[aida] Vision for Aida
janko.mivsek at eranova.si
Sun Mar 11 12:02:34 CET 2012
I'm forwarding a response from Sebastien Paquet, posted on Nabble which
has also problems sending posts to our mailing list. I'm investigating
the problem ...
23:43, 09.Mar.2012 sebastien paquet
I recently started using Aida and learning Smalltalk at the same time
and I have to say that I really believe in a great future for this
framework and where it could go.
I just did the blog tutorial that used to be on the Aida site and where
the link is broken. I squinted my eyes over the clip on Youtube and did
it. I realized that a couple things in the docs on the site and
tutorials seem to have changed. I mean code wise. Sometimes I ll do
exactly what It says and it doesn t work. I ll try something else and it
does. I guess it could be related to the versions but I m not sure. So,
I wish i could contribute and maybe write down the blog tutorial from
the video, it s really good for showing how CRUD operations integrate
with the framework. Also how to route the views.
Aida is awesome and not that Seaside is not good but coming from Avi
Bryant himself saying that its outdated as a lot of frameworks since it
was built at a time when problems were different. I think Aida is
different and can only get better especially when I hear people talking
about integrating Amber on the client-side. Wow! Technology at its best
and the future of the only real OOP language(!) looks bright.
Dne 10. 02. 2012 16:37, piše Janko Mivšek:
> Hi Bob,
> You just inspired me to write down my vision and future plans for Aida,
> so let me start with that:
> Main *Aida vision* from the start is to extend the Smalltalk philosophy,
> elegance, simplicity, power to the web applications as well while
> preserving all those beauties of Smalltalk.
> And to be on the *bleeding edge of web technology* with Smalltalk all
> the time. Which we are. I'm most proud that initial architecture was set
> good enough to be adaptable through all that time since 1996 and that we
> are not stuck even now supporting the current HTML5 wave.
> And here are the plans. Short-term, long-term? Well, we'll see :)
> 1. *Documentation*! It seems Aida core features are stable enough now
> that we can document it and that this documentation will last for a
> while without needing to update it constantly;
> 2. More *HTML5*! like FileSystem, DragDrop, Video and Audio etc etc;
> 3. More *realtime*, actually realtime anything, everywhere. WebSocket
> standard is now de-facto complete and in a year all browsers will
> support it. What I mean by realtime anything, everywhere? Any data
> you show everywhere on your web page will be updated in realtime,
> that is, immediately when data changes. And this will be switched on
> by default!
> 4 More *client*! That is, more processing moved to web client with
> Amber Smalltalk. Here the end goal is to partition your app in
> runtime where to run. All on server as now, all on client (and
> offline), or parts on server, parts on client. Client parts are
> "emitted" to Amber Smalltalk and sent to a client for execution;
> 5. More *local updating* at the client. On server only data is
> updated while all dependent views are refreshed locally on client;
> 6. *Automatic forms building* out of domain model, with help of
> forthcoming AidaFields by Nicolas Petton;
> 7. More *relational domain model* support;
> 9. Aida as provider of *REST-full web services* accessible from other
> web or mobile apps
> 10.*Cloud* support, Aida apps as both as client as service provider in
> the cloud;
> 11.*Aida hosting*, both free and commercial.
> Best regards
> S, Robert Calco piše:
>> I was wondering if you could reveal any grand plans you might have for
>> the future of Aida?
>> What I like about Aida compared to Seaside is its relative simplicity
>> and portability. Seaside uses features of Smalltalk that are not really
>> portable between smalltalks to do its magic (continuations and pragmas
>> being the big ones that I've noticed, which prevent it from being used
>> fully in Dolphin, for example).
>> It feels to me like Aida doesn't try to reinvent too many wheels,
>> meaning it's a bit more accessible to people coming from, say, Rails,
>> like myself. It has relatively more steak than sizzle, whereas Seaside
>> has relatively more sizzle than steak. I like my steak rare ("still
>> twitching" is what I tell waitresses at restaurants), so I don't care
>> about sizzle so much. That said, my favourite way to prepare steak is
>> 'black and blue' -- but it takes a rather industrial-size furnace to
>> cook it that way, but, working for a charity, I usually only have a
>> couple matches and a little kindling to start my fire pit. ;)
>> What I don't like about Aida is the documentation, which is a bit
>> scattered by topic and somewhat stream of consciousness. Seaside has
>> Aida beat hands-down in this department. There are books, well-designed
>> tutorials, lots of samples, etc. It's actually very hard, if you're me,
>> to ignore that big plus for Seaside.
>> I think Aida needs a much better, much more in-depth tutorial, that
>> really highlights its best parts, and suggests where it's going to get
>> better over time. I am not sure I like all the parts of Aida included --
>> for example, I've been reviewing the (very involved) Parties/Roles
>> packages and it encroaches a lot on my domain model, meaning I either
>> have to use all of it or none of it. But maybe I have misinterpreted its
>> design? Maybe it's better than my slogging in the code has revealed, and
>> I should use it? Some documentation about the rationale for the
>> abstractions chosen and how they're implemented would be nice.
>> What I'd like to see in Aida is a greater commitment to REST and maybe
>> XMLRPC API support (seems you're working on that one quite
>> aggressively), more of an MVP vs MVC pattern for rendering, and perhaps
>> support for rich web clients a la Amber (I see there is WebDAV support
>> in Aida, so perhaps you're already intending to go that route), and
>> better security (I really need to support HTTPS, and support API keys
>> securely). Seaside has done a good job in its HTML API, but Aida's
>> abstractions feel somewhat heavy (gazillions of methods). Some
>> refactoring of that seems like it could be good.
>> Neither framework handles layouts, look-feel, or any of the design
>> intensive side of web development as nicely as I imagine it could be
>> done. This could be an area for huge win in Aida, as I see you've tried
>> to make it easy to reskin a site by supplying a new web style. However,
>> I tried doing this with the demo app and only just mucked stuff up.
>> Again, good documentation about this important area of implementation
>> and some idea of what the end-goal is for it would be awesome.
>> I'm willing to contribute but I need to get up and running and confident
>> in what I'm doing still. It's been a bit of a ramp up. But I think I'll
>> do well with this choice for what I need to do. I would just like to
>> know what the future holds as I think there is still a lot of room for
>> innovation in web development, and it doesn't all have to be heretical
>> to be on the bleeding edge, so to speak. I'll feel better about Aida
>> knowing that it's roadmap is a good one.
>> - Bob
>> Aida mailing list
>> Aida na aidaweb.si
Svetovalec za informatiko
tel: 01 514 22 55
faks: 01 514 22 56
gsm: 031 674 565
More information about the Aida