Transitioning to Office Labs

re: Transitioning to Office Labs

The IT Being — Wed, 20 Aug 2008 11:42:04 GMT

Just found the blog, and really liked it! You are doing great job here, and keep it up like that!

The IT Being

re: Transitioning to Office Labs

MadBabble — Mon, 21 Jul 2008 16:09:26 GMT

Thanks for tips

regards

Sven

re: Transitioning to Office Labs

Craig — Wed, 11 Jun 2008 23:23:03 GMT

I hope that your presentation today on TownSquare was successful. Will this project be opened up beyond just MSFT employee participation? If so, when? Is this a project that you plan to "Codeplex" (i.e. open source), or not?

re: Transitioning to Office Labs

Chris Pratley — Fri, 06 Jun 2008 19:29:33 GMT

Joe, your thinking is similar to what we are thinking, perhaps not in exact specifics but in spirit. That sort of "out there" or "redefining" work is what we are charting in Office Labs. There is always an element of "we tried that and it didn't take off", but we all know that sometimes just small changes can make an idea fly, so nothing we've tried before is really off the books. The trick is to find how to make it take off.

re: Transitioning to Office Labs

joewood — Fri, 06 Jun 2008 05:08:10 GMT

Thanks for the long response Chris.

I've been programming with the Office platform for a long time. In fact I was a big user of Word 2 for windows (I think it was called) and even wrote a OLE server or two (with and without the help of ATL). I also work in the finance industry - and yes we have those traders that would laugh at a table in Word being their risk spreadsheet.

My point about Excel / Word is that if the platform was more modular then the engine part of Excel could run behind a canvas and the trader wouldn't even know or care. Our risk management spreadsheets could become the monthly risk reports without having to publish / cut'n'paste or embed data between applications. And that risk spreadsheet would become a real distributed process.

Your point about OLE2 (wow, haven't typed 'OLE2' for a long time) - people didn't use it I think because of performance and reliability. It was the right idea with a bad implementation. The problem is that it looks like the Office team suffered a case of 'once bitten, twice shy'. I am hoping you guys are cooking up a more approachable solution to this that involves XML, XAML and something like DataTemplates. Maybe if Silverlight becomes a first class office citizen.

And then Powerpoint just becomes a way of viewing the document/canvas.

The key frustration with Outlook has to be that it is in the position to redefine email. For a lot of people outlook and exchange is the pivot point to their working life. At the moment everything is typless and linear. An email chain can represent an interaction with a resolution that could and should be publishable as an artifact. I see outlook and exchange as a system holding many different rich sets of data and interactions but only one real way of visualizing that data - a stream of messages in a grid. I know there's outlook forms, but again - how accessible is that technology.

I realize that the foundation for all of this is there: - Excel services, Sharepoint workflow and the potential for Live Mesh in a corporate environment is immense. And that most of what I'm saying is on the plan somewhere (OneNote was a great step forward) - it just seems that (from the outside at least) the bigger the Office suite and services become the slower the development gets.

re: Transitioning to Office Labs

Chris Pratley — Fri, 06 Jun 2008 04:16:59 GMT

Joe, thanks for the comment.

Abotu Office not seeming to have changed much. You can make a good case for that, and you can make a great counter case for it. 10 yrs ago there was no SharePoint, OneNote, InfoPath, Forms Server, Excel services. All the applications have picked up many new capabilties - you just have to switch back to Office97 and try that for awhile to notice them. OTOH, the core apps are still designed to do what they were designed to do, so certainly you could say they haven't changed.

I can't find where I said there should be "user demand" for a feature in this post, but it's a good point. I've written before about articulated needs (user demands), and unarticulated needs (needs people have but don't ask for - because they don't know they have them). The former tend to be smaller things ranging from bug fixes to feature add like say, a better way to compare two documents. The latter tend to be larger things ranging from huge new capabilties, products, or "the internet" which most of us didn't know we needed and couldn't have asked for before it appeared.

There's a saying that "if you don't listen to your customers someone else will". There's another saying that "if you listen to your customers too closely you'll miss the next big thing." They're both true of course. Listening to what customers ask for is important - you'll lose your business sooner or later if you don't. You also need to look beyond what they ask for or even would seem to like to be ready for the next great opportunity. Think of the horse and buggy manufacturers, listening to their customers ask for better bridles while the next guy over is building the automobile.

Why is Excel a separate application? Is it historical, or is there a good reason? Probably both. if you ask a finanical analyst if they want to use a table in Word that had Excel fucntionality for their work they'd laugh at you. But for simple calculations that sure would be handy. In fact 20yrs ago this idea of compound documents was all the rage - that's how Object Linking and Embedding (OLE) got started. Even today and ever since the early 90s you can paste an actual Excel speadsheet range of cells into Word or PowerPoint or wherever that supports OLE and have actual Excel there in your document. People don't do that too often, and there are several reasons for that ranging from discoverability to bugs to usefulness.

I do like the idea of a canvas that you can place all sorts of content on - that was part of the OLE dream too. But maybe today there are ways to make it work better. In fact we showed some ideas around that a month ago: http://www.youtube.com/watch?v=YPrfqdl55D0

Outlook and email is a very fertile ground - I agree. Outlook supports async operations such as sending receiving and caching mail for offline use already. I wonder what your ideas are in this area?

re: Transitioning to Office Labs

Joe — Fri, 06 Jun 2008 01:46:16 GMT

I agree with your point about chasing shiny objects.

But I have to say that (from the outside view) Office hasn't changed very much in the years I've been using it.

You say there should be user demand for a new feature - but the average user probably just doesn't expect to work in a different way to they do now.

For example, why is Excel a separate application - why can't an Excel sheet be a table in a document (or a canvas). In fact, why is most of office based around printed paper - when these days most documents never make it beyond the screen.

What about email and outlook? When will it support asynchronous I/O operations? How about modelling a conversation as something other than a linear email body. How about exposing that conversation as data that can be published as a conversation - I could go on....