My last two jobs have shown me that Excel is really what you need to kill if you want to break the Office stronghold.
May I ask what specifically you think an Excel-killer would need to do to succeed? And why you think the web-based spreadsheets haven't been able to do it so far?
I worked in medical and banking...these industries seem to run on Excel. Performance is an issue, and the amount of "legacy" code running in macros/vba is staggering. You are talking about thousands of excel spreadsheet which is each like it's own little application.
I recall talking to someone at MS about .NET 4.0 and I asked what were the big scenarios -- and I was surprised to hear that Office programmability was one of the top scenarios. He didn't know exact numbers, but he said that there may be more lines of .NET Office code (VSTO and VBA) than ASP.NET.
Go and ask your random analyst friend if they know VBA, and you'd be surprised how many know it (and with it know things like dynamic typing, implicit arguments, event handlers, etc..).
"And why you think the web-based spreadsheets haven't been able to do it so far?"
A huge number of business and engineering applications integrate with Excel - all the way from supporting easy data import and export (and not just through CSV files, which Excel actually handles really badly) up to fully fledged "live" integration through local web services.
Until a web-based application can reach in through the firewall and pull data from a LOB, ERP or finance app and display it as well as Excel can then web-based spreadsheets aren't going to make that much headway.
Ok, say Google sold their spreadsheet for installation on corporate servers. That would address the firewall problem. Would it be enough to kill Excel? What more would they have to do?
I doubt it - the spreadsheet in Google docs, even at a casual glance look to be missing some key features that are incredibly useful (Trace Precedents/Trace Dependents being one that jumps out at me).
There is also the fact that there are an awful lot of "mission critical" spreadsheets out there with huge amounts of VBA in them - I can't see Google supporting that (and you probably couldn't as a lot of these things call client side objects through COM).
To be honest, I don't think users want an Excel killer. Yes it has weaknesses (why it still can't handle Unicode CSV files properly is still a mystery) - but overall most people just see it as part of the standard corporate computing landscape.
You could perhaps make the whole model more interesting (e.g. by having better support for the hierarchical multi-dimensional data that finance people in particular just love) but then you'd probably lose the wonderful simplicity of the basic 2D sheet.
[NB I've spent a lot of time building finance apps that have Oracle Hyperion as a backend - which is a specialized 12 dimensional hierarchical database that is the market leader in its particular niche].
Just another super important point about this, you are talking about two industries who, for the most part, haven't bought into the public cloud yet.
I'm not gonna say "most", because, I really don't know, but I'd bet that most (oopps!) of these companies have firewalls that block google docs and the likes. Their idea of "cloud" is to run virtualization on their own servers in their own data centers and call it a day.
There's huge security (mostly made up), regulatory (mostly made up) reasons for them not to move to the cloud.
I assume you say "mostly made up" in the derogatory sense - how come? I'm basically okay with the spirit of what HIPAA is trying to accomplish, which is not sharing my medical information with anyone I don't explicitly authorize, accidentally or otherwise.
I'm no expert, but from what I've seen, both medical and financial regulation aren't nearly as detailed and absurd as IT-Risk/Compliance make them out to be. They don't say that you can't host with AWS (or AWS private cloud), but both those departments will insist that you can't.
It really comes down to being an unknown and it being easier (and perceived as safer) to "just say no". It isn't the regulation that I dislike, it's that said regulations are used to keep large organizations in the stone age. And, to be honest, from what I've observed, it really comes down to people not wanting to learn/work and having a powerful excuse (REGULATIONS!!) to avoid doing so.
Absolutely correct. Anyone who tells you that HIPAA requires you to do XYZ is full of shit. The difficulty with HIPAA is that it sets a very basic framework that you have to fill in with your own policies and procedurs. How strictly you follow your own procedures is what will get you in hot water.
May I ask what specifically you think an Excel-killer would need to do to succeed? And why you think the web-based spreadsheets haven't been able to do it so far?