The road to Hannover

by Volker Weber

This is not going to be a popular post, but I will walk out on a limb here. I have watched the progress that IBM has made on the Hannover design, from the first Photoshop work to the latest pre-beta. And I am very, very disappointed. I felt a certain frustration for the first time in Karlsruhe in the lab. There was a lot of understanding about the things that needed to be done, while there was also a sense that they won't be done by the time Hannover ships. So here is my prediction.

Hannover will be Notes, with new icons, and a freight train of Java in tow.

Mark my words. It is better to adjust your expectations today than half a year from now.

Of course this is a matter of fixed timeframe, budget and scope. Pick any two. It is an incredible task to re-engineer Notes on top of the Workplace client technology, at the same time for two (and later three) platforms. But the interface won't be radically better, I am afraid.

The two biggest innovations, activities and composite applications as presented by IBM, most likely won't be working disconnected. Disconnected use however is the stronghold of Notes. There simply is no competition. If your applications only work while you are connected, the field is wide open.

Not all is lost. Mary Beth is listening. Read her blog. Write comments. You can shape the future of Notes.

Comments

A fair opinion I would say, I commented after the initial announcement in Hannover that the screenshots didn't look so different to the current Notes, once you'd removed menus, status bars and smarticons. You can do that with the current Notes client...but it doesn't make it more usable!

That being said, the alleged ease of deployment of the new client will be a breath of fresh air for many. Local installations of the application feel a bit dated now and easy controlled updating is still not a reality. I expect Hannover to address much of this, as well as true cross-platform support. Only the server is truly platform agnostic right now.

To close, if IBM have overpromised and end up under-delivering, then they'll be right up there with the competition. Let's just hope it ships on time.

Ben Rose, 2006-07-12

I am surprised you are worried about a post being unpopular:-).

I appreciate our frankness and candor. I have been taking a "wait and let's see what they deliver" attitude, but given the problems I experienced with the Sametime 7.5 on Eclipse Beta (yes it was beta so I am willing to cut a little sack), my expectations have been lowered.

J2EE does not make something better by definition and IBM is risking a lot taking this approach.

Christopher Byrne, 2006-07-12

Thanks for the information.

Thomas Adrian, 2006-07-12

"Hannover will be Notes, with new icons, and a freight train of Java in tow."

That's a lot of writing for a T-shirt...will need to be big...XXL? ;O)

Ben Rose, 2006-07-12

The Eclipse platform offers more than just good looks, f.x. extinsibility and interoperability, plus Notes will finally play nice with other applications in the enterprise.

Villi Helgason, 2006-07-12

Maybe the time of Notes has just come? Nothing lasts forever ...

Cem basman, 2006-07-12

Well you are right about the Java part and new icons. The key elements you are missing are the entire Eclipse foundation, a new programming model with Composite Applications, Activity Based computing, off line Web Applications, JSR168 and WSRP integration, an entire web services stack and a lot more.

You can get a lot of what is going on with Hannover by reading some of the IBM bloggers on what is cooking but Eclipse.org is a great place to start considering we push a lot of stuff down to the Eclipse platform with our contributions.

Bob Balfe, 2006-07-12

That should be "slack", not "sack" as in my original reply...

Christopher Byrne, 2006-07-12

Bob, I have been watching IBM for 15 years now. I was there, when Eclipse was announced, and I am aware of all the things you have mentioned. If you read my post, you will see that I mentioned activities and composite applications. The "key elements" are the freight train.

Volker Weber, 2006-07-12

Cem, some things actually do. And Notes' time has been called since 1995. At least.

Volker Weber, 2006-07-12

Eclipse is also the fastest growing eco system in the development community. If Notes can hook up to that, then good marketing to IBM, and thanks for all the tools.

Villi Helgason, 2006-07-13

@Vowe,

What do you mean by "freight train". I would say that freight trains are very hard to stop do to their power. Do you mean a "train wreck"?

Bruce Elgort, 2006-07-13

Hi Volker,

Something I find interesting every time J2EE and eclipse are mentioned; is that people always talk about "Eclipse as the great development tool" "Eclipse as the 'choice of developers'" "Eclipse full of wonderful tools for 'The developers'"

I never hear about "Eclipse used by 100 million Users around the world"? Maybe I've been living under a rock or something, maybe I'm missing something?

It sounds to me like the panacea for the "problems with the notes interface" is to give the users a developement tool and they'll love the flexibility???

Am I missing something or is Hannover just another marketing opportunity for Microsoft? I'm willing to be pleasantly surprised, but IBM has been giving them marketing oportunities for a while now.

After all, the only Web services I am seeing right now are "on the web". Do we really need them on our Client? As I understood it web services were the 'be all and end all' of B2B/B2C and they are normally server side, not local. OK Notes is the offline king, however offline is for travelling people now (online here and now is for the desk bound) and although that market is growing, so is the roaming connectivity market.

So I wonder, just when Microsoft is going to ever more functional menus and prettier icons with more functionality, what IBM thinks is going to make my life more functional without/reduced Smarticons or without/reduced menus.

I am willing to be underwhelmed. However I would very much like to be wowed. Bitter experience with 5/6 and 7 shows that being wowed is not in the plan. Pity, I still have my old 1995 R3 client and server with an old L-Notesl mail file on it; I also have an old R4.1 client setup. I was wowed. Something along the same lines would be nice.

I shaln't wait up.....

Oh and has anyone looked at the Yahoo mail offerings recently? If I was running a 1,000 user company, I'd register for a Yahoo account and hire a geek for six months to set it all up. At $29 per month for the Entire 2 TB service (plus website), I could afford that.

The future is not a fight with Microsoft, unless IBM want's to get into the "Live" arena. I just wonder where another huge development suite sits in the world of mail and collaboration. After all, Google and Yahoo are not just aiming at Microsoft when IBM owns more than half the world that they are aiming at.

I think the phrase is "collateral damage" :-)

Neil Thomas, 2006-07-13

Neil, you have a lot of very different thoughts.

Regarding Eclipse: There is some confusion about the difference between the Eclipse framework, the development environment and the rich client platfom (RCP). We are talking RCP here, and yes, this not yet used my millions of people, but IBM can make it happen.

Regaring web services: This is a misnomer. It does not have to do anything with the web as in www. Web services require a client and a server. It's one of the cornerstones in an architecture that IBM (and other players) peddle as SOA. If you want to build applications under this architecture, then your client better make it easy. This is what Bob Balfe means with web services stack.

Regarding the user interface: We know that there are no formal standards. There once was , but the interface standard today seems to be the latest version of Microsoft Office. That is going to be the greatest challenge for Notes, since by the time Hannover ships, the latest Microsoft Office will be 2007, and no matter what happens, Hannover isn't going to look like it, even by a very long shot.

Volker Weber, 2006-07-13

In all fairness, not even Microsoft have got a consistent interface across the whole Office 2007 suite - some apps are using the new 'ribbon' menu system, and some still have the traditional..errr...'menu' menu system :) Outlook, in particular, has stuck with menus for the main interface - you only see the ribbon in a new memo, so I don't agree that there is going to be that great a divergence between the two interfaces, and I agree with MS's apparent stance that the ribbon system isn't suitable for all situations.

In addition (and I know these are just mock-ups, but still...) the screenshots of the productivity tools within Hannover did seem to be taking some cues from the ribbon concept - not identical, but similar.

As for Hannover - I'll reserve judgement for now, although I'm quietly optimistic that the new client platform will give us most of what we need, but possibly not everything we want. In saying that, there's an IBM guy sitting not 3 meters away from me right now, and it does look like he's got an early version of Hannover on his laptop - I might just wander over and have a snoop around :)

Matt Buchanan, 2006-07-13

IBM is just repeating others' mistakes.

Stefan Tilkov, 2006-07-13

Stefan, thx for the link. Made me giggle (and nod) a lot...

Frank Dröge, 2006-07-13

Interesting conversation.

Stefan, I agree with what "Joel on Software" says about rewriting from scratch. That's why we're not doing that. Not only do you tend to end up with something worse and buggy, but it also just takes too long.

We are doing exactly what Joel suggests, keeping much, and carefully rewriting some. Trying to avoid the temptation to rewrite code just because the old code is awful and hard to understand or not in Java :-) We are rewriting some code or writing new code in an attempt to bring new functionality, improved user experience, and a new openness to the platform. It is a delicate balancing act. Whether we'll hit the exact right mix, it's hard to say. But that's certainly where I'm trying to drive the team.

I'll stay (mostly) out of the fray of things like whether Eclipse/RCP can really be a true high quality end user framework. Yes, we are really pushing the platform and driving changes into the platform in order to achive this and it is most certainly a challenge. IMHO, it has been holding up pretty well to our challenge.

Final point. To Volker's original post. Yes, unfortunately, it looks like Activities won't be fully functionally offline in initial release (though will be readable via RSS/Atom so can use RSS reader for offline experience). That's just the nature of time deadlines. That of course could change, but either way, working to rectify that ASAP without waiting for the next major release. Composite apps, on the other hand, will be available offline if the individual components work offline (such as a Notes DB with a local replica).

Jeff Eisen
Lotus Notes (& Hannover) Chief Architect

Jeff Eisen, 2006-07-13

Yes I suppose that I am a little agressive about things like "IDE" etc, however have a look at this http://wiki.eclipse.org/index.php/Rich_Client_Platform

The interesting fact is that they are stressing the fact that you "can" produce other applications than an IDE. However what I have to ask is "Is a platform designed for a programming workbench and IDE the ideal platform for a rich client"? Who knows, but I don't hold out that much hope, the requirements of a developer and a standard user are completely different.

I'm not particulary confused about web services. The model is an XML open exchange. What I want to know is why we should throw out a fast, lean, rich and extremely secure exchange method of client server for a server-server back end technology that was designed from the outset to connect disparate technologies that dont' talk?

Just exactly what web services would the client need to access that don't already exist on a server and could be accessed by a native protocol? Also if the user was disconnected, how would S/he use the web services model anyway and are we saying that the disconnected client would need to run a local server just to access local apps?

Also why would you use web services when you know both sides of the architecture? After all it's designed to connect disparate systems where you don't know the other end of the equation. It just seems so completely wasteful to me, all in the name of "expediency". I stand by my statement, web services in my opinion belong on the server side when talking to other server side apps and not on the client; in any shape or form. Otherwise bloatware would become megaware as the client tried to be everything to everybody.

On the interface side, I was dismayed by the early shots of Hannover. The very worst parts of Outlook. Screen hogging colums with crammed Iframes with unreadable data and scrollbars all over the place.

Personally I, like many others, switch off most of the Outlook "screen clutter" at first contact and return to the basic interface. If I could rip out and burn the stupid ribbon thing I would but alas it is not to be. Also the stupid reading pane which has NO global setting to turn it off is absolutely maddening.

So I suppose what I'm saying is "Please don't repeat these mistakes". I haven't yet met a user with a 26" laptop or desktop screen so how exactly will they read their mail view and see the massive preview to the side (or do people just like scrolling?)???

I have hated the bookmarks since they were introduced in Notes. I can turn that stupid Outlook bar off in Outlook, I am stuck with it in Notes. from what I can see I am going to be stuck with it forever as the Hannover team seem to like it.

So as you can see, I may be in a minority ( in some cases), but I'm prepared to be seriously underwhelmed. I expect to recieve a slow, buggy, massive memory hog which forces me to work the way the Hannover team wants me to. I hope to be very pleasantly surprised. I'm prepared to get more of the last 7 years as I don't see any indication of a change in attitude in the design team (vis a vis my rant on F5 on the Hannover blog).

So why do I support and promote this application I patently hate so much? Well it's just the best tool for what it does. End of message. Hopefully that won't change but web services on the client isn't going to help at all.

Neil Thomas, 2006-07-13

Neil, maybe you rant too much. ;-) That does not foster discussion.

Regarding Eclipse RCP: Eclipse is not designed to be an IDE. It is designed to be a framework, that you can use to build IDEs, but also other applications. The beauty is the plugin model which lets you build anything you want. Plus, it is available natively on five platforms, of which IBM will use three: Linux, Mac, Windows.

Regarding learning from Outlook: I think there is a lot to learn. And the three column view you despise is not mandatory. You can switch to the traditional two column, three pane view. I don't use three columns in Mail or NetNewsWire, but I could. They work very well in the newer wide screen notebooks and desktops. Lotus can also learn from the "categorized by days" view in Outlook, which reduces the number of columns used in the view. There is way too much cruft for my taste in the Inbox.

Regarding web services: When you talk about "native" protocol, I may be talking about "proprietary". Web services let you access disparate systems. For instance blogging on a platform, that does not talk NRPC. Like vowe.net.

Volker Weber, 2006-07-13

Jeff, with regards to composite apps consisting of replicated Notes databases, that is certainly true. But Lotus so far has positioned composite apps as a way to integrate Notes applications on the glas with non-Notes applications such as Siebel. I was referring to this scenario.

Volker Weber, 2006-07-13

Jeff, if this is true, great. I note, though, that from an end-user perspective, Notes/Domino seem to have stagnated in recent years, while everyone has been consoled with promises of something that may or may not end up to be better ...

Stefan Tilkov, 2006-07-13

Volker, I think there are many, many uses of composite applications that do not require applications or portlets that are not part of Notes. Yes, all the demos showcase a notes frame with a SAP frame and something else, but why couldnt that be three seperate Notes databases?

I do think saying "composite applications will not work locally" when the breaking point is that the external applications or portlets do not support local saving of data is pretty slippery.

John Head, 2006-07-14

John, I have made my wording a bit more precise, so you don't have to call it pretty slippery. It now says: "composite applications as presented by IBM". This should clarify.

I am also feeling inclined to include that activities won't show up in your calendar, rendering the ability to time them useless.

Volker Weber, 2006-07-14

John Head - that is an excellent point. The same goes for basic Eclipse based applications. Many developers do not componentize their re-usable view parts or portlets to be used in other contexts. With the introduction of composite applications you will see developers have a greater focus in these areas. So the next time you create that view or form in Notes you will think "can this be used by another application?".

Bob Balfe, 2006-07-17

Old vowe.net archive pages

I explain difficult concepts in simple ways. For free, and for money. Clue procurement and bullshit detection.

vowe

Paypal vowe