Have Suggestions for Symposium Session?

[David H Nebinger, modifié il y a 8 années.]

I'm considering putting together a Symposium session but haven't really decided on an ideal subject. I think it's really a problem of not being able to see the forest with all of the trees in the way...

Anyway, for those planning on attending, would you mind throwing out some suggestions for a session you'd really like to attend? Something helpful to you and other developers...

And if someone throws an idea out you like, go ahead and give it a "thumbs up" vote. More votes will mean it's more popular to the community...

[Andew Jardine, modifié il y a 8 années.]

Hey David,

If I could throw one out there. I'd love to see some practical examples of Vaadin in the wild. Perhaps start with a contrast between Liferay MVC, JSF and Vaadin, and then show an example of Vaadin in use (code and demo) -- perhaps with some technical explanations of the benefits/draw backs based on your experience with it. I know that there are a lot of people out there with posts on Vaadin, hwo to use it etc, but none by anyone with over 9000 posts in the Liferay forums! ;)

[Patrick Wolf, modifié il y a 8 années.]

Hello David,

I think that you may launch a session on « How would you see the future of Liferay portal in 2025? » Bryan Cheung's talk at Paris' Liferay symposium was interesting as it highlighted today's and tomorrow's digital era and the fact that Liferay can tackle this (r)evolution. Then what else? As a famous actor selling coffee machines would say. Today's Liferay main strengths are its collaborative and social network functionalities and also its "responsive design" capabilities. There are still damning issues coming across my mind like simplifying themes development and deployment, that can be improved and may already be on version 7.
In a nutshell, the open question of this session would be what should Liferay version 10.0 be? Maybe it should be embedded in a microchip that will be injected into people's flesh? No need to identify oneself, no more LDAP or SSL configuraiton issues, only bearers of the Liferay's microchip will be able to use Liferay portal and communicate with each other.

[David H Nebinger, modifié il y a 8 années.]

Patrick: Version 10.0 will be entirely rewritten using Angular 2.

Andrew: I do a lot w/ Vaadin portlets, just not sure if there was going to be enough interest for a symposium session...

[Andew Jardine, modifié il y a 8 années.]

David -- you have no idea how tempted I am to quote Field of Dreams right now ;)

[Patrick Wolf, modifié il y a 8 années.]

Version 10.0 will be entirely rewritten using Angular 2.

No! Are you kidding? This is out of the scope of this thread but in my opinion JavaScript has always been and still is the Achilles heel of any web applications, especially when your JavaScript functions have to traverse an entire DOM tree. On top of that, JavaScript code is hard to maintain as it usually ends up in spaghetti code.
Are you already using Angular on Liferay website? At home, the JavaScript function calls on Liferay home page loads currently in 11833 ms with an average internet connection of 1000 Mbit/s.
I think today and tomorrow's web applications main issues, not only will be design responsiveness but the main challenge will be response time. Most organisations get their web infrastructure in the cloud with poorly designed websites and portals in terms of performance which run on low resource instance types to save money. Will Angular 2 take up the response time challenge?

[David H Nebinger, modifié il y a 8 années.]

Patrick Wolf:

Version 10.0 will be entirely rewritten using Angular 2.

No! Are you kidding?

Yes, of course I'm kidding. 7.0 isn't even out yet, I don't know when (if) there would be a 10.0 but that would be "inside knowledge" that I don't have.

This is out of the scope of this thread but in my opinion JavaScript has always been and still is the Achilles heel of any web applications, especially when your JavaScript functions have to traverse an entire DOM tree. On top of that, JavaScript code is hard to maintain as it usually ends up in spaghetti code.

True, but if you use structured development techniques you can mitigate some of the spag code issues. Building on libraries to simplify and reduce the amount of hand-coding you take on will also help.

Are you already using Angular on Liferay website? At home, the JavaScript function calls on Liferay home page loads currently in 11833 ms with an average internet connection of 1000 Mbit/s.

Played with it but not much beyond that. I know that Xtivia is doing a lot with Angular-based portlets and that they have been well received.

I think today and tomorrow's web applications main issues, not only will be design responsiveness but the main challenge will be response time. Most organisations get their web infrastructure in the cloud with poorly designed websites and portals in terms of performance which run on low resource instance types to save money. Will Angular 2 take up the response time challenge?

Not sure. The direction, as I see it, is really one that creates a "dynamic" yet fat client. Apps need all of the gui widgets and support (rendering, validation, event handling, etc.) yet we want it to be dynamically rendered based on the "script" of the interactions from the server side. There will be ongoing pressure to reduce network usage as more content, media, and general dependencies (people being dependent upon the net) grows, you have the whole IoT that will be adding traffic, networking of automobiles and their uploading data, ...

Compression will only get you so far, so the only options left will be to slim down to just the script transmission. To do that we'd have to stop transmitting the big javascript frameworks (and their css) across the wire, so the browsers will need to "fatten up" to include some of that rendering logic. When you think about it, it kind of makes sense - why are all of our websites sending javascript and misc html for a nav bar that (for responsive) changes to a hamburger menu? Sure we want to style them separately, but functionality remains pretty much standard across all of the implementations.

This is all pretty much complete speculation and I have to admit that I could be off my rocker...

[Patrick Wolf, modifié il y a 8 années.]

Thank you for this disclosed information about Liferay's future.
One idea of session may be on « Let's have your Liferay on cloud nine! ». with some best practices on how to set up Liferay as a SaaS multi-tenant collaborative platform in the cloud with some insight on JGroups and RELAY2 as a bridge between n different JGroup clusters. I assume that many organisations would like to have a cluster of Liferay in the cloud scattered among different region zones.

[Olaf Kock, modifié il y a 8 années.]

David H Nebinger:

Patrick: Version 10.0 will be entirely rewritten using Angular 2.

That will be sooo 2016 by then.

Obligatory (somewhat) productive addition (so that this thread doesn't get totally hijacked): "How to manipulate Liferay's database - step-by-step instructions". But then, I'm still unsure about what to do the last 29 minutes, after stating clearly "YOU DON'T!" (That's 30 minute sessions, right?)

[Jack Bakker, modifié il y a 8 années.]

David H Nebinger:

Andrew: I do a lot w/ Vaadin portlets, just not sure if there was going to be enough interest for a symposium session...

I liked how Vaadin did it in LRNAS2014 (Boston) : at a next door hotel via a meetup where they provided beer.

I was introduced to Vaadin at LRNAS2011 (Virginia) by Artur Signell's presentation and continue to love it for AJAXABLE desktop browser admin-ish apps but still choose MVCPortlet for use cases requiring responsive across desktop/tablet/smartphone.

I would love to see a good Vaadin session at LRNAS2015(Chicago) even without beer.

More on security like Olaf has done would be good too, including riding better the EE patch train while properly maintaining hooks.

OSGi, MicroServices or not, from marketing design into theme/layout, intro to clustering, how to learn from Liferay source, ...

[David H Nebinger, modifié il y a 8 années.]

Wow, that's a lot Jack. I think, however, that I should find just one topic to cover (assuming they let me present ).

[Jack Bakker, modifié il y a 8 années.]

David H Nebinger:

Wow, that's a lot Jack. I think, however, that I should find just one topic to cover (assuming they let me present ).

I don't know man, you are experienced in Liferay at so many levels. Perhaps Andrew's hidden reference to movie quote might help guide you. A portlet tech comparison might be valuable (MVC, Vaadin, JSF) sharing common services I would expect

[Jack Bakker, modifié il y a 8 années.]

I actually think a session on how to set up and get started on a 'cool' dev enviro would be fun. I recently bought for only ~$1250 CDN a NUC that is 13.7 x 12.5 x 11.5 cm with i7 then 16 GB RAM, 500 GB SSD. Not for me but for staff, but is a rockin' box small enough to fit into carry-on luggage.

The "kick-start" dev enviro and intro with brief compare of portlet types with shared services, and very brief git/maven might be a good topic...

[Jack Bakker, modifié il y a 8 années.]

and David, you might wear a Monster's Inc "We Scare Because We Care" t-shirt. Might have some added value wrt your approach via forums