Slashdot Mirror
VisiCalc Creator Developing WikiCalc
An anonymous reader writes "ZDNet has an article about a new wiki that is trying to combine the collaboration of wiki technology and the data manipulation attributes of a spreadsheet. The creator of VisiCalc, Dan Bricklin, is working on an alpha version of WikiCalc for sometime late in February." From the article: "'It holds a lot of promise, both because it's using the spreadsheet metaphor, which is the one thing people know for working with quantitative information and because 'there's nobody better in the world to build this thing,' said Ross Mayfield, CEO of collaboration software maker SocialText. To Mayfield, WikiCalc is the answer to a problem that has been percolating for some time in the world of IT. That is, he said, that spreadsheets have traditionally been a single-user application screaming for functionality that could let multiple people edit data quickly and easily. "
spreadsheets have traditionally been a single-user application screaming for functionality that could let multiple people edit data quickly and easily.
Hence corporations all having relational databases with custom GUI applications. Spreadsheets are most useful for tabular data, which of course works well in relational database tables. While spreadsheets are great at free-form manipulation and "playing" with the data, it's the custom apps that are required to sqeeze that data into the corporation's customs workflows. For at least 20 years what corporations have been doing is creating the custom apps and having them export to more freeform data models like spreadsheets as needed. This seems to work pretty well.
But "supercharging" spreadsheets won't really be providing power to the people that need it. The people that most need power over large amounts of data have hundreds of people working in their IT departments.
Developers: We can use your help.
Think about things like budgeting/forcasting in a large operation with multiple departments, all of whom need to work on their individual sections. You end up with either lots of spreadsheets that are linked together if you're lucky, or everyone taking turns at the master spreadsheet. If they get a decent selection of formulas working this could really simplify things for stuff like that.
In the financial industry I've seen mini-applications where spreadsheets were the whole data store. Many financial analysts live in Excel. Spreadsheets can be a powerful, useful tool. But most often they grow into horribly ugly monsters that the IT department has to de-tangle and cram into custom apps.
Developers: We can use your help.
...a new tool to add money to the big boys' pockets. Saw an interview with him years ago. He tried to make it sound as if creating a program that eventually put billions into the coffers at Lotus and Microsoft but left him with a teachers salary didn't sting all that much. But, it was evident in his eyes that he was stung and felt he missed the boat that made young millionaires out of the geeks of the late '70s.
Politics is the art of looking for trouble, finding it everywhere, diagnosing it incorrectly and applying the wrong fix.
I've seen this problem here (I'm in IT). A really smart Actuarial creates a very good, effective spreadsheet. It becomes both production and "mission critical" to them. The Actuarial leaves. Oh, did I mention that there is nothing documented? And the person is a Ph.D using calculations that are not understandable to mortal man (or even a woman!). Who supports this? Especially when it breaks because desktop support rolls out the next version of Excel. Screams abound!
And I giggle because I'm a dino running on old, obsolete mainframe technology where the end user can't just slap something together and put it into production.
It does. This is obviously just an RDBMS with a web front end and hooks for writing (and probably saving) your own "mathematical" SELECT statements. Big deal.
After all, I am strangely colored.
The intricacies of spreadsheets make them much harder to edit in parallel. On a wikipedia entry it doesn't matter if one person edits something about the history of something while another person expands a section on the future. Aside from minor inconsistencies, which are easy to spot, the document is essentially the sum of its parts.
In contrast, the parts of a spreadsheet have strict dependencies that can span the spreadsheet and affect correctness in subtle ways. For example, if one person adds a row in one section, how should formulas in a different section react (do range references to the row above expand to encompass the new row or do range references to the row below expand or neither?). "Trace dependencies" functions can help but only if each editor recognizes that the scope of their edits is potentially unbounded.
The point is that it's harder to allow simultaneous independent edits because the internals of a spreasheet don't have independence.
Two wrongs don't make a right, but three lefts do.
I figure 90-95% of all the spreadsheets I see don't do any calculations, they're just used as a way to put things in columns.
This is exactly the reason that one of my favorite apps, OmniOutliner (Mac OS X) was created.
"when the Excel product manager got up on stage at MacWorld several years ago and said, "We've found that 85% of our customers use Excel just to make lists and outlines," we (Omni) said, "Shoot, that'll be our next product. We can do a GOOD job of making lists and outlines, and sell it for a lot less."" -- Wil Shipley, Omni co-founder
It seems like there might be a market opening up in the "things that people are already misusing Office for" sector.
Don't become a regular here -- you will become retarded.
And a podcast is just a RSS based distribution of sound and video media. But it is a good implementation and use of RSS, and a WikiCalc would be a great use of wiki.
Some of the best most obvious innovations seem rediculously obvious in hindsight, but that doesn't detract from their greatness (in fact, you could just say they were elegant)
Make sure everyone's vote counts: Verified Voting