Originally posted by: SystemAdmin

Hi

Can somebody please post some links or references which could help us to read through the comparison summary about the two products WTX and Contivo.

Thanks in advance
#IBM-Websphere-Transformation-Extender
#IBMSterlingTransformationExtender
#DataExchange

Originally posted by: SystemAdmin

Hi

We are starting up a project and involving data power in it. we are in a stage of finalizing a transformation engine. So can somebody tell which among this two (contivo or WTX) is best suited based on the factors:

1. ease of learning
2. cost effectiveness
3. transformation/processing speed

please post some links or references which could help us to read through the comparison summary about the two products WTX and Contivo.

Thanks in advance
#IBM-Websphere-Transformation-Extender
#IBMSterlingTransformationExtender
#DataExchange

Originally posted by: TarunB

I'm not entirely up to speed on Contivo's product set so won't comment on it here.

What I will say is that WTX has one of the longest histories in the integration sector and is ideally suited to this kind of work. From your list, points 1 and 3 are certainly strong points of the product.

I have used it in situations where £millions in transaction values are dependent on its functionality and resilience. High volume WTX processes are commonplace, whether back-end (e.g Oracle/SAP/MQ connected apps, etc) or front-end (web etc).

As for point 2 - the cost - some of this will depend on your licensing requirements and the remainder on the development, implementation and maintenance effort. Licensing should be discussed with IBM reps directly. The other intangible costs (development etc) tend to prove that WTX is easy to use, quick to deploy and simple to maintain. Therefore these costs will prove to be less than other such products.

I would suggest speaking with your IBM contact and arranging a demo or a proof of concept.
#DataExchange
#IBM-Websphere-Transformation-Extender
#IBMSterlingTransformationExtender

Originally posted by: jvanboga

Ease of learning..... Will be based on your development teams experience. When we implemented WTX (Mercator at the time) we had one person that got if fast and ran. Others struggled a little though. I'm sure that's the way it is with most products though.

One of the nice things about WTX is that there is a strong user community. Not sure all products have that.
#IBMSterlingTransformationExtender
#DataExchange
#IBM-Websphere-Transformation-Extender

Originally posted by: SystemAdmin

I guess a response from Contivo is ok? We are a partner of IBM's so this should be fine. I don't think you'll have a processing speed issue, but that is very much dependent on so many factors, I think you'll just have to run tests. Contivo performs very well for DataPower, and our joint customers have never brought this up as an issue.
For ease of use I think you'd have to get a demo or eval. This is one of Contivo's strengths, and we have both ex-Mercator and people who have used Mercator on staff. They think you'd find Contivo easy to learn and use. Certainly people like TarunB will have their preferences, I know he was with Mercator and Ascential and knows that product well, and even our people say it's very powerful. But I know the DataPower SE's can demo Contivo, so just have them demo the two products side by side.
As for cost effectiveness, I've looked at the TX pricing and really couldn't figure it out, so I don't know what is better. Contivo drives value through speed of design time, and typical customer doesn't have an issue with our pricing. But it will really depend on your needs.
Contivo also has a well integrated solution, having developed a solid XSLT transform solution for DataPower about 3-4 years ago. All I can say is that with Contivo you should be able to start implementing maps within a couple of weeks of getting Contivo, and even complex maps should get done in a few days. Well, that gives our side but see if you can get a side by side comparison and that should flush things out. Good luck!
Warren Davidson.
#IBMSterlingTransformationExtender
#IBM-Websphere-Transformation-Extender
#DataExchange

Originally posted by: SystemAdmin

The factor that would be in the forefront is getting used to these tools. Once we get used to WTX or Contivo, then transformation would not be a issue at all. Even the so said "Difficult WTX" would seem to be easy when we get used. Am i right??
#DataExchange
#IBM-Websphere-Transformation-Extender
#IBMSterlingTransformationExtender

Originally posted by: jvanboga

Have to agree with you on the pricing thing. IBM has really made pricing/billing a complexity.

I think the ease of learning any tool is partially based on your integration & development teams breadth of knowledge. In our case most of the developers came from a mainframe background. Those that had experience with DB, Web, XML and so on picked it up relatively fast.

Not familiar with Contivo but will prob download the beta to look at it. An interesting thing, to me, is that the snapshots on their web pages remind me of Biztalk.
#IBMSterlingTransformationExtender
#DataExchange
#IBM-Websphere-Transformation-Extender

Originally posted by: SystemAdmin

The snapshots might remind you of BizTalk, but don't try using us with Biztalk. They have a funny form of XSLT if you know what I mean.

And on the issue of 'ease of use', I would caution that you want to separate learning and speed. In general Contivo has been used as an upgrade to existing EAI tools, most typically platforms such as webMethods, TIBCO, BEA, Sterling etc... In those cases people knew the mapping tools well, they just took too long to complete the mappings. I'll give you a quick example; in Contivo you design the map, and before you'd send it to DataPower or any other runtime, you can do a unit test on the map. If it fails the Business Analyst can check for errors and change the map BEFORE sending it to the runtime developer. This has two benefits, one is that the map isn't being tested in the runtime, saving the back-and-forth process common in most environments. Secondly when you send the XSLT or Java transform code to the runtime environment the details of the map specification is captured AFTER all the changes are made, giving you accurate specification data that helps in future changes.

So, regardless of how well you learn the tool, certain products have features that make the process faster. It's tricky to figure out sometimes I know, but it is an important detail to recognize and try to flush out in your specific environment.
#IBMSterlingTransformationExtender
#IBM-Websphere-Transformation-Extender
#DataExchange

Originally posted by: SystemAdmin

To quote Contivo's

"I'll give you a quick example; in Contivo you design the map, and before you'd send it to DataPower or any other runtime, you can do a unit test on the map. If it fails the Business Analyst can check for errors and change the map BEFORE sending it to the runtime developer. This has two benefits, one is that the map isn't being tested in the runtime, saving the back-and-forth process common in most environments"

This feature is available in WTX right?? Any transformation engene must be providing this feature right??
#IBMSterlingTransformationExtender
#DataExchange
#IBM-Websphere-Transformation-Extender

Originally posted by: snutakki

Hai we did this comparison a couple of years ago , from the Speed perspective Mercator leaves the rest to dust. Contivo was better than other products in terms of speed but was far far behind Mercator.
Contivo is easy to learn and easy to use though.
Mercator on the other hand is intially hard to learn but once you know it it gets easier by time.

thanks,
cai
#IBMSterlingTransformationExtender
#IBM-Websphere-Transformation-Extender
#DataExchange

Originally posted by: TarunB

Anboss - RE: the previous post by Contivo, this appears to be describing a process rather than (or as well as?) the product. And yes you're correct - WTX is able to perform the same kind of process.

Generally speaking the development will be an iterative process in which you:-

1. Define the source and target data structures.
2. Define the mappings between source and target.
3. Unit test on developer's machine.
4. Tweak 1 and 2 if results from 3 indicate so.
--> Repeat 1 to 4 until you are happy with the output.
5. Deploy to execution environment.

This is a very simplified example but I would be very surprised if any integration tool did not facilitate this. One thing I like about WTX is that it is very quick and easy to make changes to maps, to test them and finally to deploy them.

I would recommend looking at the WTX presentation that Rep very kindly posted on this forum. It highlights key new features in WTX 8.0/8.1, such as the in-line Debugger and Map Profiler. I'm also impressed by the XML schema creation functionality. These will almost certainly enhance the development capability of your teams.
#IBMSterlingTransformationExtender
#IBM-Websphere-Transformation-Extender
#DataExchange

Originally posted by: jvanboga

I like WTX also. Been working with it for a few years now. The IBM WTX presentation is OK but I had issues with a number of things in it. Mostly minor.

The only one I'll mention at this time is that it's an old presentation with a new date. The look of the interface has changed.

Developing and maintaining TX applications is amazingly fast compared to other things I've worked with.

The biggest problem we had when we brought the tool onboard in 2001 was that the purchase price was larger that we had expected and all of our tool-based savings were soft dollars. A number of people here couldn't correlate a conservative 30 - 40% savings in development time into dollars.

We began installation of the tools on 1/17/2001. Had a consultant here for 1 week making sure that the install went right. He trained our network staff how to reinstall and configure if needed. He also prepared site-specific documentation. The development team (3 of us) then took advantage of 2 weeks of training purchased with Mercator (WTX). We also purchased 1 week of on site training.

By the end of February we had converted our systems so that all inbound enrollment files were mapped to a standardized format, which allowed us to reduce the amount of related mainframe code by 70-80%. It cut development time for new group eligibility files by 30% and our data center said that some of the things we did saved them as much as 10% of their time on related activity.

By the end of April we had a system in place that allowed people, subscribers and providers, on a national level, to query their benefit and claim status. By the end of June we were interfacing with our national organization.

By October we had all inbound electronic claims going thru the tool with similar savings.

By December we had most of the bugs out of our environment.

All that said, if I were to rewrite all of the systems today I'd make a number of improvements. When we first started there were a number of concepts we didn't comprehend well and our objective was to meet a HIPAA deadline not to make things perfect.

TX has come a long way in our shop. It's overcome a lot of its original baggage here over the years. However, unfortunately, there are still a few people that create roadblocks to using it where it works best, data transformation.

(Instead of a commission for this sales presentation perhaps IBM could change the icon tool back to something easier to read/view. Us old farts can't see that well anymore and it’s difficult to differentiate.)
#IBMSterlingTransformationExtender
#IBM-Websphere-Transformation-Extender
#DataExchange

Originally posted by: SystemAdmin

I work in the WebSphere TX group and helped to create the integration between WTX and DataPower so I can answer any questions about it. For those of you familiar with using the WTX Type Designer and Map Designer, you already know everything you need to create maps that will run on DataPower. Only one map setting needs to be changed to move from WTX as your runtime to DataPower. Once you have made that change, everything is handled for you within the UI using the same interface you're familiar with.

If you hit the Build button inside Map Designer, we will build the maps for DataPower and check the restrictions. Certain WTX features are not supported on the appliance but we check those for you at build time. When you hit the Run button on a DataPower map, the tool sends files to the appliance where the transformation is performed. The results are shown in Map Designer as usual. When you're ready to put your map into production, you simply upload a few files (.xml file for your map, .mts files for your trees, and the data to transform) then use it like any other transformation on DataPower.

There's no XSLT that ever needs to be edited or viewed. All of that work is handled between WTX and DataPower. The user would only work with the familiar concepts of Type Trees and Maps. Performance should be very good since the transformation runs inside the same DataPower XSLT engine other transformations do.

We think it's a very good solution. Those with WTX experience should be able to create and run a DataPower map in minutes. For those who haven't used WTX, it can be a good starting point and will save you from writing the transformations manually.
#IBMSterlingTransformationExtender
#IBM-Websphere-Transformation-Extender
#DataExchange

Originally posted by: SystemAdmin

Not all tools provide a separate test facility, but TX and Contivo do (bad example I guess). Question for you though, when TX generates the map for DataPower there are two files, do you know why this is (what's the second file for)?
I couldn't get a good answer, but hope you'd know. Also I was told they weren't generating XSLT for DataPower, so what do they generate? Thanks!
#IBM-Websphere-Transformation-Extender
#DataExchange
#IBMSterlingTransformationExtender