Which part of Vaadin.com is made in Vaadin? :)

General
1

Hi,

I would like to know which part of vaadin.com site (or maybe whole site) is made in vaadin??

I suspect that some parts are a really web application hidden behind this nice looking site:).

Thanks,
Arthur.

2

Only the
Directory
 is made with Vaadin, and the demos of course.

The main purpose for Vaadin is
web applications
 with highly interactive user interfaces, not web sites, which tend to be more or less static. HTML and non-AJAX applications are usually easiest, most efficient, and compatible for regular websites. AJAX has some limitations, such as with search engines, so it’s not ideal for everything. This is why we strongly support the use of Vaadin with portals, for example.

Nevertheless, it is possible to make entire web sites with Vaadin, if they are very “application-like”.

3

Great, thank you :))

I saw one site that is made on Vaadin and it looks like ‘classic site’.

I hope that it is only a case of time that search engines will somehow change to support sites fully based on java (applet), flash and js+ajax.

4

I hadn’t noticed it before, but it looks like Google can indeed index Vaadin applications. For example, if you enter “csvalidation” in Google, one of the first hits is
http://vaadin.com/directory#!addon/11
.

You can make a Vaadin application that provides a non-Ajax version of some URL just for indexing. See, for example, the
cached page
. The problem has been how the servlet gets the URI fragment, as it is
not passed to the server in HTTP request
.

Google
seems to support Ajax indexing
 for URIs where there is an exclamation mark token after the hash. You just need to provide a non-Ajax page for the URL where where “[tt]
#!
[/tt]” is replaced with “[tt]
?escaped_fragment=
[/tt]”. For example:
http://vaadin.com/directory?_escaped_fragment_=addon/11
. I think you can implement that easily with a
ParameterHandler
.

Google would have trouble navigating Ajax apps, but the servlet could perhaps provide also navigation and of course Google can use URLs from non-Ajax pages.

5

The Google indexing support of Directory is based on
this article
. Basically it’s done by implementing a
Filter
 that checks the request for the
escaped_fragment
 parameter and outputs the plain HTML version of the content if the parameter is found.

As the article suggests, I also tried out rendering the plain HTML version using
HtmlUnit
. I even got it partially working but not fully and the portal environment made it even trickier, so the current version was decided to be done without HtmlUnit. I’m still planning to later do some experimenting on using HtmlUnit to produce the HTML snapshot of a Vaadin application.

6

Teemu,

is this filter are in vaadin distribution or maybe I can find it in directory?

If not what you think about creating this kind of extension. Also it could be connected somehow with compiler that will generate rotots.txt file and redirects search engine to special file with special conntent.

And all that done in java using simple api.

(I have serched and I could not find any support in Vaadin for robots.txt, I think it’s time to create support for this subject as add-on).

Thanks,
Arthur.

7

I would guess that Teemu referred to [tt]
javax.servlet.Filter
[/tt] interface.

I think the indexing support could be enabled as is done in the shiny new
indexing example
. It has a Vaadin application that, in addition to the Vaadin URL, provides an indexing URL for search engines:

[list]

[]

http://magi.virtuallypreinstalled.com/book-examples/indexing#!mars
 - the Vaadin app with a state specified by the URI fragment
[]

http://magi.virtuallypreinstalled.com/book-examples/indexing?_escaped_fragment_=mars
 - the HTML page for search engines

[/list]Haven’t seen yet if this actually works in Google.

8

Hi, I read that the Directory is made with Vaadin. Is its source code available somewhere?

9

the source is vaadin copyrigh dude, may be you may pay it or donate for vaadin. just asking vaadin developers and marketting may be… :smiley:

10

The vaadin.com website is based on Liferay Portal, which can integrate Vaadin applications, like the plugin directory page.

11

The Directory application is not open source and includes code for integration with other internal systems of Vaadin Ltd.