| Author |
Message
|
| smdavies99 |
 Posted: Mon Jan 07, 2013 9:44 pm Post subject: But the 'requirement' is... |
 |
|
Jedi Council
Joined: 10 Feb 2003
Posts: 6076
Location: Somewhere over the Rainbow this side of Never-never land.
|
This is a general point
I have seen a lot of posts in recent months where the OP is 'required' to do something using a mehtod/node that would not be the first choice for other experienced developers.
for example,
'I must do this using a JCN'
Sometimes a solution can be found but there are others where the 'requirement' is clearly wrong. I know that in some cultures questioning what you have been told to do is not normal practice but the thing with Message Broker is that the old saying 'there is more than one way to skin a cat' really holds true.
Beating your head (As a developer) against a brick wall when trying to do something when you have been shown by 'us here' that there is a better way to do it is not a good way to enhance your job satisfaction levels.
I'd like the thoughts of the community as to how we educate the people who tell the devs that they must do this in a certain way. Perhaps those architects/designers might like to sign up to this forum? They might learn a thing or two.
_________________
WMQ User since 1999
MQSI/WBI/WMB/'Thingy' User since 2002
Linux user since 1995
Every time you reinvent the wheel the more square it gets (anon). If in doubt think and investigate before you ask silly questions. |
|
| Back to top |
|
 |
| kash3338 |
| Posted: Mon Jan 07, 2013 10:33 pm Post subject: Re: But the 'requirement' is... |
|
|
Shaman
Joined: 08 Feb 2009
Posts: 709
Location: Chennai, India
|
| smdavies99 wrote: |
| I'd like the thoughts of the community as to how we educate the people who tell the devs that they must do this in a certain way. |
Good point smdavies.
One way to do this is to show/prove them the solution with the use of best nodes possible than the one that is part of requirement. It would not be possible to implement the complete solution and prove this, instead we have can a set of POC's (Proof of Concepts) for different scenario's which shows what kind of nodes best suits the requirement type.
Documentation of these POC's along with the links to the IBM documentation which support these would be great asset for every project. These should be in place during the early stages of a project so that during the design phase it helps a lot to take proper desicions.
| smdavies99 wrote: |
| Perhaps those architects/designers might like to sign up to this forum? They might learn a thing or two. |
To drag people into a forum is not a easy way I feel. It should be of their own interest. Only thing that can be done on this is to have a separate links to certain imporatant design decisions that were discussed in this forum or for few important topics which can be shared with these architects as reference. |
|
| Back to top |
|
|
| zpat |
| Posted: Tue Jan 08, 2013 3:10 am Post subject: |
|
|
Jedi Council
Joined: 19 May 2001
Posts: 5866
Location: UK
|
Yes, a business requirement is what outcometo achieve.
The architect should decide which way to do it at the design level.
The developer should decide how to do it at the code level.
I quite often hear things like "the requirement is to FTP a file from server A to server B".
My answer is no, the requirement is for application A to communicate with application B. The non-functional requirements that I then automatically add, are to do this in the best way to provide a continuous, secure, flexible, recoverable, reliable, scalable operation.
The answer is usually MQ with a sprinkling of WMB 
Last edited by zpat on Tue Jan 08, 2013 3:49 am; edited 1 time in total |
|
| Back to top |
|
|
| nathanw |
| Posted: Tue Jan 08, 2013 3:25 am Post subject: |
|
|
Knight
Joined: 14 Jul 2004
Posts: 550
|
To quote Cool Hand Luke
"What we've got here is a failure to communicate"
The requirements as said by zpat are to carry out an action.
The action can be done in several methods, so the initial design should be revisited once developers get involved and then discussed and the best fit solution decided upon.
I have in my past been involved in projects from scratch and then come into projects later one.
Every time I have said is this the best solution?
Sometimes it is and others, even though I have not liked it, it isn't.
Designs can be fluid and change based on input and changes to architecture etc and the developer / pm / team lead should always be in the loop so that there is 2 way input and therefore the best solution can be made.
As a developer amongst other things I have sometimes had to bite my lip initially and then knuckled under but not before expressing problems and discussing potential issues.
Basically, if there is a problem or a change discuss it!!!!
Designs are not set in stone in my view and can always be changed so that the best fit is made.
_________________
Who is General Failure and why is he reading my hard drive?
Artificial Intelligence stands no chance against Natural Stupidity.
Only the User Trace Speaks The Truth  |
|
| Back to top |
|
|
| mqjeff |
| Posted: Tue Jan 08, 2013 3:30 am Post subject: |
|
|
Grand Master
Joined: 25 Jun 2008
Posts: 17447
|
| nathanw wrote: |
| Designs are not set in stone in my view and can always be changed so that the best fit is made. |
While I agree, there's nothing less useful than the new guy bringing up all the same issues that the team has rehashed ten times over the last few months.
"Yes, we know this is ugly. Yes, we know. We told them that. We just need you to *build* *this*". |
|
| Back to top |
|
|
| smdavies99 |
| Posted: Tue Jan 08, 2013 3:36 am Post subject: |
|
|
Jedi Council
Joined: 10 Feb 2003
Posts: 6076
Location: Somewhere over the Rainbow this side of Never-never land.
|
| nathanw wrote: |
As a developer amongst other things I have sometimes had to bite my lip initially and then knuckled under but not before expressing problems and discussing potential issues.
Basically, if there is a problem or a change discuss it!!!!
|
All good points Nathan and is the way we do it (most of the time)
However the point that I was making is that in some cultures/organisations this is just not being done. Hence the stream of posts that begin 'I am required to.... using xyz' AND the OP does not seem to be able to take advice from the good,bad and ugly who post help here
_________________
WMQ User since 1999
MQSI/WBI/WMB/'Thingy' User since 2002
Linux user since 1995
Every time you reinvent the wheel the more square it gets (anon). If in doubt think and investigate before you ask silly questions. |
|
| Back to top |
|
|
| nathanw |
| Posted: Tue Jan 08, 2013 3:59 am Post subject: |
|
|
Knight
Joined: 14 Jul 2004
Posts: 550
|
mqjeff I agree and have been on both sides of the fence there
_________________
Who is General Failure and why is he reading my hard drive?
Artificial Intelligence stands no chance against Natural Stupidity.
Only the User Trace Speaks The Truth |
|
| Back to top |
|
|
| lancelotlinc |
| Posted: Tue Jan 08, 2013 5:59 am Post subject: Re: But the 'requirement' is... |
|
|
Jedi Knight
Joined: 22 Mar 2010
Posts: 4941
Location: Bloomington, IL USA
|
| smdavies99 wrote: |
| 'I must do this using a JCN' |
... or 'I must do this using a mapping node because the IBM sales rep said you could develop message flows without writing code.'
The choice of nodes should be left to the developer not the sales rep or the management.
_________________
http://leanpub.com/IIB_Tips_and_Tricks
Save $20: Coupon Code: MQSERIES_READER |
|
| Back to top |
|
|
| NealM |
| Posted: Tue Jan 08, 2013 9:13 am Post subject: |
|
|
Master
Joined: 22 Feb 2011
Posts: 230
Location: NC or Utah (depends)
|
Regarding JCNs, a little history. When I used to be part of IBM's team in mentoring folks at customers new to WMB, sometimes IBM couldn't even get in the door because the prospective company had decided it was a Java-only shop and didn't want its people to have to learn a "proprietary language". Thus the addition of JCNs, etc to gain entrance into these shops.
Since then, some have seen the light when they realize how straightforward ESQL can be, and have opened up to various cat skinning techniques. Others however haven't and will only allow Java. It has become a sort of religion, and the unfaithful had better not try to spread their heresy. |
|
| Back to top |
|
|
| Vitor |
| Posted: Tue Jan 08, 2013 9:53 am Post subject: Re: But the 'requirement' is... |
|
|
Grand High Poobah
Joined: 11 Nov 2005
Posts: 26093
Location: Texas, USA
|
| lancelotlinc wrote: |
| smdavies99 wrote: |
| 'I must do this using a JCN' |
... or 'I must do this using a mapping node because the IBM sales rep said you could develop message flows without writing code.'
The choice of nodes should be left to the developer not the sales rep or the management. |
And yet in many instances it's not being so left. For all the reasons discussed here. So to return to the start of this thread; what's the best way to tackle this?
_________________
Honesty is the best policy.
Insanity is the best defence. |
|
| Back to top |
|
|
| lancelotlinc |
| Posted: Tue Jan 08, 2013 9:56 am Post subject: Re: But the 'requirement' is... |
|
|
Jedi Knight
Joined: 22 Mar 2010
Posts: 4941
Location: Bloomington, IL USA
|
| Vitor wrote: |
| lancelotlinc wrote: |
| smdavies99 wrote: |
| 'I must do this using a JCN' |
... or 'I must do this using a mapping node because the IBM sales rep said you could develop message flows without writing code.'
The choice of nodes should be left to the developer not the sales rep or the management. |
And yet in many instances it's not being so left. For all the reasons discussed here. So to return to the start of this thread; what's the best way to tackle this? |
Assert your authority as the Subject Matter Expert and provide sound justification for your rationale.
Although this strategy would be hard for those that are not experts.
_________________
http://leanpub.com/IIB_Tips_and_Tricks
Save $20: Coupon Code: MQSERIES_READER |
|
| Back to top |
|
|
| PeterPotkay |
| Posted: Tue Jan 08, 2013 6:18 pm Post subject: |
|
|
Poobah
Joined: 15 May 2001
Posts: 7722
|
Fight fire with....water, not fire.
When they come to you and say "Do XYZ, don't ask why, just do it." going back to them with "I think it should be ABC" is the same as "I want ABC, don't ask why, just accept it". Now you're in a urination competition and everyone digs in.
Explain in detail, in writing, why ABC is better. Or, it just might be that you don't know for sure - so do both, test, evaluate, think  , and then come to a conclusion.
_________________
Peter Potkay
Keep Calm and MQ On |
|
| Back to top |
|
|
| fjb_saper |
| Posted: Wed Jan 09, 2013 2:30 am Post subject: |
|
|
Grand High Poobah
Joined: 18 Nov 2003
Posts: 20756
Location: LI,NY
|
now that everybody has had their nice say let me add some fuel to the discussion:
Part of the misdirected implementations may well come from a few things IBM:
Think about the differences in broker mode: adapter mode for instance.
Know that some broker modes will only accept JCN transformation nodes and not allow you usage of all nodes.
Thus we might see requirements to achieve some outcomes on the "cheap" when a full version of the broker would be better suited to the tasks needed.
Just my $0.02 
_________________
MQ & Broker admin |
|
| Back to top |
|
|
| zpat |
| Posted: Thu Feb 21, 2013 3:13 am Post subject: |
|
|
Jedi Council
Joined: 19 May 2001
Posts: 5866
Location: UK
|
Just to add my view based on experience. Message flows that stick to MQ and ESQL usually work without any problems.
Message flows that use Java and/or web services tend to have problems more frequently.
The big problem is new developers who are never properly inducted or trained in WMB. All they know is Java. Give them JCNs and they will try to code everything in Java and also use non-IBM JAR files all over the place.
Personally I would have a rule saying use ESQL only, unless it cannot do the job - but some people are convinced that IBM only add new features because they are better than the old ones. Wrong! |
|
| Back to top |
|
|
| smdavies99 |
| Posted: Thu Feb 21, 2013 4:50 am Post subject: |
|
|
Jedi Council
Joined: 10 Feb 2003
Posts: 6076
Location: Somewhere over the Rainbow this side of Never-never land.
|
| Quote: |
| all they know is Java |
That in itself is dangerous. Broker is not an APPServer and coding stuff in the same way as for WAS/JBOSS etc is asking for big trouble.
I agree with the ESQL unless there is no other way rule although two .Net coders I know would venture to disagree with you.
_________________
WMQ User since 1999
MQSI/WBI/WMB/'Thingy' User since 2002
Linux user since 1995
Every time you reinvent the wheel the more square it gets (anon). If in doubt think and investigate before you ask silly questions. |
|
| Back to top |
|
|
|
|
