A Case For Wonder/D2W (was: How would you handle such situation?)
A Case For Wonder/D2W (was: How would you handle such situation?)
- Subject: A Case For Wonder/D2W (was: How would you handle such situation?)
- From: Anjo Krank <email@hidden>
- Date: Wed, 3 Aug 2005 23:27:33 +0200
Am 03.08.2005 um 20:36 schrieb Florijan Stamenkovic:
In the end you might end up with a model independent template.
Something like D2W :)
Exactly. So I wouldn't really bother with any home-brew-stuff and use
what is provided. I know I sound like a broken record, but what the
heck...the time used to learn D2W is well worth it.
<tiny-success-story>
The place I work, I'm always introduced as the "guy doing the backend/
admin". This is blatant lie, as I haven't been actually doing
*anything* of reasonable complexity in that regard for the last
couple of months.
Our app(s) have currently about 130 entities, growing in a rate of ~4
a month. There have been *major* reorgs of some parts of code, all
entities are editable with access control and while the framework EOF
code has grown from 25k lines to 60k lines, the backend has grown
only from 6k to 12k lines, most of which is used for a few larger
custom components. The Backend app has about 250 pages you can
actually navigate to as of last count (which is quite some time ago)
and has a lot of "hidden" screens, which you can access if you have
the right privileges but I haven't really bothered to specify and
create navigation items for.
It currently supports German and English as UI languages and could
support a lot of other languages without any real work except for
people doing a translation of some keys.
There are tons of other features beside editing like extremely cool
and complex reporting in html, excel, pdf, gfx (courtesy of
DynaReporting, ERXLSTWrapper and ExcelGenerator - and FOPObjects for
PDF rendering, which is so trivial that I didn't even bother to
include it into Wonder). Like debugging stuff, newsletter sending,
app monitoring etc all done with these mere 12k lines.
Now, adding new features is provided by adding a few entries into a
navigation list, (probably) writing some new rules and localizing
some keys. That's all. Most of the stuff I get asked for takes more
time to write commit messages than to actually do them. And the day I
decide that I want my interface revamped I change about 5 templates
and adjust one css file and I'm done.
And the reason this works is because of the D2W rule system, which
allows you to specify default rules and then specialize from them and
Project Wonder, which does *so* many things for you I won't even
start to sum them up.
</tiny-success-story>
The bottom line is: D2W is well worth the any effort you put into
learning it - provided that you have actually a halfway complex
problem in the first place:)
However, given that you can practically give away a serious backend
app for free - thus beat a lot of non-WO shops on price, it may even
be worth it on smaller projects. I've also done two of those and both
took only a few days to implement - even tough they had about 25
entities and about 80 screens each.
So there, enough bragging about me:) Are there any stories anyone
else could share?
Cheers, Anjo
_______________________________________________
Do not post admin requests to the list. They will be ignored.
Webobjects-dev mailing list (email@hidden)
Help/Unsubscribe/Update your Subscription:
This email sent to email@hidden