[Nitro] The status of Nitro/Og - some personal thoughts

[Michael Fellinger]

Hey List,

I've experienced a bunch of problems while working on OgFile/OgImage/Thumbnail 
for Og...
mainly, because there is no documentation on the way the things are supposed 
to behave - luckily POLS (Principle Of Least Surprise) hurried to my resuce 
at most times... but there are some parts that need some thinking over...

Ok, basically i would like to plead to everyone to get back to the roots.
I don't want to know what exactly an Og-Object does, i don't want to take care 
about the fact that it's a Database under the clothes...
But it's being hard to forget about it, differences between the stores are one 
of the main-problems, while you're mainly secured from bugs in psql (when you 
don't do too fancy stuff) mysql and kirbybase are really left behind - also 
sqlite has some drawbacks.

Now, i don't think we're perfect yet, and i cannot really complain since i'm 
no db-guy and cannot work on the adapters without breaking everything... :(
but Og is supposed to work the same way on every store, i see there are some 
limitations, especially with KirbyBase and i respect the work that has been 
put into every single adapter so far, it's only that i cannot get rid of the 
feeling that i write my code for only one adapter - losing one main-feature 
of Og... :|
that gets me down to the ground very much about the limitations of Nitro/Og.

Now, i still haven't quit complaining, i know - but that is one point i cannot 
put enough weight on, and one of the main-reasons we're losing users. (a fact 
that everyone here should have noticed already)
I cannot go out on the street and preach Nitro/Og when i'm not comfortable 
with it myself.

My opinion on Nitro is another one, i think it's quite mature now, still has 
some downs but overall it has more energy in it than every other framework i 
know. To me it looks like a raw gem, waiting to be polished and get some neat 
facets :)
One of these facets is the new javascript-helper - never has javascript been 
that easy (correct me when i'm wrong) and tailored to your application!
The Videos are really pointing out the beauty and simplicity of Nitro - a 
really big thank you to george and all of the other guys who worked with him.

But there are still some minefields in Nitro - things that we pros seldom 
think about - stuff like nitrogen-bugs, missing tutorials (i really have to 
update nitro-in-flames), lacking documentation, unreasonable defaults or 
missing options, no explanation of the transformers and things like that...
It makes getting started a worthwhile, but hard procedure and according to 
almost every company out there, time is money - no time to learn nitro, more 
money to earn with rails.
I'm trying to help on that with RDOG, and there are some ideas around to make 
a source-editing-wiki as a fork of spark to get an easy way to edit 
comments/code of nitro/og/glue and the examples - also there was an idea 
about including some tutorials in the tests, to make sure we're notified if 
something breaks and can correct it.

I've been a bit disappointed by the development in the past few months - but 
i'm willing to forgive everyone because i know that it's hard to find time to 
work on everything... i, myself have only little time and my contribution to 
nitro was more in manners of working with the community... trying to find new 
ideas and bringing some interesting approaches into the game.

Ok, this has almost become a summary of all my experience with Nitro/Og, 
please don't think that i want to get you all down, i enjoy my work with 
nitro, i get along with Og, and i have lots of good experiences with it.
Take it more like a kind of status of Nitro/Og, because the thing we lack 
most, is a larger community - the project has grown, but where are the 
people?
We're visited almost everyday by new users, they try nitro and ditch it, 
mostly for reasons that could be easily fixed but are a pain for everyone who 
wants to learn much in the little time one has.
So take this as a call to the weapons, we need more quality, tests, and maybe 
even some documentation :)
This is a call to everyone who finds one of these small stones that make 
newbies or even pros stumble, put it somewhere in the wiki, mail it, put it 
into the rubyforge-tracker, send a bugfix - every little effort is well 
appreciated and needed.

Sorry, i didn't plan that mail to become that long... but i had to say that.
~~~~manveru
-------------- next part --------------
A non-text attachment was scrubbed...
Name: not available
Type: application/pgp-signature
Size: 189 bytes
Desc: not available
Url : http://rubyforge.org/pipermail/nitro-general/attachments/20051228/eb45f157/attachment.bin