Well it is interesting to imagine what the future of a product will be that you love very passionaltely . I am of course talking about Biztalk . Ever since WF was announced and was shipped there was a question
Is WF going to replace Biztalk .
Is R2 going to be the last release of biztalk .
What is after R2 V next . What after VNext .
I have heard V Next is going to be based on WF .
That is a lot of questions and No I am not trying to answer them nor am I making any statements . This is just some random spaghetti thoughts that I feel I need to put out of my head .
Well for one I believe BizTalk as a product is NOT going away . It is very much here and it will be here to stay .I am no way discounting the argument that WF lets you do many of the things that Biztalk does today .But my point is how easy is it for you to do something that is already available in Biztalk . BizTalk today is much more than an orchestation engine . It has a very rich adapter story . Building a workflow engine is one thing having the surround story where you can integerate that engine easily with other systems is a totaaly different ballgame . Writing an EDI adapter for your project to connect using workflow project using hand built WF vs what you can achieve with BizTalk and the EDI Adpater that ships out of the box is two different set of efforts and they are very much different . Of course when you see the 23+ adapters shipping out of the box which includes SAP , MQ , Oracle, Peoplesoft , JD Edwards to name a few it is easy to see why I feel that way . Another importnat reason is today enterprises want to make thier integeration as smooth and integerated while being very cost effective . For this they need an integerated help desk . Case in point some enterprise EDI customers who had the Covast accelerator was not happy with the covast support that they recived compared to the level of support that they get for BizTalk . Considering EDI is not going anywhere with the advent of new standards like Rosettanet and eb-XML on the contarary EDI is growing in momentum and is growing about 6% YOY . So these customers will definiely buy a more integerated suite like Biztalk and not probably build or buy external party adapters .
The next release of Biztalk after R2 code named V Next is based on .net 3.0 stack where the WF that you build in Biztlak and the Visual studio will both be the same in fact the designer could be a single one . Also with adpters moving to the new WCF Framework . One could potentially see future adapters built on WCF being consumed by Biztalk ad well as WF . So will biztalk be replaced then , in future . Probably not again becuase Biztalk also does guaranteed delivery and has a high availability model buuilt in which is scalable vertically and horizontally which is again hard to replicate . But if you are a small shop and you want to send some SAP data to your big customer . You probably can use your SAP adapter based on WCF to send data while compromising on lot of enterprise class features at a much lower cost .
Now with R2's EDI support there would be massive RFID deployment at the edge and there should be some new RFID SKU like the previous partner edition whihc wil make the Biztalk footprint much larger . I know there are other RFID early movers out there but the reason I am willing to bet on BizTalk is I remember back in the 90's every game that I used to play from Wolfenstein to Jazz Jack Rabbit you need to install the sound driver for every game . same with a word processing you needed to install the driver for every printer which means the app maker has to figure out support for each printer /sound card vendor . But microsoft came out with a unified printer and sound driver model for which you could write your app and now you can plug and play any printer that you like . This is pretty much the model on RFID now so the RFID printer manufactures I am sure is gonna love that . The fact that RFID will be packeged with Biztalk is a master stroke that I gotta give it to Microsoft .
So in relaity both WF and BizTalk is going to remain playing to their own individual strengths .
WF in the smaller integeration efforts with lower cost for trimmed basic functionality and BizTalk giving a true integeration platform or the Service bus for the much larger enterprise.
5 comments:
I'd appreciate it if you leave out comments like "Case in point many enterprise EDI customers who had the Covast accelerator was not happy with the covast support that they recived compared to the level of support that they get for BizTalk". Considering that more than 10% of BizTalk users WW also use Covast and that Covast is being used by 35 of the top 100 BizTalk users I find this a rather bold statement that just simply is NOT true.
Hi Gijsbert
I have changed my text to say some EDI customers instead of many EDI customers . I know customers who had problems trying to put covast in a farm mode and who ran into issue. This was later addressed in a fix. I still believe that Covast 2006 compared to R2 will have more features and probably a customer migrating out of biztalk - covast 2004 should go with biztalk covast 2006 . But considering Microsoft is claiming this will be a very comprehensive product question is Should a new Biztalk customer go with R2 or BTS 06 + Covast 06 ? Maybe you could have a good post on it
Thanks Abhilash.
I just posted an entry on this very topic! Check out http://www.indafield.com/BizTalkNotes/?p=31.
By the way, using Covast in a farm mode used to be "sub-optimal" with the 2004 version. Since 2006, the architecture has changed significantly and completely adheres to the BizTalk pipeline architecture and therefore has natural failover and load balancing facilities.
This is an interesting blog particularly as you have just been to the MVP meeting in Redmond and know full well the plans for BizTalk V.Next. I know its hard when under NDA but this sort of half hearted info does the community no favours.
It is common knowledge that BizTalk V.Next will replace the orchestration engine with WF (with 3,5 out in Nov. I would have touht it would be at least that version)
Also it is not clever to go Covast bashing. We in Europe have been forgotten by the R2 EDI launch - no Tradercom ect. and so will look to the excellent Covast team to support us.
My intention was not to go Covast bashing thought it could be read that way but was to question the Covast value post R2 .There were some shortcomings in farm mode see previous posts.
Things like no support for TRADERCOM which is very popular in UK makes the covast value add. Covast is moving up the value chain. So in reality smaller to mid range EDI implemntations could use R2 potentially where as big EDI customers ( especially European :-) would use Covast ) .
Your point on NDA is spot on,they do make some post too edited and then you realize people who are not under NDA know them anyway
:-). You have your profile turned off in blogger BTW. Are you from the UK ?
Post a Comment