Eusprig 2007

Another superb event. There seems to be an increasing buzz around the conference, Eusprig and the spreadsheet risk world. Which is more like the End user Computing (EUC) risk world these days.
All the vendors and industry analysts (it is now an industry and it does have analysts!)( EUC compliance/quality) are reporting significant increased activity. Maybe its time to dust off XLAnalyst and release a new version.

Its always a relief to get get confirmation of a feeling. I was third to present, and the third to mention the explosion of Access databases, glad its not just me.

Dean Buckner from the FSA, always provides a frank and fascinating insight into the world of regulators and the world of the city, big banks and traders. Every time I attend Eusprig I wonder if I should stop avoiding the city and get involved with the big scary stuff down there. Then I squeeze into a sweaty, creaky, rammed tube, stand up for the full journey and remember why I left London. My mates transport their sheep and cattle in better conditions. Mind you the journey south by train is pretty good these days (thanks Virgin trains).

As usual, virtually all the papers were excellent, and I think pretty much every one gave me something to think about.
I now have a ton of follow up to do.
Highlights for me, apart from far too much beer on Thursday night:
A suggestion in Steve Powell and Barry Lawsons paper that maybe the real ‘risk’ was the time lost working with sub-optimal spreadsheets. Errors are certainly there in live spreadsheets, but the impact is often not dramatic. A fact re-affirmed by Dean, when he mentioned that most processes are performed multiple times, by multiple people, often in different ways, and that this seems to catch many errors. Makes me wonder if there is some mild chaos theory thing going on.

New TLA, courtesy of David Colver of OPERIS:
MUT. I so wish that could have been MUTT. (woof)
(Model Under Test)

A couple of people mentioned something which I kind of thought a bit, but hadn’t worked out properly. Peoples actual spreadsheeting ability is often inversely proportional to their perceived ability. Basically many intermediate level users think they know most of a small subject. Advanced users realise that they know a lot, but they also realise the topic area is massive. Puts me in mind of ‘The Humble Programmer’ by Dijkstra.

also got a couple of new products I need to take a look at:
Resolver– looks well neat, combines a grid with Python, with much tighter integration than Excel/VBA. It has proper data sheets just like I was on about a few months ago. Are you guys stalking me? (joke – but the Resolver folks were keen for feedback and ideas, so grab the demo and get playing). I’m going to learn Python especially.
Expresso – a way of collaborating with spreadsheets – if I had any mates I’d try this ;-)

Someone also mentioned a trend of sharing chunks of spreadsheets via googledocs. Which is just another online service which I can no longer access because I forgot my password. (I’m also locked out of the Microsoft partner extranet for the same reason, and AccountingWeb, and the Eusprig yahoo group web interface, and a ton of other things). I’m looking forward to AccountingWebs soon to be published report on password managing software, I need it (assuming I can access it of course!!).

‘Frankensheet’ struck a chord with a few people, as did this slide:

crap formula

I’ll put the full slides on the Codematic site in the next few days. Mind you its just mainly pictures of none spreadsheets.

Eusprig 2008 will probably be in Greenwich/south England, despite our attempts to get it back to Dublin, so be sure to clear the first couple of weeks of July to attend. I think this is a ‘must attend’ event for anyone serious about spreadsheet development. Not just for the excellent lecture content, but also for the chance of informal discussions around spreadsheet quality with some of the leading thinkers around.

One of the saddest things I heard was that some banks/financial institutions wouldn’t attend because they didn’t want to ‘appear’ to have a problems with spreadsheets. HELLOOO? We KNOW you have a problem, just own up then we can fix it together.  

 If you were there leave a comment with your thoughts. (what happens in the pub, stays in the pub right? ;-)



8 Responses to “Eusprig 2007”

  1. dermot Says:

    re why we don’t get more big errors from sub-optimal spreadsheets – I think our intuitive reasonableness checks save us from the big errors most of the time, and if the answer looks reasonable, we’ll often accept it so the (smaller) errors never get found

    re ability, I think it’s quite simple. Users start by learning the technical side of Excel, and then think they are experts. Real experts find out that there are whole other worlds like testing, UI design, programming, error control, etc and that they will never master everything. To paraphrase an old actuarial saying, “an Excel expert who is only an expert in Excel is not an Excel expert”.

  2. Marcus Says:

    Hi Simon,

    Great update – looking forward to the slides. I’m planning on attending next year.

    “third to mention the explosion of Access databases”

    I have a theory, probably more a sneaking suspicion: The business hates the IT dept and wants as little involvement with them as possible.

    I’m working on a couple of tactical solutions at the moment, Y’know the ‘temporary’ solutions which live on… forever. In each case the business has asked that the solution should not involve the IT dept for either development, deployment or support. This is for a variety of reasons, but essentially it’s once bitten…

    For most of these solutions though Access is only the repository, not the GUI. Excel forms (via an add-in) are the usual GUI. What’s your experience been?

    “that most processes are performed multiple times, by multiple people”

    Yes I’ve witnessed this too. Often this is related to authorising a transaction or model validation where the spreadsheet model effectively gets signed-off up to an ultimate level of seniority.

    “peoples actual spreadsheeting ability is often inversely proportional to their perceived ability”

    You bet. This article, entitled “Unskilled and Unaware of It: How Difficulties in Recognizing One’s Own Incompetence Lead to Inflated Self-Assessments” might be enlightening.

    Click to access psp7761121.pdf

    “banks/financial institutions wouldn’t attend because they didn’t want to ‘appear’ to have a problems with spreadsheets”
    Some have labelled this as ‘ignorance is bliss’. Others call it happiness. Happiness, as we all know, is being too stupid to know what to worry about ;)

  3. Harlan Grove Says:

    So Resolver is Python’s answer to Office Web Components?

    Have fun w/ Python. It ain’t bad, but it’ll be a big change from VB[A].

  4. Ross Says:

    >>”So Resolver is Python’s answer to Office Web Components? ”

    No way, it looks much more than that. One of the devs has a few blogs that are worth taking alok at:

    Looks intresting that Resolver thing, let us know what you make of it simon.

    thanks for the updates!

  5. Simon Says:

    yes and yes, I totally agree, like the saying.
    Look forward to seeing you next year then. Business v IT – agreed totally (see the garden/field slide). Excel front end, yes totally, although I avoid Excel forms – VB com add-ins right? thanks for the link, I’ll read it tonight.
    anything thats a big change from VB/A is fine with me, the bigger the change the better. Been doing a load of C# recently, just opening the VBAIDE makes my heart sink. (Even though it is sooo much better than C# for working with the Excel Object model).

  6. Marcus Says:

    Hi Simon,

    Funny, I don’t avoid Excel forms – the majority of the add-ins are xla but I also do COM add-ins. The clients often like/prefer the xla’s for a couple of reasons:

    1) They understand it – it’s just part of Excel.

    2) In a locked down environment, deployment is very easy. In some environments, COM add-ins have to be wrapped up in a distributable for NetDeploy (or similar)

    3) The ‘what if I get hit by a bus’ syndrome. They perceive that I’m easier to replace with another VBA developer than a VB6 one.

    4) The code is invisible (compiled) – they often feel more comfortable with being able to enter a password and see real code.

    5) Some banks have put a freeze on VB6 development as it’s no longer supported by Microsoft (which ironically hasn’t stopped me from developing some DLL’s in VB6 for one such client).

    However, when I’m doing xla’s I really miss object arrays and user controls.

    Cheers – Marcus

  7. Patrick Says:

    Hi all,

    Simon, thanks for the excellent talk at EuSpRIG – very funny as always…..

    Harlan, just to make it clear: Resolver is a complete application like Excel and is not just a web publishing tool although there is an optional version to do just that.

    Resolver Systems

  8. Don Price Says:

    Hi Simon,

    An excellent post on an excellent conference, and your presentation was as humorous as it was serious (as expected).

    I also liked Resolver, and thought it was a breath of fresh air to see the demo and listen to the very keen presenters, who could have been going into the lion’s den at EuSpRIG but turned out to have tamed us all! Well done to them and I hope we see them next year.

    Apart from your talk, I liked David Colver’s – he is always worth his rations and speaks as he finds. His methodology for audits is interesting and he obviously lays down a specific law to his staff (loved the comment “it’s a dismissable offence to send out a spreadsheet unformatted for printing”).

    Keep up the excellent blog!



Leave a Reply

Please log in using one of these methods to post your comment: Logo

You are commenting using your account. Log Out /  Change )

Twitter picture

You are commenting using your Twitter account. Log Out /  Change )

Facebook photo

You are commenting using your Facebook account. Log Out /  Change )

Connecting to %s

This site uses Akismet to reduce spam. Learn how your comment data is processed.

%d bloggers like this: