Why can't I use pools, lanes and milestones in a Sub-process?
According to BPMN, a sub-process (of the "embedded" type) should not have pools, as these are dependant on the parent process and subsequently have visibility to the parent's global data.
In order to make use of pools, lanes and milestones diagram elements
in a sub-process, you need to define that your sub-process is a reusable sub-process.
Reusable sub-processes allow inclusion of pools, lanes and milestones elements. Note that
in BPMN, reusable sub-processes are called "Call Activities".
For more information, refer to Converting to reusable sub-process.
EDIT: According to Anatoly's feedback, we will treat "lanes in sub processes" as an improvement suggestion. From this feedback, an idea is created as: allowing-lanes-in-sub-processes.
Objection regarding lanes.
I treated this issue just like you but after a conversation with Bruce Silver and thorough study of the spec have changed my opinion.
The standard does say that embedded subprocesses don't have pools or better to say they have an implicit pool inherited from the root process. However it doesn't say that lanes are not allowed.
So please allow lanes in (embedded) subprocesses and also at the top-level process that don't have explicit pool.
Based on statistics of hundreds students passed through my BPMN training, I can witness that lanes in subprocesses is important for at least half of them. There are very respectful organizations e.g. Gazprom that have made usage of lanes a part of the corporate modeling style guide.
Switching from embedded to reusable (global) subprocess is nothing but a hack and should be avoided.
This request raises another issue: currently Modeler depicts pools and
lanes the same way. This is actually a mistake - the labels in lanes must not be surrounded by a box.
Thank you for your clarifications.
This topic will be treated as an improvement "idea".
Hello folks, unfortunately one year later this functionality is still not available in Bizagi. Greetings
I have the same problem here in my company. This restriction make me change the software. Pls tell me when this solution will be available in BIZAGI?
I'm not in position to speak for Bizagi (guess it's a matter of priority); from user's perspective it shouldn't be a big deal as soon as there is the workaround that Daisuke outlined. Just utilize reusable subprocesses - what's the problem?
By default sub-processes are created as embedded. In BPMN this type of sub-processes have no pools or lanes. Embeddedsub-processes do not allow you to add any pools or lanes to ensure conformity with the BPMN standard.If you need to include pools and lanes you can use Reusable sub-processes that according to the BPMN standard can include pools and lanes. These type of sub-processes are predefined activities. That is, they are independent and thus are created individually. Then, they can be related to the sub-process shape. In BPMN this type of sub-process is called a "Call Activity", in Bizagi Process Modeler we call them Reusable.
Could you please support your statement "embedded subprocesses have no pools or lanes" with a quote or reference to the current BPMN spec?
Thank you in advance.
There is a book called "The Business Process Modeling Notation BPMN 2.0 Destilled" of Patrice Briol who explains the same point of view as I did. Please visit this link.
Thanks for the link but it's nothing more than just an opinion.
I was on the same position once but after discussing it with Bruce Silver and other colleagues and more thorough consideration changed the view.
The truth is that embedded subprocess cannot have a pool of its own because it inherits it from the parent process.
But there is no objections in the spec nor the common sence is against drawing this inherited pool on the subprocess diagram page nor to split it by lanes depicting subprocess performers.
I'd appreciate if you could argue this by a reference to a spec which is the only authoritative source. Please consider the matter more thoroghly than just relying on a single book.
The BPMN standard defines an embedded sub-process as an "Inline Block". Its main purpose is to group together in a sub-process several process elements that can be used at a specific point in a process. This is why embedded sub-processes do not have pools.
If you need to use pools and lanes, then we recommend you use a re-usable, independent process.
As I wrote in the previous message, the embedded process can't have the pool of its own, for sure. However it may have lanes to depict process performers and it also may have other pools e.g. to depict external entities/participants.
Being a staged BPMN trainer I witness that lack of lanes in embedded subprocesses is #1 issue that almost every Bizagi Modeler users hits upon.
Switching to reusable subprocesses is a good workaround indeed but not a solution. Hopefully I don't need to explain that there is significant difference between embedded and reusable subprocesses.
From my perspective, you are trying to defend the undefendable position in this thread. It doesn't improve product and company reputation, really.
FWIW I fully agree with Anatoly.
Bizagi obviously confuses
1) "an embedded subprocess cannot define its own pools"
2) "an embedded subprocess cannot re-use pools/lanes from its parent (or include an external pool like "black box" - which is also not supported by Bizagi as I understand?)
While 1) is true, 2) is certainly not and also not defined this way in the standard.
I also agree that re-usable subprocesses are no good workaround.
I don't wish to give an opinion either way on the standards of an embeded vs re-usable however i do wish to identify a my needs to have some consistency between them in Bizagi Modeler. My difficulty lies with how the web documentation displays them differently. My preference would be to allow both the embeded and re-usable sub-processes to both have the option to be collapsable in the left navigation menu, and the option to have the sub processes appear in a modal window.
Hi, we are moving our BPMN diagrams from another tool into Bizagi. I have diagrammed the following scenario in several other tools but have found it more difficult in Bizagi.
We have a parent diagram with sub-processes 4 layers deep. Each layer has tasks/sub-processes that may be carried out by either a single actor or several actors. In other tools I have included lanes, or a pool with lanes in each diagram layer, and labeled each lane with either the one resource that carries out all the tasks/sub-processes or left the lane title blank to indicate that several resources carry out the tasks/sub-processes in the lane (which will be generally narrowed down to a single actor in one of the sub-processes). This layering can be repeated until you get down to the detail needed for a single actor.
To do this in Bizagi today and to get access to the lanes that I need, nearly every one of my sub-processes has to be a reusable sub-process. These are not really reusable sub-processes; so, I would lend my voice to those requesting that lanes be included with the palette for normal sub-processes.
Not a Bizagi user, but:
Pools obviously not in subprosses
Lanes: BPMN is not that spesific about lanes. Lanes can be used for any partitioning of the activities of a process - even phases.
I'm using lanes consistently for specifying actors. In that case using a lane in a subprocess would logically correspond to a sublane of the lane above.
Logically ok - but a bit messy..
But I do not see BPMN going against lanes in subprocesses.
I'm surprised at a tool preventing such.
Ups, in my above comment, I meant partitioning (not participation), thus
Lanes: BPMN is not that spesific anout lanes. Lanes can be used for any partitioning of the activities of a process - even phases.
OK - found the edit link, and corrected the above post.(This is really not easy on a phone :-) )
Comments have been locked on this page!