efficient parts, things will be very different. See the graph below:

In an open advertising model, the inventory silo, placement silo, and parameter silo are controlled by many different organizations, which can interact with each other and create advertising mashups. Today Google provides these silos under the following services:

• Inventory Silo: AdWords
• Placement Silo: AdSense
• Parameter Silo: PageRank, Google Analytics, Gmail, FeedBurner etc

The most crucial part of the advertising network is the inventory silo. Therefore, in order to make the open advertising model a reality, a company (presumably Microsoft) would have to jumpstart things by opening up its inventory silo - so that others could use that inventory to create new applications for placement, parameters, etc. What would happen is:

1. Companies would give away parameters that can be used with the open inventory. In return, they generate revenue when their parameters are used to place ads.
2. Publishers would no longer be stuck with "contextually relevant" ads, but could instead use any of these parameters or mash them up for optimal results on their pages.
3. New 3^rd party companies would emerge and they would make the whole process easier and more efficient for everyone involved.

The graph below summarizes the ecosystem I’ve just described:

With this model, a web page that consists solely of a Flash game is not stuck with Google’s "contextually relevant" option. It can mash up a bunch of parameters, or get help from 3^rd parties and choose the best option for itself.

As shown above, this creates a whole new economy for parameter providers. It also opens new doors for 3^rd parties for matching inventory with providers, mashing them up, analyzing and finding the best solutions for advertisers and publishers. It would also enable advertisers to make bids by filling out some XML files and allow the best ad to be displayed on particular page in the best way for a particular visitor.

Gain for Everyone

This model would not only eliminate the dominance of a single network and create a whole new economy for a lot of players, but also it could prevent a possible bust in the Internet industry.

Today’s online advertising is far too linear. Whoever clicks gives the same amount of money to the publisher and the network. That is, under the current model, the click of a person with limited purchase power is worth the same as a click from Bill Gates.

But this unfair model is not sustainable. Because what it does is to shift money from real production to vaporware. The economy can only get better if clicks on ads produce real results. And that can only be established in an open model in which everyone participates, and all parameters are run in a fully competitive and flexible environment.

Conclusion

Yahoo! is too risky for Microsoft because of the size of the deal and possible inefficiencies. So my advice to Microsoft is to let Yahoo! remain as the online media mogul. Let Google be the search giant. Instead, focus on shaking up the advertising industry by pushing it toward a democratized structure, make it work for everyone, and weaken Google's business model before it takes a bite from your Office and Windows revenues.

Then yesterday, Fried announced two more new Backpack features: messages and newsroom. Messages is just what it sounds like, a message board where users of the same Backpack project can talk to one another. While newsroom is an activity feed for the Backpack. For any 37Signals fan these features should sound familiar because they already exist in one of the company's other popular applications, Basecamp. Basecamp is a great groupware tool that we rely on daily to manage our activities here at ReadWriteWeb. Backpack is starting to feeling a lot like Basecamp in a different skin.

With the addition of messages and an activity feed, both apps now sport more or less the exact same feature set. Both have lists, messages, files, and writeboards. The main difference is in the way those things are displayed and how much control users have over them. While Basecamp breaks everything out into separate pieces, Backpack combines them all on a single page in any configuration you want. Both approaches have their merits, but is it necessary that they exist as separate apps?

The implications of the upcoming changes haven't been lost on users. "Cool. So Backpack is the new Basecamp with a better calendar," wrote user Jim on the 37Signals blog. "Does anyone else now feel like Backpack will have too much," chimed in Tim. "With the announcement of these new features (which are great), the difference b/w Basecamp & Backpack is starting to blur."

When others echoed the sentiment, Fried responded. "Basecamp is your project management tool, Backpack is your company intranet," he wrote. "Basecamp and Backpack are entirely different products for different purposes. We use both for very different things."

But while Basecamp and Backpack still have some major differences when it comes to things like permissions handling, which drastically effects the use cases for each, they do now share most of the same features. That brings me back to the question about whether 37Signals has lost focus with the app.

When I saw Jason Fried speak at the BIF-3 Collaborative Innovation Summit last fall, Fried was asked by the Wall Street Journal's Walt Mossberg, how do you avoid feature creep? According to Fried, the key is the ability and willingness to say no. "You have to be a hard ass," he told Mossberg. So with Backpack going from a simple, single person organizational tool to what Fried now describes as a "company intranet," has 37Signals lost focus and succumbed to feature creep? Are they not eating their own dogfood, so to speak?

The answer is a firm maybe. When talking about feature selection in their popular book about their software development methods Getting Real, the company writes that when building a web application you should start with just the single core feature. Say "no" to all the others. But that doesn't mean to you always have to say no to new features. "Start off with a lean, smart app and let it gain traction," they write. "Then you can start to add to the solid foundation you've built."

That is more or less what they've done with Backpack. Started lean, let is get traction, then expanded it to where they thought it should go. But as an application Backpack has clearly evolved beyond a simple organizational tool to something that is starting to resemble their Basecamp application (which we, incidentally, already basically use as our intranet). Whether they've fallen prey to feature creep will really be measured by the response of their users. If people remain confused over the difference between Backpack and its cousin Basecamp, then the answer is yes, somewhere they stopped following their own advice and said "yes" to one too many feature that was better suited to one of their other applications. But if people embrace both apps as complimentary offerings, as Fried clearly hopes they will, then they've made the right decisions.

37Signals offers a range of applications, from simple, single-function apps like Ta-Da Lists (to-do lists), Writeboard (collaborative word processor), and Campfire (group chat) to more complicated apps like Basecamp (project management) and Highrise (group contact manager).

In the past, 37Signals has often integrated its simpler products into its more complicated offerings. Basecamp has Campfire chats, Writeboard shared writing spaces, and to-do lists clearly based on Ta-Da, for example. Along with their slow shift toward full OpenID support, this points to the potential for a modular, create-your-own app system from 37Signals, where users could pick and choose which of the company's applications to install. That's completely speculation on our part, but we really hope that's the direction 37Signals is headed.

What do you think? Has 37Signals lost focus with Backpack? Or do you still see differences distinct enough between Basecamp and Backpack that you could see yourself using both? Do you wish the functionality of both applications was merged into a single app? Let us know in the comments below.