Web framework or Web Application Server?

Janko Mivšek wrote:
  

Dear all,

Well, time to decide what to use for Aida. Should it be Web framework or 
Web application server? And we need to decide quickly, logo is on the 
work, Wikipedia entry also...
    

Regardless of the decision. I believe that whatever we decide should be 
well defined on the website as there are many differing views on the 
definitions of these phrases.

Personally, I like Web Application Server.

  

Nico's stance:
I think that we should be clear about that : Aida is a web framework,
Swazoo is a web server. Aida is clearly not a web server, and I think
that the sentence "Smalltalk web application server" is very confusing
for newcommers (It was the case for me when I discovered Aida...) : What
is Aida and what is Swazoo?
    

I can understand Nico's position here. This is why I believe that on any 
material which describes AIDAweb it should be defined as to the meaning 
of the term.

There is much differing and ambiguity in the world of web development.

For example what does "Full Stack" mean? Rails and others use such an 
expression to describe their framework. But the combination of 
deployable options in such "Full Stack" frameworks is dizzying. Web 
server..., Database..., Templating language..., etc.  So much for "Full 
Stack" and having quality decisions already made. :)

Nevertheless, I believe a good definition attached to Web Application 
Server may take care of Nico's situation and where he and others come from.

  

My stance:
Swazoo is web server while Aida is web framework. But
because Aida comes always with Swazoo, I think we shall name Aida as Web
application server, which is web framework + web server.
    

Point. Can AIDA serve a web app or Smalltalk objects without Swazoo?
To my understanding no.
So while AIDA is not a web server. It does require a single specific web 
server as a part of its framework architecture. It isn't optional. We 
can't swap Apache, nginx, lighttpd, or any other web server for Swazoo. 
We can put any of them in front of AIDAweb/Swazoo but we can't replace 
Swazoo with any of them.

To me this is a big distinguishing difference between AIDAweb and RoR or 
any other "Full Stack" framework.

I am not certain how RoR works and whether or not if you use Apache you 
still have to use its internal web server or not.

  

But on the other side, RoR and Seaside are declared only as web
frameworks and as we like to be on pair, we also use only "Web
framework" for Aida when appropriate.
    

Seaside can claim to be a framework if they wish. But it is just as 
dependent on Komanche's as AIDA is on Swazoo.

AIDAweb and Seaside are much more like Zope than Ruby on Rails or other 
such frameworks.

 From the Zope site at:  http://www.zope.org
Zope is an open source application server ...

From:  http://www.zope.org/WhatIsZope
What is Zope?
Zope is an open source web application server ...

More detailed descriptions available at the above pages.

You can put Apache, etc. in front of Zope, but you don't have to.
You can use an external database, but you don't have to.

Read the above 2 pages on Zope and see what you would like to use for 
descriptions or marketing.

But of course nothing, absolutely nothing is as well integrated, tightly 
coupled, out of the box available as AIDAweb or Seaside. :)

Zope is a close as I think we get to a Smalltalk like system, out in the 
file based world programming languages. And as such its advocates and 
opponents have the same likes and dislikes as are used for and against 
Smalltalk.

  

Also if you see title of our website on Google, you'll see a long
"AIDA/Web Smalltalk Web Framework and Application Server". That's
because people search sometimes as web framework, sometimes as web
application, sometimes as web application server etc.
    

In the end, to me, AIDAweb is for building and serving web applications. 
Yes, it does contain a framework, but it is far more tightly coupled to 
its components and parts than most.

Is a framework?
Has a framework?

Is an application server?
Has an application server?

Nothing will be perfect.

But I vote for Web Application Server and a quality definition as to 
what that means for AIDAweb.

Don't know if this helps. But hope so.

Jimmie
_______________________________________________
Aida mailing list
[hidden email]
http://lists.aidaweb.si/mailman/listinfo/aida
  

Jimmie,

Thank you very much for your thoughts and they are on the right 
direction, at least from my viewpoint. I'll use your questions below for 
my ESUG presentation to show the vision of both Aida and Scribo. Well, 
I'll try to show and let discussing the vision here as well if time will 
permit.

About question from subject, let we therefore stay with Smalltalk Web 
Application Server.  Technically it is much more appropriate as you and 
Nicholas pointed out, while me and Nico had some worries from marketing 
standpoint. Also on Wikipedia we are among Web frameworks, there is even 
not a serious page for web app servers!

Best regards
Janko

Jimmie Houchin wrote:
  

nico wrote:
    

Hi,

Of course I also understand Jimmie's point. But I still think that it's
a mistake, at least for marketing reasons.
      

Understood. :)

However, there is no clearly defined use of the phrases "Web Framework", 
"Fullstack", or "Web Application Server". The definition unfortunately 
varies depending on the tool to which it is being applied. Which is why 
I believe that on the "About AIDAweb"... page(s) that a good description 
as to whatever expression we use means.

With regards to marketing. Who are we marketing to? What marketing 
battle do we hope to win?

If Seaside v. AIDAweb/Scribo, then I don't think it is a problem. Each 
are different and distinct with different personalities. And with regard 
to Framework v. Application Server, Seaside is in the same boat as us.

I believe that with regards to marketing we have a lot more to overcome 
than ambiguities between Framework v. Application Server.

Why use Smalltalk?

That is the first barrier to entry.

Once that is overcome, then we have somebody who can listen and 
understand the whys behind Seaside or AIDAweb.

We will never win with people who come to AIDAweb who are looking to 
swap in and out there favorite pieces and parts and tools for the job. 
Its hard enough to talk to the Seaside community and they are one of us. 
Janko posts on the Seaside list about adding a feature to Swazoo and 
gets blasted. Why add this feature? We have Apache? (All bow down to 
Apache)  Ugh!

One thing I love about AIDAweb is Janko's vision for a complete 
Smalltalk stack for web development. Turtles all the way down, as Avi 
says. Nice!

We are getting closer to offering a good and reasonably compelling tool 
for the developer. I would love to see us get to a better out the box 
experience for those who aren't programmers (yet) but are very good with 
computers and software and we can enable them with quality software and 
a good web UI to get the job done.

The app I am currently working on is currently implemented in 
SharePoint. Yuck! But it was chosen (not by me) because it enabled a 
certain set of people to get a job done. AIDAweb/Scribo as of yet 
wouldn't quite enable them in the same way. But it can, and I hope it 
will soon.

Out of the Box experience is the one thing which has so strongly tried 
to pull me to Plone. Being the end-user of an application (CMS) instead 
of a developer with a CMS is very appealing. It is very enabling.

I am very much a power user. I very much prefer using an application, to 
developing the application. I am hoping that Scribo will reach the place 
that it enables people like me. I know Scribo isn't there yet. And that 
I might have to help to get it there. But if there is no desire for 
Scribo to enable application users. Then I may be in the wrong place and 
need to go to Plone who very much does. But I care about the back-end 
technology also which is why I currently prefer AIDAweb/Scribo.

Any way, I believe marketing to people looking for a Framework will 
frequently draw people in only to see them leave when they learn that we 
don't embrace or enable their favorite tool. They will see our vision as 
too narrow and non-inclusive. Whereas I find it liberating to not have 
to worry about all that other technology and what to pick and how to 
assemble and get working together.

A quote from the Seaside list:
"It took me at least 3 times as much time to get Apache up and running 
with Seaside as it did to actually write the Seaside application."


So the question really is:
Who are we marketing to?
Who do we want to become a part of this community?

What is AIDAweb?
Who are its users?

How do we get the word out to them?

Regardless of which phrase gets chosen, it will need to be well defined 
for the context of AIDAweb.

To me, we want people who are:
Opinionated :)
They love Smalltalk, or are at least open to learning and loving Smalltalk.
They love Turtles all the way down.
The more they can do with the chosen tool, the better.
The less they have to look elsewhere to solve the problem, the better.

And the nice thing is ...
That if you want to put Apache in front of AIDAweb, you can.
That if you want to use PostgreSQL for persistence, you can.
Nothing stops you. Nothing even slows you down.
It just isn't the required path of the tool.
You are fully enabled without such tools or requirements.
Nothing else is required out of the box.

So what is our vision statement?
Who are we?
What do we want to be?
Who do we enable?
What do we provide?
Where do we want to go?

The better we can answer all of these questions the better we can choose 
who to market to and how.

One of the fantastic things to watch about the Plone community is how 
they choose to organize themselves. How they go about deciding a vision 
for the future of Plone. A vision for Plone, its development, its users, 
its community.

I know the AIDAweb community is small. But we don't have to think small. 
Without a vision the people perish. Without a vision, no one can come 
alongside and become one of us. With a vision clearly expressed and 
written down. People can come alongside, take a part of the work and run 
with it. Without such, people are only somewhat walking in the same 
direction. If their works happen to complement the whole and help the 
community, great. If not, oh well. No common vision. No common goals. No 
common achievements.

I am not a strong developer. But hopefully I can be a strong motivator 
and help where I can. But without the vision, I can't say that AIDAweb 
is going the direction I want to go. It currently isn't at a place I 
want to be. But can I help it get to where it wants to go? And do I want 
to go there also?

If I can't answer those questions, then I am better off with Plone even 
if the back end development is a little harder. Because of community, I 
will have much less of it to do. Only those content types which aren't 
available out of the box.

My apologies for such a long email.

I am just trying to exhort this community toward creating and moving 
towards a common, well written vision as to who we are, what we are 
doing, where we are going and why you want to join us in this adventure.

Jimmie
_______________________________________________
Aida mailing list
[hidden email]
http://lists.aidaweb.si/mailman/listinfo/aida