Total Immersion

Because its having a distorting effect on the rest of the industry. I’m afraid I don’t buy the argument (nor would I have to, but you get what I mean) that all software should be free, as Richard Stallmann would so dearly like. Any more than I would agree that all music should be free, or indeed that my plumber should pop round tomorrow and fix the dripping bath tap. It’s a laudable goal of course, as is world peace and the nirvanic state where everyone just gets on. But its just not going to happen, because various elements of human nature - good and bad - won’t let it.

To me, and unlike what the “try the latest distro” Linux User cover disk would suggest, open source is far more about commoditisation than diversification. I find it hard to believe that there is a place for new operating systems which try to compete on features - as long as we build systems around the Von Neumann architecture, there have been operating system constructs around since 1969 (that’s Unix, folks) and indeed before to support them. I don’t want to ignore z/OS on the mainframe - but let’s remember its precisely because the engineers of a few decades ago got so much right that they are so full of themselves now. Windows is also fine - its an OS which cuts the mustard, both on the desktop and on the server. Bt half the reason I believe that Vista tripped up, was that it did not offer anything sufficiently compelling to the majority, even if its security and manageability features far surpassed those of Windows XP. Didn’t anybody tell Microsoft how hard it is to make a business case for security and manageability?

So, open source offers a commoditisation route: if something is algorithmically so straightforward now, and its a question of evolving it in line with the hardware, then open source offers the answer. No point in paying for something that is already done. There are several advantages: the source is openly readable, which makes it potentially more future safe than anything proprietary. Development continues, in an evolutionary manner, and is funded and resourced across the community, which also provides a proactive support base. Its a model which gives us the LAMP stack - thats Linux, Apache, MySQL and which every programming language you can think of that starts with P. And there is money to be made - but out of services, not so much the software licensing.

And here’s the kicker. When it was realised that the real money was to be made out of services, that’s what had the biggest impact on the rest of the industry. Red Hat started to rake it in due to the fact that corporations wanted to know they had the same levels of support as with their proprietary application base - a fact which triggered Microsoft’s ill-advised “Get the Facts” campaign. IBM started to recognise the role of F/OSS (free and open source software) as on-ramps onto what were at the time more enterprise-ready platforms - Linux to AIX, MySQL to DB2 and so on. And Oracle just started to buy everybody it could get away with, as it always does.

Meanwhile we have Sun, which came surprisingly late to the party. Sun’s going through an open source epiphany at the moment, which is just dandy - though I’ve been spending a lot of time thinking about just how successful they will be. Sun’s heritage with software has been dodgy to say the least - it had a good start with the Catalyst catalogue and a pretty healthy software channel back in the Eighties, but that was in the days when the hardware manufacturers called the shots. Things started going a bit ropey in the early Nineties, when a number of big software plays (developer tools and network management) started to wither on the vine. Java came and should have been Sun’s big success, but the Internet came next and took all the attention away. While Sun was being the dot in dot-com, it forgot to be anything else.

It could be argued, quite successfully I am sure (though I will not try to here), that Sun has turned to Open Source for two reasons. First, it had no other choice, as it was no longer seen as a credible player in the world of proprietary software and it had burned its bridges with the flat-rate licensing deals brought in earlier this millennium. Second, one place Sun does have a growing reputation among its own customers is in services. Today’s open source models are all about building a services revenue stream, and I wish Sun success in that.

In doing so, Sun, IBM and indeed Oracle have embraced open source and integrated it into their business models. There’s one last area of course that open source can be used, and that’s as a competitive weapon - the only major company which is yet to embrace open source in the same way is Microsoft, preferring still to approach open source from the point of view of interoperability, not as an integral part of its software platform. Personally I think this is a mistake, but - let’s be frank, what the bloody hell do I know. Microsoft’s ultimate responsibility is to maximise its shareholder value, just as the rest of the majors. I have no doubt that they have done the maths, just as the others will have done.

Which comes back to the first point. If all software should be free then that’s great, but I don’t see IBM , SAP, Oracle or HP open sourcing any of its core moneymaking platforms. With good reason, from their perspective - its not in their commercial interests to do so. However it is in the commercial interests of some players to knock the competition for being proprietary, even while being quite happy to retain a significant proportion of the proprietary software market for themselves. Its a dangerous strategy - ask any of the bigger companies how they see the impact of open source on their own software base in a few years time, and they’d be hard pushed to give a straight answer. Fortunate for them that this industry has a very short memory, nobody will notice when they change their minds.

It’s all good fun isn’t it. Perhaps that’s the biggest reason why I’m interested in open source: it’s not the software itself, though that appeals to my geeky side; nor particularly wanting to consider the community driven development process, though that is a phenomenon in itself and worthy of attention. Nah - its watching the big guys duke it out in what is in fact a global game of paintball, with all the ducking and diving, short-lived alliances and backstabbings, and where the nature of the code maters little more than the colour of the paint.

Anyone who thinks modem speeds are a thing of the past (hi Joe, not a criticism) can’t spend much time working mobile, using GPRS, throttling back the arrival of the 4Meg file in Outlook so that he can get onto the Web and post a blog. 56Kbps doesn’t seem all that distant, because that’s pretty much exactly what I’m on right this minute. In central London, no less.

Yesterday I had a rather interesting conversation with Dale about how perspectives on information management can vary according to the provenance of the people involved. At the highest, most visionary level Information Management can be defined (and indeed, is, by IBM) as getting the right information to the right people at the right time. That’s simple enough to have people nodding sagely or shaking their heads in a “well, yes of course” kind of way.

While the “what” might be simple, the “how” can very quite considerably. Essentially there are 3 camps:

- the Business Intelligence (BI) crowd - with a structured data background, these people discuss normalisation, cubes and master data.

- the Content Management (CM) crowd - historically working with documents, they talk in terms of taxonomies, workflow and rights management

- the collaboration crowd - coming up from file-based environments and delving into Intranets, for these it’s all about desktop access, email and office integration

While each group of people may be highly computer literate, each will tend to talk to its own using a certain language, philosophy and (dare I say it) psychology. I know this to be so, based on experiences talking to groups of each type. I could waffle on about that for a while but hold that thought - what’s perhaps more interesting is how this impacts specific vendors, again, based on their provenance.

The obvious examples are:

- IBM, with its DB2 heritage, still very much in the BI camp despite having acquired FileNet

- EMC, having purchased Documentum 4 years ago and falling rather naturally into the CM camp

- Microsoft, its Windows-as-the-platform heritage yielding a firm position in the collaboration camp, with its Sharepoint “flagship”

Savvy, companies marketing execs in these companies will be able to pitch at the visionary level. For the rest however, its not so much that they don’t get the existence of the other areas, more that they don’t see the point of dwelling on them. They’d be right to surmise that there’s so much going on in their own areas, they’re kept too busy to step across into the other areas. The end result however is that we end up with three communities, not one, the behaviours of each dictated by their own technological heritage. For a stark example of this, consider what part Lotus has to play in IBM’s Information On Demand vision (here’s a clue: not much).

Does it really matter? I believe it does, not least because of the inefficiencies of reinvention and the requirements for integration across the piece. Perhaps it matters most of all because our research tells it so - organisations really do want an integral view of their data assets, of all kinds, and they are frustrated by the inability of vendors to serve it up. I do understand how hard it is to build a scalable repository, but that doesn’t mean we need to stick with models and architectures that are some 20 years old. Neither am I sure we can afford the luxury of preaching to the converted within our own comfort zones. 

P.S. I’ll leave the decision about where Oracle fits, as an exercise for the reader

A long, long, time ago, in a vertical far, far away, I once used to spend a lot of my time doing various things with UNIX - as an administrator or a developer. Somewhere along the route I appear to have become waylaid - I now spend most of my time (literally, I feel) in front of a Windows machine, like one of those office workers I used to have to support. How the tables have turned.

Don’t get me wrong, I’m quite satisfied with what I have, though some things could of course be better. Much like, I suspect, the majority of office workers out there. However (and probably unlike the majority of office workers out there), there are some things I do miss about UNIX:

- text manipulation - the number of commands (sort, uniq, awk, etc) available to muck around with strings, extract them, compare them, munge them together and so on

- pipes - with the above, making it very simple and elegant to develop command lines that could do some very powerful things

- the command line (of course) - but not just for text stuff, also just to make it very easy to move around directories and move things around

- the knowledge (though this had to be learned) of where admin information was being stored, generally in text files which could  be easily changed (though this was a two edged sword!)

- finally, the general feeling of control that comes from having an operating system in which everything was configurable, even to the extent of tuning the kernel…

… so I guess the question is, if it was so great, why don’t I still use it? Bluntly, the answer is that it wasn’t - so great, that is - at least, not for I have needed computers for, for the past 10 years. This is possibly more an indictment of me than of any specific technology, in that I don’t have the time or the inclination to spend my time tinkering with software, when I should be getting on with other stuff - writing, generally. Also, and again to be blunt, office apps have (until recently) been poor or stupenduously expensive for UNIX, as anyone who tried WordPerfect for UNIX will testify.

All the same, that doesn’t stop me missing such things as the above - particularly when desktop tools from Windows don’t cut the mustard. If I wanted to dedupe my email contacts in UNIX for example, I could do so with 2 or 3 commands piped together in a single line… whereas in Windows, I have the option to filch around the Internet for a piece of freeware (slow), or write a bit of code myself (unlikely - and option 3, to actually buy something, is beyond me completely). Of course, all is not lost as I could just install Cygnux for many of the above benefits, but somehow, I just don’t think it would feel the same.

“Never comment about those things you know nothing about,” is the recommendation - so I won’t remark on SAP nor, particularly, about Business Objects, though I have had dealings with both at various times. Aside from questions about what BO means to SAP (maybe I should rephrase that), interesting to me was what this means for the wider market.

As I have written previously, information management and, for that matter, service management are like different tracks up the same mountain, with fissures in between caused largely by historical scenarios - even the separation between structured and unstructured data comes largely from the adoption of two types of (incompatible) data repository. The result is that we have, today, a number of larger companies that offer large-scale, solution type software, and meanwhile smaller vendor companies whose job it is to add functionality or enable integration between the bigger stuff.

There’s a whole stack of acquisitions we’re seeing right now, which are potentially partially to do with the liquidity of the market (i.e. companies have cash to spend) but equally, there’s a consolidation phenomenon taking place. The big players - IBM, Oracle, EMC and now SAP are all filling gaps in their own portfolios - like blotters, absorbing the smaller players as they go. This is healthy, particularly for larger companies that want to rely more closely on a smaller set of vendors; but also for the vendors themselves, who can look to deliver more integrated solution sets and service packages to go with them.

P.S. Still waiting for the HP acquisition of Opentext/Ixos! No I don’t have any insider knowledge, but to me it would be like a key fitting in a lock.

Goodman Martin just forwarded to me a most excellent blog post about whether ITIL implementation could be replaced by a placebo, based on (say) astrology. Could it be done, and would it be effective? I’d have to answer yes, but that doesn’t diminish the value of ITIL itself.

Instead, what this comes down to is process. Management initiatives have been legion probably since Plato suggested everyone should get out of the cave (John Gray, eat you heart out). But what often yields success in management initiatives, is the process that is worked through and delivered upon, rather than the specifics of the initiative itself.

In some ways this is similar to therapy. Some advocates of homeopathy say that a great deal of the benefit comes from having someone with whom one can share one’s problems, and the fact that one is told at the end to stick a couple of small, sugary pills under one’s tongue is just one element of the overall experience. It is a tricky one - because if it is ultimately seen to be true, it does suggest that snake oil salesmen were maybe not quite such confidence tricksters after all. And IT marketers… maybe that’s a step too far!

And so to ITIL implementations, which will undoubtedly require some of the following:

- strategy definition

- discovery of “what’s out there”

- reviews of existing processes

- interviews with key stakeholders

- training

- definition and use of metrics

These characteristics are not new, and they are not specific to ITIL. What they do offer however are opportunities to engage, to review and to update the organisation (IT or otherwise) on the latest incarnation of best practice. Right now, in IT management circles, this best practice revolves around ITIL - which, let’s face it, is a pretty good starting point. But even if ITIL is being picked up by an organisation in a faddish way, the process still offers such opportunities.

Of course, the question then becomes, how useful are activities such as those listed above? In my consultancy days, I can remember being brought in to work on business process modelling exercises, but when I went to interview key people, they would tell me it was the third time that year they’d been interviewed, for different initiatives. At which point, any such efforts become counterproductive.

Overall then, should we have IT astrology improvement programme? Well, potentially. But only if it goes through a process that will make a difference to the organisation, all by itself. Indeed, an initiative without a correct process is probably no initiative at all.

Online backups of desktops and laptops are such a no-brainer for so many small companies and individuals - aren’t they? Markets are all about supply and demand, so if this were true, there would be a mass of different options available. But there isn’t, which suggests that either the conventional wisdom is wrong or something that needs to be in place, just isn’t.

I started playing with online backups a while back, when I was introduced to the company Connected.com, way before it was acquired by Iron Mountain (a move that continues to flummox me). At the time the bottleneck on such capabilities was the available Internet network bandwidth, but then broadband arrived and took that problem away. I continued to use Connected.com for quite a while, but then after one laptop upgrade I never got round to reinstalling it. These days I’m running a RAID box in the spare room at the home office, and backing up our computers to that on a nightly basis. Hmm, no online backup. Why?

The main answer is probably that the quantity of data that is changing, despite broadband, exceeds the bandwidth available to back this up. For me this is about email - to keep some kind of control on my email load I make extensive use of offline folders, which are now several gig each in size. While Connected.com professed to do clever things with email data files, my personal backup windows were growing far too big. Meanwhile, personal photography and video capture habits are growing a large quantity of multimedia files, and to back them all up (40 gig and counting) would break the back of all but the most expensive online backup services, not to mention my ISP fair use policies.

Meanwhile, there is the question of usability. It pains me to acknowledge that many small organisations and people aren’t running backups - or at least, are facing the risk that if they have a hard disk crash, they could lose all of that customer data, or family snaps, or whatever. I really, dearly want this problem solved. While Connected.com was eminently usable by me, it still fell into the trap of assuming that the user was computer literate, and so could only ever be working with the minority.

And so, to Mozy. EMC’s acquisition is the first time (to my knowledge) that an online backup service has entered the fold of a major vendor whose business is based on effectively delivering storage solutions. I’ll let EMC blow its own trumpet on that one, but let’s face it, you would hope that if anyone can crack the code it would be a company that had set its core business on it!! Immediate caveat, no, I don’t believe EMC’s going to get it right just because they set their store (sic) on these things. However, one would hope they stand more of a chance than, say, manufacturers of washing machines, or indeed, companies that have built their businesses on providing secure offsite locations for holding large wads of paper and boxes of tapes.

To resolve the issues of bandwidth and usability together, Mozy needs to be delivered as an integral part of any small company’s information risk management strategy. I deliberately use this term rather than backup strategy, because lets face it, backup is the answer, but not the question. Not all information was created equal, and not all information is subject to the same risks - so, given the fact that we have different ways of backing up and protecting information, we should be able to pick and choose which mechanism is appropriate for what type of information.

Looking specifically at usability, however, such gubbins needs to be kept under the bonnet. Consider a specific example - the emails I have sent over the past few days are quite likely to see responses (I hope), and I would probably like to refer back to them. Meanwhile, while I may want to refer to documents I created as part of older projects, the chances are I won’t want to change them. In terms of risks - the chances of a house fire in the next week, are 52 times less than the chances of a fire over the next year (and potentially increasing further - if my son’s demands for fire poi are ever heeded). In other words, I would dearly love to know that the information I have just created is protected in some immediate way, and I am pleased to have my older data protected, but they don’t necessarily need the same mechanisms: if my recent data is backed up in-house that’s a pretty good start. I may change my mind if I’m working remotely from the office

The bottom line, for me anyway, is that Mozy is a feature, not a product. The product I’d like to see in small businesses is one which exists as a client on every computer, and which can then deliver a co-ordinated backup strategy that meets the needs of each individual. If disaster (of whatever form) should strike, then emphasis switches onto the usability of recovery tools, so that individual files (or indeed, the entire environment) can be rebuilt. Mozy by itself may offer a tool for individual use that can offer significant protection, particularly for IT-literate individuals with lower data transfer requirements or high bandwidth availability. For it to be sustainable into the future, and pass the “upgrade and reinstall test” it will need to become an integral part of the backup toolset, in-house and external.

I’ve just installed this handy little widget on my personal blog. For those (anyone left besides me?) who hadn’t heard of it, reCaptcha is a word verification tool reputedly to guard against comment spam. I’ve only just installed it so its success is yet to be proven - however it does have a spin-off benefit. Text strings are taken from publications being scanned (currently) for the Internet Archive, this is what the reCaptcha guys have to say about it:

“reCAPTCHA improves the process of digitizing books by sending words that cannot be read by computers to the Web in the form of CAPTCHAs for humans to decipher. More specifically, each word that cannot be read correctly by OCR is placed on an image and used as a CAPTCHA. This is possible because most OCR programs alert you when a word cannot be read correctly.”

I believe comment spam is still coming in through the back door - but I have Akismet to help with that (and thanks to those guys). Meanwhile, here’s a tool that is not only keeping my front door clean, it’s making a literary difference as well. Hurrah to that.

I’m assisting Matt Deacon this evening at the IASA meeting in London. While I’m not an architect myself (”I’m a lover, not a fighter” springs to mind, though I have no idea why), I do get a lot from hooking into the people whose job it is to make IT work in their organisations. More than just keeping me grounded, its also a great environment to test ideas and find out what’s really going on.

I’ve been meaning to write this for a while, but as usual I couldn’t find the time to write the full story. So I won’t but at least I’ll give the highlights! It just struck me as an interesting example of how connections yield, well connections, and potentially results.

A long time ago, I’d been thinking about writing a book about how to do IT right. Not that I had all the answers, but on my travels I’ve met plenty of people who have suitably inspired me, and I wanted to distil their collective knowledge in some way. I even started writing bits of it, but it never really reached critical mass.

Meanwhile… one day I was feeling a bit sneaky. Sudoku was growing in popularity, and I realised it was only a matter of time before people brought out bigger versions. How about an alphabetic version, I thought, a 25×25 grid based on letters? It could be called - alphadoku! A quick check revealed that the web site www.alphadoku.com was free, so I, ahem, purchased it.

Almost immediately I felt guilty - how could I sit on a web site without doing something useful? So, I set about producing one of said puzzles. A little while later, it was done - and I posted it up. One day, I thought, I would get round to writing code that could autogenerate alphadoku puzzles (note: started, but never finished - yet!)

So, I left things as they stood.

A goodly handful of months later, I got a phone cal from Wiley, the “for dummies” publishers. Was I interested in writing an “Alphadoku for Dummies”? Yes of course, I said. Unfortunately, Wiley went away to think about it some more, but when they came back they had decided the market was probably no longer in the ascendant - which may have been a good job considering my coding skills.

However, I did ask - “while you’re there, I’ve been thinking about this technology book - interested?” Perhaps out of their guilt this time, I was put in touch with the right people. Almost immediately I realised how crap the book would be if it was just from me - by no coincidence, I was at the time working with the guys at MWD, whose opinions I valued (and continue to value) enormously, as well as those of my past and present colleague Dale Vile. So, I proposed we jointly wrote the thing, and for better or worse, everyone agreed.

So, from registering a Web site, we have a book - “The Technology Garden“. I thoroughly recommend it of course, and it wouldn’t be a quarter of what it is without being a team effort between the four of us. Still, ain’t it interesting what unexpected acorns can grow into?