Once upon a time there was a product called WordPerfect Office. It was DOS based and it contained a basic menuing system and one other very important feature, and that was E-Mail. For those of you how are not intimately familiar with the history of the product here is the Wikipedia entry: Novell GroupWise. Give it a read for some context.
So whats my point... GroupWise is losing the Enterprise E-mail battle on several fronts:
The biggest problem that I see is that there seem to be two very distinct camps at Novell these days. We have the people who develop for windows who are doing a great job. We have the people who develop for *nix and they to seem to be doing a great job; however, the two of the must be in separate universes or at least galaxies.
The GroupWise Clients are being developed in two very different languages, and that is the problem. One group is using Java, and Java at the desktop / UI level just does not work, sorry all you Java fans but it just does not, its slow, the UI elements are horribly implemented. The other is using C ( I like to call it C-- ) but hey, I am a programmer and I like just plain C or Object Pascal, but that is beside the point.
On one hand we have a GW Client using Native windows components, on the other we have a mashup of Swing, Awt, god alone knows, because I don't and I don't really care, but the thing I really care about is the problems it causes with GW being accepted in mixed OS environments. The Windows Client always leads the Java Client. Why IS that so? I don't know, but it is detrimental to GW's future, this I DO know. What must be done is to seperate the underlying engine from the UI and then build the UI's for each OS that GW wants to compete in. The OSX Native UI is simply stunningly beautiful, Linux's Gnome or KDE interface is solid as is Windows, build to their individual strengths, just do it.
Integration. I have said in replies to Dean Lythgo's blog entries that we must completely and utterly abandon OLE, it was a broken technology before it even hit the street. It has to go and it has to go NOW! If any of you have ever tried to develop to the GW API, you know the pain, you know the frustration. MAPI is over, MAPI is dead, not even MS uses it any more. ALL major 3rd Party vendors that integrate with Outlook / Exchange do with either VBA or .Net, that is the reality, nothing more nothing less. We need the API's to be concrete, with bindings for C, Maybe C and C# ( aka .Net ) and thats it because the rest of them, Java, Python, Perl, yadda yadda yadda, will ALL use those bindings. The API calls MUST be simplified to the point of being able to make single call to insert AND reference a Gw Message Type or Document in the Libraries.
Battles are won in many ways, but at the end of the day, the Army with all of its various Units MUST come under a unified command structure. GroupWise needs a "General Patton". An individual with guts to get out in the field and tell the troops how to fight this battle and kick their butts up the hill if required. That individual must in no uncertain terms give the battle orders and see that they are followed.