[Seaside] Thoughts about WikiCalc?

[Andrew Catton]

You asked Avi, but I'll offer my 2 cents anyway :)

On 11/14/05, Darius <squeakuser at inglang.com> wrote:
>
> Avi,
>
> Any thoughts about Dan Bricklin's WikiCalc? It seems like an attempt to
> fulfill
> the long tale of simple app. dev. similar to Dabble DB.


Dan's definitely going for the long tail too, and everything we say should
probably take into account the fact that he developed one of the first and
most successful long-tail product types ever. I also think that a good
online spreadsheet (or something like that) is a very valid product, though
not what Dabble is aiming for (we're after the *mis-use* of spreadsheets :))
FWIW, I actually think the current approach to WikiCalc doesn't go as far
down the tail as we do with Dabble.. more on this in a second.

WikiCalc's potential failing as currently configured is that it seems to be
> document/page based as derived from wikis. It seems to me that it would be
> a
> better solution if it was more cell based, a finer granularity for the
> user
> interface, storage, mashing, publishing, and distribution.


Right, I certainly agree with you there.. in Dabble we have take 'Entries'
as the base unit (rows in an RDBMs or objects in Smalltalk, though we have a
more flexible type system than both of those :)). The nice consequence of
this is that having a wide variety of views on the same data becomes a
trivial thing to think about and use. Also, when you need a system to track
and manage data, real relations between things are kinda nice :)

(That level of granularity could more easily be supported with a
> Seaside/Squeak/Magma solution in my opinion. But, WikiCalc has the
> advantage
> that it can more easily live and spread in a FTP, Perl, XML server world.
> [sigh])


Back to the long tail.. I agree that for a model where people download
software, install it on a server somewhere and share it with collaborators,
the perl, etc. side of WikiCalc is a huge advantage. However, if you go a
bit further down the tail, where you don't assume an IT department, person,
or even anyone who has the ability or interest in going that trouble, this
ceases to be much of an issue. As Dabble is meant as a hosted service, we
expect our users basically not to care about the underlying technology,
provided they can use the system and also get data in and out of it in
whatever ways they need. They just go to the right URL and there everything
is. For what it's worth, as a developer myself, I find there are a great
many long-tail apps even I wouldn't bother to build and run if it involved
more than clicking 'Create New Application'..

Cheers, Andrew
-------------- next part --------------
An HTML attachment was scrubbed...
URL: http://liststest.squeakfoundation.org/pipermail/seaside/attachments/20051114/fc894b45/attachment.htm