Now, following the example, can we achieve the same thing with the help of Vaadin ?
ie does someone know if it’s possible to create multiple web sites in such a way, managing web domains, sub domain, web space ? As Vaadin would only be one ingredient, what would be the whole stack needed to provide the researched features to achieve such a goal ? As i’m interested into Glassfish, i have to search also on this side i guess but if someone already know that my goal is impossible, that would save me the unecessary research time :lol:
It were just an example, because of SEO problems, Vaadin seems not well suited (and built) for that need, that would be rather a multi application offering than a multi blog offering but you get the idea.
If this is really a blogging service, I would create the front-facing static blog part of the service with something else than Vaadin. Any lightweight MVC framework would do.
Apache Play sounds like a good alternative, but I do not have first hand experience with it. I would use Vaadin as an interface for bloggers to author their blog posts, manage their blogs and of course for administrators to manage the service as whole. I might also use Vaadin for posting on comments threads, registration and commenter account settings (if there are any). This hybrid model would have the following benefits: front facing “page oriented” part of the site would be just plain HTML and indexable by search engines. Furthermore, reading the site would be stateless. Then all the user interactions (authoring and administering) would be easy to develop with Vaadin.
Back to the original question - how to address multiple domains. I do not know - have not tried to do it. Still I am sure that there exists many alternative pieces for dynamic configuration of virtual domains for the (Java) web server and dynamic configuration of DNS server.
Blogging were the underlying example for exposing the context. With a blog, we are tempted to use a framework such as Apache Play for read part (serving pages only) and a framework such as Vaadin for edition part (authoring pages) because things can be quite easily decoupled this way in a blog that we know to be a simple blog and not a ria blog :).
My situtation is a little more …richer. It’s more like a ria content management system with some blogging capabilities, making use of portlets with pushed content (among other widgets) and inplace editing. So things are not decoupled anymore and unhappily, i don’t see where we could say “this is for play” and “this is for vaadin”. Unhappily because i’m still looking for a solution regarding SEO and stateless you mentionned; with Vaadin and GWT in general, it’s a problem.
Thanks for referencing again Play framework to me. It allows me to give it more attention this time and this may become the whole solution for the project i’m currently working on.
On another level, do you see Vaddin adapted for a project like
Betfair ? that is a betting exchange web site.
If you decide to use a hybrid solution, please report back on your experiences. I find this combination very interesting and hearing experiences would be great.
Why not. It is very interactive and SEO is not a problem for most parts.
And just to state the obvious - in most cases adding SEO to a Vaadin application is easy. As an example, see Vaadin directory. There are just two really really simple HTML pages created to make it fully searchable and to give it good visibility on Google. To try it out, turn off javascript and navigate to
http://vaadin.com/directory . These simplified pages are very efficient from SEO point of view. Also try to search any vaadin add-ons on Google to see that they end up in the top of the results page in the most cases.
In fact, i don’t see how to use an hybrid solution on this project except by searching complexity, so it would be this framework or this one. Some things i liked to see in the Play framework :
Yes, my question were in fact more related to performance possible issues when you have millions of nervous people facing the web site :))
Yes, all information related to SEO and Vaadin interest me
By searching “vaadin ckeditor” we obtain a lot of google code and vaadin forum links not Vaadin directory apparently.
So you would see Vaadin well adapted for SEO after all ? What is really interesting in the hybrid approach (Play/Vaadin) except the nice style exercice ?
When we do not have yet subscribed to a pro Vaadin account, are we allowed to see the votes on suggested add-ons ?
Really strange. When searching for vaadin tinymce, we get directory linked as too 1 result. I am not sure what is the difference between them and why google has not properly indexed ckeditor addon.
I would not say “well adapted”. I would say easy to add support in most cases
For a site with complex indexable page structure it would make sense. For a static (read only) content Vaadin doesn’t add value.
You mean votes on features? The list of votes is not visible, but list of voted features should be visible.
You’re welcome, tou’re also welcome to vote for this kind of features if you want
and if you want one on portal :), this jquery adaptation is a nice start http://www.trilancer.com/jpolite2/
That’s the main need i’m looking for on the gui part, that’s partly why i suggested it in the dev part of this web site, feel free to vote for it
Maybe easy to add
Yes i guessed that and good chances i will learn to better understand this by learning Vaadin
Yes, as you surely know, there is often this kind of “issue” when using portal capabilities as mentionned here by the way regarding Jpolite http://www.trilancer.com/jpolite/#t6 which finally manage a way to deal with it =>
-Before : “Poor SEO support by design, most content invisible to Search Engines”
Now : “Static Modules are supported now, just insert code into one of the containters…”
I tried to explain in my “portal suggestion” what is my goal but maybe this kind of “illustration” could be more explicit => Concrete5 CMS demo http://www.concrete5.org/about/trial. In a way, i could say that i would like quite the same thing with portlets. So, by following what you say, i’m perhaps more close to “complex indexable page structure” than “static”.
Today Google seems to find that add-on just fine. (but ranks google code pages over it). Maybe it was just a temporary problem in Google.
I think that implementation of a
PortalLayout component might not be that hard. Unfortunately it is not scheduled yet and thus probably wont be picked up by the core dev team in short future. Maybe you would like to implement it yourself and publish the results at
directory ?
Yes, at a time or another, it seems i won’t have choice, it’s just a question of skills as for the moment i’m on the learning path java / bonita / orientdb (which i contribute on), i think i will begin vaadin in the next few days. Do you think starting to learn vaadin 6.x give would some kind of bad habits with a release 7 maybe introducing many changes ? Or should i wait a little to start learning with this fresh new branch ?
Coming back on portal, Portaneao Posh http://www.portaneo.com is also interesting. Yes it’s php but they announced a java version while ago http://www.portaneo.com/websites/en/portal/index.php, i will contact them to know more. When time will come i will ask for what you think the best adapted choice for Vaadin in terms of “integration”.
Kynao, maybe there is a problem with your HTTP proxy: it seems to insert funny yellow circles every two words in every post and makes your posts quite hard to read. B)
You’re right, i’m afraid the proxy is my brain, i’m working on it.
Note i managed to not introduce yellow circles this time.
Thanks for helping me to improve my behaviour.
Vaadin 7 will not be that much different. We try to get it to be as backwards compatible as possible without making bad compromises.
Also - you (or anyone else) should not hold their breath waiting for Vaadin 7 release. The development has not really started yet and we probably can not ship even an early developer preview by the summer (as we have planned) Why? The team is busy with other things like Vaadin 6.6 and Visual Editor 1.0. Both of them should be
coming out really soon now .
There is also a obvious choise: Use of Java portal. Vaadin plays really well together with most standard Java portals. My recommendation would be Liferay that supports Vaadin particularly well (their tooling includes direct support to Vaadin and Vaadin framework is pre-installed in their portal distribution…
So Portaneo answered me, the java version has not been started because of the lack of sponsors.
Yes, i also thought about Java portals so i tested the most famous quite a long time ago and they all had an heavy behaviour for which i would be absolutely not confident at all with an internet audience. Maybe things have changed but i think these products are much more built for intranets, i mean would you see a Facebook using Liferay ?
There is also the features side. For a portal, i look for lightweight portals like Netvibes, igoogle, Eskobo (if you knew it when it were a portal), Pageflakes, Protopages… some framework propose some widgets in respect to this approach like ExtJS or Smart GWT.
Along the years, i were waiting for a modern portal approach and to my mind, Exo/Gatein is the only one to have it by now but i guess as the others, it may be not en engine for an internet audience.
Why? The team is busy with other things like Vaadin 6.6 and Visual Editor 1.0. Both of them should be