Web based HMI - your thoughts

You are talking about Web Browsers only.
But i think this is not the right technology for HMI/SCADA, because it uses a connectionless protocol.

Our project
http://pvbrowser.org
uses a connection oriented protocol on top of the TCP transport level.

We have developed a completely new browser in C++ that uses Qt widgets instead of HTML. The client and server exchange strings that are terminated by a newline. The browser will interpret that string and call the according Qt method. In the reverse direction the browser sends events (also strings) to the server, where this is handled in a event loop.

There are no "round trips" within the protocol, thus getting a higher performance.

Our solution provides much higher update rates than can be achieved with connectionless http protocol.

But "http://" is also available within pvbrowser, because it can use WebKit. You can link from "http://" to our "pv://" protocol and vice versa.

You can integrate any protocol for data acquisition into pvbrowser, because we use a separate daemon for data acquisition, linking by "shared memory" and "mailbox" to the pvserver.
See:
http://pvbrowser.de/pvbrowser/pic/prinzip.png
Many PLC and fieldbus protocols are already implemented like this.

Regarding the ActiveX/broswer comments - I don't think splitting hairs in browser market share matters much. IE has strong market share, as does Firefox, and there's plenty left in Opera, Safari, etc. So what?

What is important is that ActiveX is an old technology that, quite frankly, sucks and presents gaping security vulnerabilities. Microsoft is pushing Silverlight. JavaFX and Flash are strong. Even AJAX works wonders with any of the above browsers and NO plugins. Hopefully HTML 5 delivers and knocks the field down. All of these can be made to work with any of the leading browsers on any platform.

Why endorse an old Microsoft technology that they're not interested in?

-----
Nathan Boeger, CISSP
http://notanotherindustrialblog.blogspot.com

I'm sorry, but you using statistics from 2008 - As of June 2009, Explorer has 40.7% and Firefox has 47.3%. With the recent EU ruling regarding bundling of IE, many expect the gap to widen.

Please refer to:
http://www.w3schools.com/browsers/browsers_stats.asp

Don't trust studies/sites that are sponsored by Microsoft. W3 is an INDEPENDENT organization and is the largest web developers site on the net.

Actually the numbers I was quoting came from a study done independently of MS in May of 2009. But you seem to have your mind made up. W3 has a significant bias toward Mozilla...as significant a bias as the MSMUG has toward IE8.

I'm done with this thread.

Walt Boyes
Editor in Chief
Control and Controlglobal.com
www.controlglobal.com
Mailto:wboyes [at] putman.net
Read my blog SoundOFF!! At www.controlglobal.com/soundoff

In repy to Walt Boyes: W3Schools is not W3. W3 is web standards committee that Microsoft happens to be a part of. W3Schools is a web site with focused content that earns money from selling advertising.

The audience for the w3schools site is web developers, who are not typical of the market as a whole. W3Schools however is oriented towards older style web development for maintaining simple business apps, which means their stats are likely not even typical of web developers in general. They probably have a higher proportion of users with MS IE than is typical for web developers.

You should have noticed by the way that W3Schools is running everything using ASP (Microsoft's product for dynamic web sites), which is rather unusual for a public web site. If the site operators themselves have a bias, it is towards Microsoft.

However as I said before, measuring browser share is very difficult. I think that your original numbers were not too far off however as an overall average for the US market. MS IE has been losing market share at about 3 percent per year for the past 10 years, and that will probably continue for the foreseeable future.

In repy to <b>Michael</b> and <b>Walt</b> Can't we all just get along!?

Gentleman - browser market share isn't worth arguing over. Let's keep it productive here.

Let's talk about the future. So what about HTML 5...?

----
Nathan Boeger, CISSP
http://notanotherindustrialblog.blogspot.com

In reply to Nathan Boeger: Mr. Boyes and I are not disagreeing with each other. I've been saying that his market share figures are roughly correct for the US, and provided some additional detail. I did point out that he was confusing W3C with W3Schools, but I don't think that was central to any of his arguments with other parties.

As for HTML 5, I'm looking forward to doing things with it. What is going to matter is not so much what is in the formal W3C spec, but what the vendors other than Microsoft agree to do informally. Microsoft played dog in the manger during the standards process and wouldn't agree to more than a minimal change from HTML 4 because their browser is technically so far behind everyone else's.

However, everyone else (Mozilla, Apple, Google, Opera, etc.) is getting together outside of W3C to agree on the features that will really matter. These will become the defacto standards that advanced web applications will be written to use. Another thing that will matter a lot but isn't part of the spec is the fact that everyone (except Microsoft) is introducing new, much faster Javascript systems. Modern web applications will be able to use Javascript in much more complex ways.

Something else that will matter (that isn't in the official spec) is that SVG (vector graphics) will be going into regular HTML (except for MS IE of course). Everyone (except MS IE) already supports SVG in xHTML, but I think it will be more widely known and used when it is in regular HTML.

I've been using SVG plus Javascript and CSS for HMI and I think it's great. You can do pretty much anything your imagination allows, including live interactive graphics. I think that is the direction that most HMI applications should be going in. I think that people in our business who don't learn how to make web pages are going to be like the people who knew relays but never figured out PLCs.

Here's something else. How about viewing and editing PLC ladder diagrams via a web browser? Why do we need to install large complex programming packages which always seem to be out of date just when we need them the most?

I have a soft logic project (http://sourceforge.net/projects/mblogic/) which currently can display off-line ladder diagrams in a web browser using SVG. For on-line troublshooting I will be adding live animation of the ladder using the same techniques I am already using for the HMI. I'm currently working on an integrated web based IL editor, and it looks to be very easy to do. I also have plans to do a web based ladder editor using similar techniques. So, all this is certainly quite possible, and not even that difficult.

Did you see the Google Wave video when they demoed it? If you didn't, it's well worth seeing. They did some really amazing stuff using HTML5 features that already exist in major web browsers. It's hard to believe they are working in a web browser. Google Wave is for e-mail, messaging, and collaboration, but I think the web browser programming techniques they showed can be applied to industrial applications as well.

I think the future of almost all industrial software is going to be based on server applications (on PCs and embedded systems and as hosted apps on the vendor's web site) and web applications running on standard web browsers. The problem that the major vendors are going to have will be figuring out how to convince people to pay $2000 each to look at a web page. I think the change in business model has the potential to really shake up the whole industry.