LotusLive Irony...

One little remarked piece of information about LotusLive (dont look to me for information on this - I'm sure the little yellow bubble is hopping with information) - which I think I concluded is this:

LotusLive is the first time that Lotus have actually implemented a multi-client hosted infrastructure - for themselves.

Yeah, IBM Global Services have implemented lots of single-customer solutions (dont for a second assume that this information was made available to anyone else within IBM, believe me). And of course huge Lotus hosting partners (Chris Miller at Connectria, the chaps at Prominic in Chicago, Applicable in Bristol here in the UK to name but a few) have always had the ability (and the success) to host multi-client solutions.

But.

This is the first time (IMHO) that Lotus have done it themselves. Not IGS. Not Consulting. Not IBM themselves.

So - what can we expect ?

  • World class support, obviously. IBM do that and do it well. You might end up talking to an underpant gnome halfway around the world, but you will get support, and lots of it.
  • Hosting in IBM data centers. So between all the mouldering black racks of iSeries, pSeries and zOS boxes, lots of happy little IBM servers will be whirring away. No idea where these hosting centers will be, of course, but why should you care.
  • A somewhat fragile onboarding and offboarding process. They've just been working on this for a little while, and will lack - at least in v1- any of the complex experience-led onboarding that the hosting partners will have developed over the last 10 years. And knowing IBM in the way that I do, expect about 80% of it to be well thought out, reliable, and robust. Expect about 10% of it to be in flux, and expect to install an OS/2 box, a 3270 emulator and punch card device to get the last 10% of it running. Hey, it wouldnt be IBM without some old fashioned magic, eh?
  • Flexibility. Thats a good word. So lets imagine the vast committees within IBM develop a new product. Its not the skunkworks-led team that did the original PC. This will be a whole herd of lumbering committee animals, each defending its turf at the great watering hole that is the product development meeting. So there will be some extraordinarily fantastic bits, some normal bits, and some bits that could only have been invented by an animal on class-A drugs with four back legs. It'll be interesting to peer through the lies vapour of the marketing materials to see what is what.
    I think what I'm getting at is that if your after an IBM hosted environment, with IBM world-class support and hardware, for a fixed price, then perhaps lowering your expectations in terms of things like flexibilty might be an idea. Hey, dont let me stop you *asking* for impossible things - but dont hold your breath.
    IBM can be as flexible as an iron bar a mile thick when it wants to be. And sometimes its less flexibile. Just ask your local Business Partner about how flexible IBM is. Lets see where on the hardness scale this particular attribute fits
  • Upgrading. Congratulations, you've just outsouced the mundane bits to LotusLive. And you know - there's a few odd things happening on that particular version. When will it get upgraded ? Well. You have no control over that anymore. Remember? That was hard to do ?
    Your CEO is screaming that his latest BlackBerry's not syncing B/OS v6 contacts with his Lotus Notes 8.5.2 mailbox (for instance). Your level of control over this will be the same as your level of control over an outsourced envrionment: The square root of bugger all.
    However, you *have* managed to get rid of all those whiny admins who just litter the place. Good for you. Go to the gun-section of Buy-n-Large and get that Kevlar vest for the next IT review board.
  • Enterprise scalabilty and roll-out. Yes. Finally, since Lotus will be running an Enterprise system at last (and not just basically ignoring its biggest customer - IBM's - demands) - we will finally get a fix for all those really irritating little things that make Lotus Notes 'challenging' to roll out to Enterprise. Note that I didnt say 'Difficult', nor 'Impossible' (like Exchange). You just have to have a good crew running it.
    Features like:
    • Wouldnt it be cool for the users not to be bugged when their mail client fails over to the cluster mate ? Well, that apparently is fixed (I've not seen 8.5.2 yet, so I'm operating in the same level of hype and enthusiastic speculation as you).
    • Wouldnt it be cool for your notes client to know all the servers you normally connect to (as it does), do a quick - ohh - network latency ping - as it can (The C-API call has been there since v5!) and then open the *nearest* (That is, the server with the lowest network latency) replica ? Wouldnt that really make your road-warriors life cool ? Especially if it worked on the replica page too ? With failover ?
      Guess what. Thats been requested since forever. Here's hoping that since Lotus are now having to support their own customers, this quick fix makes the triage list.
    • Imagine if the password box on your notes client had a checkbox to 'change password'. So when you entered your password, you were then immediately prompted for a new one. Wouldnt that be cool. Instead of:
      1. Entering your password
      2. Hitting file, security, user security
      3. Entering your password again
      4. Clicking on the 'change password' prompt
      5. Entering your password for the THIRD time
      6. Entering two new passwords
    • Want to actually be able to put icons on the Workspace? Or change the server that they point to? Wouldnt it be cool for IBM just to buy PanAgenda and just make their product part of the toolset?
  • On the plus side, we all know Lotus Domino scales extraordinarily well, IBM host data centers extremely well, and IBM know how to put a support organisation together. No. Really.
  • Not like the upstarts from Redmond who make server software with the strength of a chocolate fireguard. Or the chaps at Google - who are very brilliant - but dont seem to have the business sense nor experience to actually put together a product a serious enterprise might actually want to buy.

    In fact, I see the three vapour-salesmen like car salesmen:
    • IBM to a certain extent is like Volvo. "Boxy but good". Your not going to scare the kids at the lights with one, but you will want to put your kids in it and drive across some of the worst roads in the world (North America) in it.
    • Microsoft, I see more as a GM sort of deal. You get what you wanted at your price point. You'll have to replace it every other year, and keep some gaffer tape around, because you know at your pricepoint and complexity level, things will just keep falling off. No, really.
      Hard to love.
    • Google. I think they think they're Lexus. They're certainly smart enough to be Lexus. And we - by the time we kick the tyres, think more 'North Korea'. But they're only a heartbeat away from their startup position. All very smart but no business intelligence. And do we really want to give up our privacy for free product? Dunno.
  • I think it'll be great for someone in development finally to have someone within Lotus screaming at him for a fix.

    We business partners have had this for the last 15 years, and its done us the world of good in terms of blood pressure and stress level to absorb this level of pure hatred from end-users over bug fixes, and then be basically ignored when we raise the PMR's.(*cough* 64-bit Lotusscript failures in Domino)

Right. So. LotusLive. Cloud meets enterprise. You can host all your boring mailbox gubbins, flaky sametime infrastructures, cool and froody other stuff, and yet still keep your apps servers with all your funky (old, unstable, unsupported) applications - and your users wont know any better.You'll save a ton of cash constantly upgrading your old servers to do this, and all the weekend work that your admins used to charge you for.

And - ignoring my cynicism here - IBM will support you. No. Really. They do. They wont rip+replace you like the Borg do, they wont just completely drop the product as google might do.

Its a brave step. One that I'd advise anyone making to consider properly for some time. Given the right set of starting conditions and assumptions, it'll really help your enterprise.