Ok, so with 2.0, we have an opportunity to do everything from scratch. We need to figure out what we want to do.

I'll start by asking everyone this: What are we trying to 'fix' with the web? For me, I see Lucid as making it easy to make your own set of web services available in a unified manner.

Steven mentioned to me an idea that he had. It was basically something that would let you install traditional webapps, like wordpress, status.net, and others, through Lucid. Lucid would aggregate everything from these apps, providing a unified login, dashboard, and search system. Personally, I think this is an awesome idea, but if we do this, I'm not sure where Lucid apps, which would use the Lucid apis completely rather then their own set of libraries, would fit in that picture. I'm also not sure how we'd be able to handle the database complexities that each webapp would require.

Another idea is that we could help developers make their own web apps by providing a rich set of apis, which is what we've been doing all along. Some things we could add in is some sort of SMS integration, as well as some of our other ideas such as offline support, and the mobile phone client.

As for the backend, we have a set of frameworks we can use:

• PHP/Zend
• Persevere
• Narwhal

Narwhal I just found recently, I think we could find it useful. You can also use it in conjunction with Persevere.

That said, we need to figure out how we're going to structure this thing. This can come after we decide what we want to do on the frontend, of course.

Note: you get bonus points for drawing up slick looking mockups.

To be honest, something key for me, would be being able to share and collaborate in real-time documents, applications and even the entire desktop. If I have a Text Document stored in my Lucid installation, I want to be able to generate embed code for it, making me able to embed it anywhere on the Internet, read-only or read-write. I want to be able to generate a URL to give to a friend, so we can explore the net and use the desktop together, collaborating on stuff. If I could do that in 2.0, I'd be happy.

jaymacdonald Wrote:

To be honest, something key for me, would be being able to share and collaborate in real-time documents, applications and even the entire desktop. If I have a Text Document stored in my Lucid installation, I want to be able to generate embed code for it, making me able to embed it anywhere on the Internet, read-only or read-write. I want to be able to generate a URL to give to a friend, so we can explore the net and use the desktop together, collaborating on stuff. If I could do that in 2.0, I'd be happy.

Hmm, that's actually not a bad idea. I really like that. There's also the whole file sync thing that we wanted to do.

psychcf Wrote:

jaymacdonald Wrote:

To be honest, something key for me, would be being able to share and collaborate in real-time documents, applications and even the entire desktop. If I have a Text Document stored in my Lucid installation, I want to be able to generate embed code for it, making me able to embed it anywhere on the Internet, read-only or read-write. I want to be able to generate a URL to give to a friend, so we can explore the net and use the desktop together, collaborating on stuff. If I could do that in 2.0, I'd be happy.

Hmm, that's actually not a bad idea. I really like that. There's also the whole file sync thing that we wanted to do.

Yeah, cool story bro.

Hello, I chance upon all these as I was checking out how others are contributing to the Cloud concept, and how others are doing compare to what I am doing myself; I would like to contribute my own two cents on this. You are only 2 of you, I would've pitch in as I'm refining my own Web Desktop Toolkit so you might not care, http://blog.thinkonline.ca to take a look.

I've read a lot here, I thought this is an interesting alternative to eyeOS. Although it does have the potential, Lucid fails quite short from my expectation. There are many small issues that should be addressed, and my thought of Cloud is not necessary aggregate others in one go, but to provide bridges to those Web 2.0 APIs, my view is that they are quite disparate, for example, learnt how Facebook APIs can be remotely invoked is very different from Wordpress, your pouring of ideas is no more than brain-storming on something that requires further refinements. For example, dragging windows have conflict, that I believed to have solved nicely myself. Then moving onto Office Tasks, Word processor is pretty much the only thing here, end-user would not care you have lofty goals with quite involving work, only on the sparseness of the Desktop, my suggestion is purely on refinement, and targeting on a small workgroup setting, instead of lofty goals intended for thousand people at once. But if you wish to have something interesting to do, Lucidize the entire Lucid Desktop website is a good idea. APIs will naturally form as common denominators are ID'd along the way. Most APIs might even be no more than lucid.* branded wrapper for PHP/JSP/ASP, so on. With special attention to data-storage.

In just two words: "pipe dreams" so far, I hope I did not offend any of you but that is what I see here.

actsai Wrote:

Then moving onto Office Tasks, Word processor is pretty much the only thing here, end-user would not care you have lofty goals with quite involving work, only on the sparseness of the Desktop, my suggestion is purely on refinement, and targeting on a small workgroup setting, instead of lofty goals intended for thousand people at once.

I think we may actually use external word processing services like google docs for the time being. Eventually we want to do something with dijit's editor, but right now we don't have enough resources. Besides, we should be focusing on the core, not on the apps. This way we can make it easier for developers to write great apps in general.

actsai Wrote:

But if you wish to have something interesting to do, Lucidize the entire Lucid Desktop website is a good idea. APIs will naturally form as common denominators are ID'd along the way. Most APIs might even be no more than lucid.* branded wrapper for PHP/JSP/ASP, so on. With special attention to data-storage.

Well, we were planning on something like this for 2.0, but on a different site. I don't want to let the cat out of the bag, but you'll see in time. 2.0 will have a pretty big shift with the APIs, you'll also see what we kick out with that in time.

actsai Wrote:

In just two words: "pipe dreams" so far, I hope I did not offend any of you but that is what I see here.

Don't worry, we love criticism, otherwise we wouldn't be able to figure out what to fix.

Hello all, We are currently working on a huge bank web application. I am the lead developer and Technical Project Manager. We decided to use PHP as a programming language and Java Script (Dojo) for the UI. This is an extremely big challenge. “Why PHP?” – Somebody may ask. Why not Java or .NET (#C) or something else more "secure"? Well, I strongly believe that everything is about the Design Patterns principles organization and discipline when you write the code. If the language has OOP support you can build almost everything. The answer “Why PHP?” has one answer: “Show me something that others can do, that PHP5 can’t.” We are using Zend Framework. This is one of the best frameworks, if not the best. This component library is “use on will” and extremely powerful. It allows us to use Agile Software Development or Extreme Programming Software Development Methodology. It includes OOP, Unit testing and refactoring. Zend Framework (ZF) has excellent support for Dojo. Our front end is a regular Web 1 application. The users are not going to spend too much time there. They may check their statements and leave the application. So having the old web application web style is OK. The pages reload and we use Dojo for the forms and to make the pages nicer. We are using unobtrusive Java Script. The back end is another story. The people from the office will spend all day long a front of this application. UI design principles dictate to use some kind of a Native Desktop looking application. I was trying to find the solution for long time. I knew the Java Script ( Dojo ) is the key and the solution. Java Script (with Dojo or another toolkit) is the way to go. Why to use Flex (Flash, AIR) or Silverlight? I don’t like the proprietary runtime environments for Microsoft Silverlight, Adobe Flash. I don't like even the runtime of Java. We have Java Script CSS and DOM. I had the idea to use the browser like the ultimate UI interface since long time. The idea was not very clear. I was thinking to use the browser tabs to simulate the Desktop application taskbar. But I had in my mind the idea of a desktop. So I naturally came to the idea of a desktop application to solve our problems. The first application like that, I have seen was created using Ext JS. It was (http://www.qwikioffice.com). I was fascinated. Only 2-3 days later I was asking myself is there something similar build with Dojo? This is how I came across Lucid-Desktop project. It is very interesting! But after the initial excitement I started to ask myself questions. Why there are not that many contributors? Am I only the one who believes this is the way to go or this is an evolutionary dead end? Why do you use Dojo 1.0? Why don’t you use Dojo 1.3 or 1.4? Can you keep up with Dojo changes? Maybe you are using already a newer version on Dojo in the repository? Or maybe there is something wrong with Dojo and the way it evolves? Not having compatibility with the previous version is good and bad. We have created already some applications for the backend of our application and they have been created with Dojo 1.3. They will look very nice in the Lucid-Desktop environment, but can we use them? I think we cannot if Lucid is using Dojo 1.0. I am strongly convinced that applications like Lucid-Desktop are the way to go. We are witnesses of a new leap. All computers in Internet (the cloud) are one supercomputer and the web Desktops or webOS are layer on top . I think Cloud Computing's Web Desktop is the way to go. It is a shame and will be too bad if Lucid-Desktop project is about to die. I would like to use this Web Desktop in our current application. Are you planning to use Dojo 1.4 in Lucid-Desktop project? Is the project evolving? It is necessary to use Agile Software Development (OOP, Unit Testing, and Refactoring) Methodology.

I think using PHP/Zend is a great idea. The newest Version of Dojo toolkit for the UI (RIA) is perfect.

In the future, this would be better in a separate thread, since this thread is planning for 2.0. You may also want to add paragraphs so it's easier on the eyes.

Anyway, yes, we use the latest version of Dojo. At the moment we have 1.3.x, but we'll push it up to 1.4 in the next.

It is kind of unfortunate that we have so few contributors, but I suppose it could be worse. We're working on things for 2.0 right now that may attract more developers. We do have several hundred installations though, and we get talked about regularly as well.

Anyways, you could definitely write a Lucid app as a frontend to your application, using Dojo widgets. Lucid will let you write large scale applications in an organized fashion. You could also write an app that would just display an iframe pointing it to your bank's webapp, the level of integration is totally up to you. In the case of the latter, you could probably implement some system to automatically log the user into the webapp as well (see the twitter applet code for info on how to do something like this, lucid.xhr can store passwords on the server for extra security).

Also, since this is a bank application, you might want to look over the login code to make sure it meets your bank's standards, maybe even send back a patch if anything is wrong. Everything should be perfectly fine with the login code though.

Keep us posted, this sounds really interesting!

WOW! That is great news! I am very happy that you are back on the project working on version 2.0!

But, why Python??? No!!! Please, don’t do that! That is a huge disappointment. I think you are making a mistake.

You just will benefit from the huge PHP Zend Framework community. Can you answer the question “What Python can do, that PHP (Zend Framework ) cannot?”. Zend Framework is a rapid development environment. You will be building your application faster and easier. You will step on a robust, well tested platform and huge community.

Yahoo and many others have based their systems on C, PHP, Java Script combination. It is all about a good software engineering, using the design patterns and principles. Any OO programming language can do the job. Zend has a huge growing community! It is the best component library and powerful framework.

I like Lucid ideas very much! I took the risk to use it as a platform for the backend of our application. We have started to work on building a lot of applications for Lucid 1.0 already. We are using Zend Framework and Dojo. I already have one post on your site ( Post #172 by wingman007 1 month, 3 weeks ago ). I was almost sure you will decide to continue the project with version 2.0 and you are going to use Zend and Dojo. This is the PERFECT combination. I was a bit nervous about the fact that there were not too many developers involved in Lucid. Also I was not sure which direction this project is going. Did we make the right decision to bed on this project as a platform? It has a huge potential! It can be used as a universal platform for the backend of all CMS, CRM etc. systems. It offers the comfort of the desktop UI to manage your business on the web. But the advantages of using Zend Framework and Dojo are obvious and I was thinking that this is the only right choice for you and you will make it.

The choice is yours!

I just cannot stop thinking about the Web Operating Systems, Web Desktops and Lucid.

The evolution is a spiral. We always return to the same point, but on another level. I remember the time writing CGI scripts on C back in 1995 thinking about the browsers as the universal UI platform. My dreams are becoming true 15 years later. There are a lot of examples in our evolution like that.

Is the idea of Web Operating Systems and Web Desktops evolutionary dead end?

I don’t think so. Following the principle of the spiral we can easy come to the conclusion that this is the next step of the evolution. There were local computers operating systems.

Is the idea of Web Operating Systems and Web Desktops too advanced today?

I think it is just about the time. There a large number of Web Operating Systems, Web Desktops. Also the standards are evolving. ECMA 5, HTML 5 that make the web browser very powerful. We will not need any more runtime environments like Adobe Flash, MS Silverlight to play video audio draw vector graphics. All this becomes very easy and accessible.

Why I like Lucid?

Lucid is using PHP and Dojo.

Very popular and easy to learn and use technologies. Dojo and PHP have great community support and evolve very quickly. Who has survived? Only the guys with easy to use simple technologies with GREAT COMMUNITY support. PHP is like a Volkswagen. Easy to use simple and extremely popular. Dojo will become.

The PHP is becoming the language of the web. Since version 5, PHP is mature OO language and continues to evolve. Zend Framework is build on top of PHP and it becomes an industry standard too. Everybody is using it. For examples CMS’s Joomla, Drupal and lately the CMS’s build on top of Zend Framework Digitalus, Cahaya. More and more software for the web is based on Zend Framework.

Java Script is the language for the UI on the client site. The new ECMA5, HTML5, CSS3 standards are giving it incredible power. We will not need technologies like Flash, Silverlight or Java to do incredible things in the web browser. We still need a browser normalization platform to hide the browser incompatibilities, but even this eventually will go away in the future. But for now Dojo has a great community support evolves very quickly and is becoming the best Java Script Toolkit.

Lucid can play the role of the Java Script Framework. Lucid brings the structure of the application.

My vision about Lucid is to become a platform for managing the backend of other web applications. The opportunities are endless. For example have you seen the backend of Joomla or Sugar CRM? They are like operating systems. There are a lot of small apps running to manage your frontend. Lucid can play this role. Lucid can glue together all this small apps allowing the user to mange his CRM, CMS, BMS etc.

The Lucid apps are “…almost too easy to create…” :). This is great!

Who will survive?

The projects with bigger community, using popular and easy to use technologies.

Please, use easy accessible, popular technologies. Soon or later they will prevail and survive. I think these technologies are PHP + ZEND Framework and Java Script + Dojo.

Please, let me know how can I help Lucid?

Well, Python is used more then you think. For example, Raindrop, Bespin, and even this site were written using Python. I think the great thing about the web is that it's not dominated by one specific technology; just like we have different browsers, we can all use the language that we prefer.

We'll definitely take this into consideration when planning out 2.0. I think we're starting to shift our plans a bit so that we can do a PHP version briefly after the release of 2.0, but we'll see how things pan out with time.

I think that this webOS would so much better if it had session management, more user settings(roles, more items for the user app(but make it optional), an icon for each app, a java-script progressbar underneath the login form instead of the redirecting text, ssl support, change the default folders for users, block/bann users or ip's, don't use commands that can be blocked by hosts for security reasons like some web operating systems do, Separate the menus more and give users a menu for: profile; theme; and other settings, easily (through the appearance app) change text and some other css options, create launchers with icons through the launcher creation dialog.

racoonware Wrote:

I think that this webOS would so much better if it had session management, more user settings(roles, more items for the user app(but make it optional), an icon for each app, a java-script progressbar underneath the login form instead of the redirecting text, ssl support, change the default folders for users, block/bann users or ip's, don't use commands that can be blocked by hosts for security reasons like some web operating systems do, Separate the menus more and give users a menu for: profile; theme; and other settings, easily (through the appearance app) change text and some other css options, create launchers with icons through the launcher creation dialog.

I think so too

1 We need a PowerPoint app 2 He's right we do need more settings...