Fresh Content on SharePointJoel.com SharePoint Ads
Subscribe in a reader
I hear this question at least at every conference, and I'm sure there is no right answer. Of course I have an opinion. I have lots of opinions, and I am biased. So I don't see this as a chicken or egg debate. There are some very clear reasons why you'd want to not hold up your server deployment with planning for Office 2007 or deployment. Neither has a dependency, with some caveats or considerations listed below.
Straight out, if you already have some flavor of Office on the client, or even not, you need not hold up deployment of WSS or SharePoint Server. If you're debating WSS 2.0 or WSS 3.0 or SPS 2003 or MOSS 2007, that is a no brainer. You have to choose the latest and greatest. There are a hundred reasons on that one, but I'm not going to go into that here.
The key piece of information for a successful client and server integration experience is to understand what the "light up" experience means. The fair, good, better, best whitepaper, which goes into the features of office and what you get with the various flavors goes into compatibility and features quite a bit. The true light up experience is best explained as features are simply not there and won't present issues. You simply won't see the data information panel in Office 2003 clients, you won't see the ability to publish your Excel workbooks in Excel 2003, you won't see the Offline and RSS in Outlook 2003. So when you do install, boom, lights up... You now have the richest experience for publishing, and interacting with the full featured experience.
Below is a quick list of things to consider. The thing you'll notice they all have in common is the power user/business analyst/designer. If you segment your users in a way that the author/contributor/power user gets more tools and there is a way of updating those clients first. They will benefit from being upgraded first. Rolling out the broader functionality to the rest of the users does make sense. Indeed sharing the wealth to increase productivity obviously makes sense, but what I'm suggesting here is the group of users that know what features to take advantage of will *really* see a power boost and quickly become more productive.
1. The most important aspect in terms of deployment to understand is the design aspect. FrontPage 98, 2000, 2002/XP 2003, etc... Do NOT work with WSS 3.0 or SharePoint Server 2007 (MOSS 2007). Your users will get errors if they try to open pages with the any version of FrontPage. With this in mind, if you plan to have the set of web designers, those building your masterpages, or managing workflow should get the SharePoint Designer 2007 for use with editing the pages, creating workflows, creating data forms, and creating master pages.
2. Infopath 2003 can be used to create forms that are published in forms libraries, but it is *Required* to have Infopath 2007 if you want to build web based forms to leverage the capabilities in SharePoint Server 2007 Enterprise edition. If you end up with mixed Infopath environment you'll like the backward compatibility wizard for ensuring backward compatibility or ensuring the form is web capable for zero client footprint.
3. If you're planning to do anything with Excel Services (another enterprise edition feature set), you will want to have those that are publishing the excel workbooks using Excel 2007, but this isn't usually the first thing you do.
By going client first, you gain the "best" experience from a client perspective, but how much did your users suffer by not having a collaborative platform, or how much did IT suffer by not being able to either upgrade or take advantage of the new features like the recycle bin.
If you're looking for things that simply are broken, or simply don't work, you can hunt and find them, but they are really difficult to find in my experience. Office Web Components on the client and server may be one scenario where you need to figure out if it's something you need to support and working through that one.
Other than some functionality your users may not even be aware of, going with server first and rolling out the client either after or concurrently has it's benefits. In recent conversations and in a recent thread I was told "The ONLY reason I can justify moving to Office 2007 desktop is the business value of the new Office in the context of how it leverages productivity features in MOSS." This from a CIO trying justify deployment of the office client features. I'm sure I could give him a bunch of reasons with the client on it's own, but I do have to agree when you have SharePoint Server or WSS, the RIO and REJ get's much, much easier.
My top 3 client features that I actually use are 1.) Speed of search in Outlook 2007 2.) Web 2.0 Features in Office - publish blogs from word, RSS feeds Offline SharePoint lists and Offline SharePoint files in Outlook. 3) File size and extensibility in new file formats. I didn't mention the ribbon in my top 3, but it's something I use every hour I'm on the computer. I didn't realize how much I use it, until I was looking at Office 2003 the other day. It was a shock to the system. I get a lot of measurable increased productivity and time savings out of powerpoint 2007 and Outlook 2007. It's interesting to note that you CAN upgrade Outlook 2007 to take advantage of the new productivity without upgrading the rest of the Office client. OneNote 2007 does rock! You need to check it out if you haven't.
I'm not a massive Vista fan yet, I use it quite a bit. I myself and a RAM hog, so if the machine doesn't have at least 2GB (4GB preferred), it's a no starter. My top 3 there are 1) Optimized network stack (faster to save and copy) 2) Save as dialog for saving to "My SharePoint Sites" when used with Office. 3) File UI and previews. Can't say I see my desktop, so I don't use gadgets, and I still use alt tab vs. windows tab, guess it's habit. I most definitely would not hold up a SharePoint deployment until all users had Vista.
So in a nutshell since this is dragging on. Don't wait till you deploy your clients with Office. You can do them both at the same time if you want, but don't hold up the SharePoint deployment or upgrade. You should consider SharePoint Designer deployment in your web designer community and reducing the FrontPage footprint. It is fine to have co-existince in the server world. At Microsoft we still have WSS 2.0 and SPS 2003 sites and farms that exist with specific business units that will want to be upgraded, but with their applications it will require some hand holding. The mass hosted environment and 3 TB of data was upgraded prior to RTM. Mixed clients - OK. Mixed Servers - OK. FrontPage 2003 with WSS 3.0/SharePoint Server 2007 - BAD. SharePoint Designer 2007 with Office 2003 client apps with WSS 3.0/SharePoint Server 2007 - OK!
Note: If you are using really old versions of office like XP and before and you're reading this. Don't use this as an excuse to again hold off on upgrading your clients. Windows 95 and Office 97 were amazing products in their day, but welcome to the 21st century.
As usual you have good insights about this stuff!
My personal experience is that when you give information workers the new office suite, they are frightened at first..So when you are giving them training about office 2007 (which you should anyway), involve the integration part of MOSS. Get them to save their documents on a document library, get them to write a blogpost from their Word 2007...
This way, they learn the new way of work they will use for the next X years, and it takes away all the scary things of getting to know a new platform.
This does require to have MOSS implemented first, but this can be done with a bunch of key users who are needed to create the portal..
Great Suggestion. I think users like to get warm fuzzies about client upgrades AND server upgrades. By simply having a brown bag lunch, a 2 hr presentation with some hands on and very demo heavy experience, the users will be much quicker to adopt, and quicker to transition. I'm all for creating a sandbox or "demo" environment that is simply super low quota and branded with data that says, demo purposes only with NO SLA. That way users know the environment isn't "production" and they can experiment and try things out.