Web Dynpro Java – Is it SAP’s latest modern legacy technology?
Hello All,
The decline of Web Dynpro Java ( WDJ – Java. Web Dynpro Java – Thats a bit of 007 ) as SAP’s preferred UI technology is well documented. From the Teched in which SAP basically disowned WD Java, we WDJ developers have been fighting hard to keep it alive( That really makes me feel like a soldier 😎 ). I have myself written blogs to enthuse faith in this technology and also retain the belief in its latent potential. However, its time for me to resign to the fact that WD Java could seen be a thing of the past.
Divine Intervention or Internet-inspired intervention: Modern Legacy Systems
During a random internet surf, there was a video on Modern Legacy systems. The presentation basically highlights the importance of a legacy system. It also suggests a change in our attitude towards a legacy system. We are required to appreciate the need and usage of legacy system for in this ever changing world, it does not take long for a system/technology to become legacy. Well, it tingled my SAP love interest – WD Java 😳 .
WDJ is a stable UI technology. Now, even its alternatives by SAP seem more motivated by non-technology factors. So, agreed WDJ may not be used in the new systems but there is no reason for it be replaced in the existing systems. Re-writing the apps may sound simple but they are far more expensive than maintaining the existing ones. Re-writing may not always produce the desired effect and soon that could also become legacy, so why do it? Finally, its about the trade-offs.Is it worth the effort? Does it make life better? and at what cost?
Now, as WDJ developers we have an obligation to leave a good legacy behind. That’s is critical for its survival. Well, my intention is to suggest the we need to retain the necessary unless a suitable/viable/disruptive alternative arrives.
Some pointers to leave a good leavcy:
1. Documentation – Now that’s a surprise!
– We need to document all the aspects of the existing system from the user lists, dependency lists, fail-over/backup plans, system dependencies, architecture diagrams and more. Remember: Code cannot speak for itself. We need to write it down.
2. Tests – Document the tests
– Not just automated unit tests but all possible/probably testing phases to be performed and well-documented.
3. System Design documents
My resolution: Hopefully to leave a good legacy and wish you one too.
My request: Leave a good legacy behind, for legacy is not only about old.
All the best.
Regards,
Sharath
man,you see, it is not that hard to admit it.
please stop to confuse other
don't insist just for your personal belief
Well, its not to confuse and its neither about my personal belief.
Its just not be blinded and always have an open mind. An open mind is to accept on both sides of the coin.
The point in this blog is about moving forward but in a hurry to move forward, we tend to forget about the things we leave behind.
As a legacy system/technology which will be used by customers for sometime, its our obligation to leave in best possible manner.
The same applies to all technologies and not just WDJ.
Regards,
Sharath
Hi Sharath,
Nice blog. I dont think WDJ is done and dusted. There are huge migration challenge now in moving to NW 7.3. Even if there is a tool for NWDI you know how long it gonna take.
I do think any UI which cannot run on mobile device will problems to survive in coming days. Also the WDJ framework was too heavy to my knowlege. The amount of java classes it generated was never looked a clean job.
As usual sap has release loads of new jargons ..Fiori,/SAPUI5 - Give a try !
-V
Absolutely right Vivek.
With new technologies, old ones will have have to be moved out. But, not completely. As you rightly said, migration and maintenance of existing systems will always exist.
Just like customers who are still on R/3 4.6C. 🙂
Regards,
Sharath
will you still encourage new customer to implement 4.6c?
if not, why you still recommend people to use wdj?
Jun,
You are missing the point.
You are right. We will not recommend 4.6C and maybe WDJ for new customers. But, they are customers who have thier business on these technologies. It will stay for some time ( for various other reasons ).
Hence, a need for people with exp. in those technologies is required. That is my argument.
Regards,
Sharath
i think you are still encouraging people to use it, says it is not dead.......
Well, I am not encouraging people to use it, per se. But, I am trying to clear the air. Yes, it will not be used for new ones but some may continue to use it.
A sense of panic has been set in, mainly due to half-baked information. The whole truth is bitter but not all gloom. 🙂
Regards,
Sharath
I have said it before so I may be repeat myself here, WDJ is far from retired. It's a proven framework, there are still lots of new developments taken place in the (EU) market and as such there is amp[le work for a WDJ developer.
WDJ is indeed not further developed by SAP, but it is certainly not end-of-life.
Thank you Robin.
In Asia Pacific, we are lagging behind mostly due to poor communication with the customer.
As a customer, he/she will be more worried about new from SAP. As consultants its our duty to clear the air but we have failed. Instead, we have been carried away by the news and have propagated the fear( if I may say so ).
Regards,
Sharath
Hi Sharath,
very nice thought provoking blog.
You know my thoughts on abstracting the Web User Interface from the Critical Business System.
If WDJ is on the decline, then there will need to be WDA 'CE's for abstracting WDA composite applications and user interfaces away from the critical business systems.
Time will tell.
Best regards,
Andy.
Thank you for the feedback.
WDA for CE seems far fetched at this moment. The way things are going, UI5 could be a suitable alternative.
But definitely interesting time ahead. Patience would be a good virtue. 🙂
Regards,
Sharath
Hi Sharath,
if you don't want the WebUI running on the critical business system, then you need some kind of application server infront of the critical business system, and if WDJ is not the preferred direction then the way will be an abap stack used as an web application server, ie, an ABAP 'CE', abstracting the Web UI away from the critical business system and using RFC to connect to the master data of the critical business system, like in the most secure PLM solutions.
Time will tell, we shall see.
I agree WDJ is not dead yet.
Andy.
the information i got is the UI5 is far from productive......
Okay. That is not so good news for all new UI5 enthusiasts within SAP.
In my opinion, SAP has tried to fix something which was not broken. Knowing SAP, they will never close their doors, which is the right way to grow in business.
In the scenario where UI5, not being production ready we are limited to WDA and WDJ(SAP Business Apps).
Interesting to watch the developments in this domain.
Regards,
Sharath
Hi Jun,
Unless you can provide me a couple of sources stating SAPUI5 is "far from productive", I would kindly advice to refrain from stating such things. To me, your quotes are exactly what Sharath means by "poor communication with the customer" and " we have been carried away by the news and have propagated the fear".
The last think I know and have seen is that SAPUI5 is absolutely, genuinely productive. Just take a look at all the SAP HANA, SAP HANA Cloud and Fiori applications.
Or have a look at this blog post by the very CTO of SAP, Bjoern Goerke http://bgoerke.wordpress.com/2013/05/13/is-this-cool-or-what/ where he states:
one question, have you ever used UI5 in real life project to taste it?
Jun,
It is difficult to have experience on any new technology. The customers will be hesitant to invest and would ideally want to "wait and watch".
It all depends on the roadmap by the product vendor and SAP in this case have a clear roadmap for UI5 for atleast next 5 years.
With that in mind, it would be hard to argue against UI5 solutions. They may not be fully ready now, but SAP will ensure that they are ready sooner than later.
Regards,
Sharath
Hi Sharath,
that question is for Robin.
BR
Jun
Got that. 😯
A couple actually, one using HANA DB and one using MDM
Never understood how HTML5 is showcased as the next big thing in UI. All customers know what happened to sencha when iOS 7 come out. Also devices does not support hardware acceleration for HTML, it will be real challenge. Interesting to see the future of SAPUI5/Fiori/HTML5
-Vivek
I wholeheartedly disagree with you of it being a challenge 🙂
Just have a look for instance at this blog http://scn.sap.com/community/developer-center/front-end/blog/2013/11/12/performance-comparison-of-abap-webdynpro-and-sapui5-with-neptune-application-designer where they compare a standard cached WDA screen load time & bytes received to it's cached SAPUI5 equivalent:
0,9 MB and 20 sec responsetime (WDA) vs 885 bytes and 55ms (SAPUI5), respectively
Yes, the initial load of the full SAPUI5 framework is painfully slow on first request, but subsequent requests perform blazingly fast
WD now roots back more than 10 years. Browsers were limited and roundtrips common as everybody worked on LAN. Even after SAP introduced AJAX in WD, there is still a lot of information transmitted. Short: today everything is faster than WD.
If you want, you can use jQuery and a plugin like data tables to receive impressive speed when navigating inside a table. Just load the data once, let the browser do the rest. But that's HTML5, JSON, AJAX, and has nothing to do with SAPUI5. SAP just "benefits" from HTML5.
That said, it doesn't really matter what you use for your UI, as long as your architecture supports current and future apps: make it service oriented (like REST), and try not to mix UI and logic too much like you are more or less forced to do with WD; and many do with Javascript in HTML5. For WDJ this means: use EJBs, WS or JCo, which allow you to later switch the UI, to something like VC5 (err, status? Yariv Zur ), SAPUI5, AngularJS, etc
If you design / code today a WDJ that that is a monolithic block, where persistence layer is part of your WDJ code, data handling is done completely in the controller and you treat your application server only as a DB abstraction layer, stop now. You are creating a monster. Even with 2020 still being far away, business apps live 5 years or longer.
I agree. Unfortunately I have seen (and still see) a lot of monsters created in WDJ; it's as if CAF has never existed 🙂
And of course, no one forces you to use SAPUI5 -- like you said, you can use any frontend you want -- it's just that for SAP it's the prefered way forward
For many WDJ devs CAF never existed. They developed for 7.0 and then went straight to 7.3. So, yes, that explains a lot of WDJ apps out there, but I blame that WDJ devs too often are not Java devs: EJB, JPA, JSP are unknown to them. Worse: it was never needed by them (BAPI based model).
Well, it's not like WDA devs and the applications they code are better.
" WDJ devs too often are not Java devs: EJB, JPA, JSP are unknown to them. Worse: it was never needed by them (BAPI based model). " - You have hit the nail on the head. We have never needed and haveneither bothered about it. 😛
VC5 is planned(!) to be released as part of SP11 of NW 7.31 / SP06 of NW 7.4. Early next year. Keep your fingers crossed.
what's the SLA on SAPUI5 releases on iOS or Andriod. HTML5 will normally have issues with new OS updates from device platforms.
Also I forgot to mention the monster NWDI I think one of the reason customers found it really hard was NWDI track and its upgrade, this caused WDJ to be very unpopular with consultants and endusers. I have spend days fixing issues with tracks - really a one of the hardest product I have ever worked. The guys who developed it - clap clap!