This is a great compendium of practical experience taking organizations
agile with Lean methods. It includes methods where the agile direction
is oriented more from the business than from teams doing a bottom up
approach alone.
You can also see selected chapters on our site
http://www.netobjectives.com/lasd
Alan Shalloway
CEO, Net Objectives
Achieving Enterprise and Team Agility
Congrats Mary and Tom. Big accomplishment.
Looking forward to picking up a copying and reading some interesting stories!
All the best - Jonathan
--- In leandevelopment@yahoogroups.com, "Mary Poppendieck" <mary@...> wrote:
>
> Our new book, Leading Lean Software Development, is now available here
> <http://www.amazon.com/exec/obidos/ASIN/0321620704/poppendieckco-20> .
>
>
>
> Enjoy!
>
>
>
> Mary Poppendieck
>
> 952-934-7998
>
> www.poppendieck.com
>
> Author of: Lean Software Development, Implementing Lean Software
> Development, & Leading Lean Software Development
>
We're an XP team (of 10 people) trying to go a bit more lean, and I don't know what to do with multiple work items in progress. With XP, the end of the iteration was the time to make sure everything was done-done, to cut a release, and deploy (I do oversimplify, I have worked on teams where we deployed daily, but not my current team). Yet with a kanban system, I get the impression I should be ready to deploy as each card is finished... but there are other story cards underway, that may be half finished--so I don't think I'm in a spot I can really release after every card... and since we're talking about eliminating iterations altogether, I don't know when we'll get to a fully-tested done-done code base. What is your lean team doing to be confident the code is ready to release? I don't like using branching, or flags in the code, because it adds overhead. I can't see how limiting the work-in-progress slots to 1 would be effective on a team of 10. Still, maybe that's what you're doing--tell me what advantages/disadvantages you see...
1. Branching strategy (yes I know you don't like this) - create a feature-branch that, on commit, gets merged back into a simulation of trunk (another branch) that gets commits from other feature team commits (i.e. separate work items) in progress. This simulation branch is where your traditional CI goes. First feature team "done" with their feature promotes their changes to trunk and feature teams in progress pull from trunk. Yes, this is a lot of apparatus. Tools like git make this *slightly* less painful.
2. Possibly other branching strategies... folks?
3. Composite / micro-kernel architectures - solve isolation issues at a component level. This requires a little upfront thinking on the architecture front and small, isolated components. We see these considerations in language/runtimes such as Erlang, OSGi, etc. This obviously flies in the face of BDUF, but I'm a big fan of thinking about architecture and design constraints upfront. Just because we're making broad decisions about constraints doesn't mean we're over-engineering; certain JIT architecture points will be larger than others.
#3 - is sure to incite some controversy, I'd think.
On Thu, Nov 5, 2009 at 9:47 AM, D. André Dhondt <d.andre.dhondt@...> wrote:
We're an XP team (of 10 people) trying to go a bit more lean, and I don't know what to do with multiple work items in progress. With XP, the end of the iteration was the time to make sure everything was done-done, to cut a release, and deploy (I do oversimplify, I have worked on teams where we deployed daily, but not my current team). Yet with a kanban system, I get the impression I should be ready to deploy as each card is finished... but there are other story cards underway, that may be half finished--so I don't think I'm in a spot I can really release after every card... and since we're talking about eliminating iterations altogether, I don't know when we'll get to a fully-tested done-done code base. What is your lean team doing to be confident the code is ready to release? I don't like using branching, or flags in the code, because it adds overhead. I can't see how limiting the work-in-progress slots to 1 would be effective on a team of 10. Still, maybe that's what you're doing--tell me what advantages/disadvantages you see...
We use Subversion, for which branching is cheap. If you branch (or tag - it is
effectively the same thing) at each complete build, say after each check-in to a
central repository, you then have an identifiable release candidate that you can
DECIDE to make ready for deployment and release to your users (whatever that
takes). That branch is where you'd apply any fixes to that release, of course.
In other words, each branch would have an agreed "policy" so that everyone knew
whether it was simply created by a regular build, or DECLARED as a baseline for
a release - something that people were actually going to use.
I found this advice from Perforce very helpful:
http://www.perforce.com/perforce/papers/bestpractices.html
I hope you feel that that is on topic and somewhat helpful.
--- In leandevelopment@yahoogroups.com, D. André Dhondt <d.andre.dhondt@...>
wrote:
>
> We're an XP team (of 10 people) trying to go a bit more lean, and I don't
> know what to do with multiple work items in progress. With XP, the end of
> the iteration was the time to make sure everything was done-done, to cut a
> release, and deploy (I do oversimplify, I have worked on teams where we
> deployed daily, but not my current team). Yet with a kanban system, I get
> the impression I should be ready to deploy as each card is finished... but
> there are other story cards underway, that may be half finished--so I don't
> think I'm in a spot I can really release after every card... and since we're
> talking about eliminating iterations altogether, I don't know when we'll get
> to a fully-tested done-done code base. What is your lean team doing to be
> confident the code is ready to release? I don't like using branching, or
> flags in the code, because it adds overhead. I can't see how limiting the
> work-in-progress slots to 1 would be effective on a team of 10. Still,
> maybe that's what you're doing--tell me what advantages/disadvantages you
> see...
>
> --
> D. André Dhondt
> http://dhondtsayitsagile.blogspot.com/
>
> Support low-cost conferences -- http://agiletour.org/
> If you're in the area, join Agile Philly http://www.AgilePhilly.com
>
We're an XP team (of 10 people) trying to go a bit more lean, and I don't know what to do with multiple work items in progress. With XP, the end of the iteration was the time to make sure everything was done-done, to cut a release, and deploy (I do oversimplify, I have worked on teams where we deployed daily, but not my current team). Yet with a kanban system, I get the impression I should be ready to deploy as each card is finished... but there are other story cards underway, that may be half finished--so I don't think I'm in a spot I can really release after every card... and since we're talking about eliminating iterations altogether, I don't know when we'll get to a fully-tested done-done code base. What is your lean team doing to be confident the code is ready to release? I don't like using branching, or flags in the code, because it adds overhead. I can't see how limiting the work-in-progress slots to 1 would be effective on a team of 10. Still, maybe that's what you're doing--tell me what advantages/disadvantages you see...
Some thoughts...
* Feature Branching - depends on your CM tool, and how well it deals with merging
* Latent Code Patterns - allow WIP to be released by developing it in such as way that its not visible until its Done.
* Have a Release Cadence where you batch up features into a release
I think you should also consider what others said, but... After one card is finished, are you talking about still having half-finished features (ie: if you click here and there the system does nothing) or broken features (the system crashs with a new feature and/or breaks something that was implemented before)?
I consider "ready to release" as "the end-user is fully able to use the system", which often translates to "all previous tests (unit and acceptance) are still passing".
Would your software be, in any aspect (functional or non-functional), less useful today than it was yesterday? If so, you´re not ready to release yet.
Hope this helps. Regards,
On Thu, Nov 5, 2009 at 9:47 AM, D. André Dhondt <d.andre.dhondt@...> wrote:
We're an XP team (of 10 people) trying to go a bit more lean, and I don't know what to do with multiple work items in progress. With XP, the end of the iteration was the time to make sure everything was done-done, to cut a release, and deploy (I do oversimplify, I have worked on teams where we deployed daily, but not my current team). Yet with a kanban system, I get the impression I should be ready to deploy as each card is finished... but there are other story cards underway, that may be half finished--so I don't think I'm in a spot I can really release after every card... and since we're talking about eliminating iterations altogether, I don't know when we'll get to a fully-tested done-done code base. What is your lean team doing to be confident the code is ready to release? I don't like using branching, or flags in the code, because it adds overhead. I can't see how limiting the work-in-progress slots to 1 would be effective on a team of 10. Still, maybe that's what you're doing--tell me what advantages/disadvantages you see...
>I consider "ready to release" as "the end-user is fully able to use the system" Our continuous integration system catches regressions, if any, so in general we're always adding value. The problem I have, though, is that we like to do commits every few minutes, so story card 1 will infect story card 2, and when we pull trunk, we've got functionality that, even though it may be hidden, deactivated, or inaccessible, it's still there, potentially impacting the quality of story card 2, but we may not even think to do exploratory testing in that neighborhood since we're validating #2. With our end-of-iteration exploratory testing practice, we'd be hitting all the areas touched that iteration--but if we don't have an iteration, we don't know what was touched...
On Thu, Nov 5, 2009 at 5:04 PM, Karl Scotland <kjscotland@...> wrote:
* Feature Branching - depends on your CM tool, and how well it deals with merging
We refactor mercilessly--often making renaming changes that hit a dozen classes here or there. Do you have CM tools that can handle these kinds of changes? (we're using IDEA and svn and hudson)
* Latent Code Patterns - allow WIP to be released by developing it in such as way that its not visible until its Done.
This is our most common strategy--but I don't like the overhead of going in later to remove flags, switches, options, whatever--but it seems the least costly, anyway, to me.
* Have a Release Cadence where you batch up features into a release
Interesting idea--but doesn't that violate the principle of stockpiling/inhibit feedback on quality?
By the way, we too use svn, and I don't like applying patches to both trunk and a branch. I don't want the overhead!
With our end-of-iteration exploratory testing practice, we'd be hitting all the areas touched that iteration--but if we don't have an iteration, we don't know what was touched...
I see... I don´t know your specific enviroment and problems, but can you move the testing forward in the value chain? I mean, in your example, define the acceptance tests of the story card 1 and 2 (which may potentially impact each other) before finishing any one of then and creating a build system/conditionals that automatically enables/shows each one when their tests pass sucessfully?
I'm studying the possibility of going to the Lean Journey next year (http://www.enna.com/japantrip/index_affiliate71.asp). I'd like to hear a little bit from you guys if it's worth the investment (around USD 8,000.00 + air fares), since I'm not a manufacturing person, I'm from the world of Lean Software Development.
From my point of view, it is worth. All the key concepts are there, and having the opportunity to see a huge company with a in-depth lean culture is priceless. But this is my point of view, and I'd love if you guys could give me some insight.
Has someone from this list ever took part of this event? If yes, how does it contributed to your day-to-day practices? What did you took from there?
Bent Jensen of Denmark leads a study trip to Japan specifically
aimed at people in software development.He is arranging the
third annual trip the week of April 19, 2010. A visit to Toyota, as well
as other "lean", high-tech companies will be on the agenda. If you would
like more information or to be on the mailing-list for the trip, send an e-mail
to bent.jensen@.... I was
on the trip last year and can highly recommend it.
Mary Poppendieck
952-934-7998
www.poppendieck.com
Author of Lean Software Development, Implementing Lean Software
Development, Leading Lean Software Development
From:
leandevelopment@yahoogroups.com [mailto:leandevelopment@yahoogroups.com] On
Behalf Of Igor Maciel Macaubas Sent: Thursday, November 05, 2009 10:45 AM To: leandevelopment@yahoogroups.com Subject: [leandevelopment] Lean Journey
Hi all,
I'm studying the possibility of going to the Lean Journey next year (http://www.enna.com/japantrip/index_affiliate71.asp).
I'd like to hear a little bit from you guys if it's worth the investment
(around USD 8,000.00 + air fares), since I'm not a manufacturing person, I'm
from the world of Lean Software Development.
From my point of view, it is worth. All the key concepts are there, and having
the opportunity to see a huge company with a in-depth lean culture is
priceless. But this is my point of view, and I'd love if you guys could give me
some insight.
Has someone from this list ever took part of this event?
If yes, how does it contributed to your day-to-day practices? What did you took
from there?
D. André Dhondt wrote:
> * Latent Code Patterns - allow WIP to be released by developing it
> in such as way that its not visible until its Done.
>
> This is our most common strategy--but I don't like the overhead of going
> in later to remove flags, switches, options, whatever--but it seems the
> least costly, anyway, to me.
Andre,
Can you think of ways to reduce that overhead? When I've used this
strategy, I've generally found ways to minimize the places that require
a run-time decision, such as the selection of which object factory to
use or which injections to make.
- Geoge
--
----------------------------------------------------------------------
* George Dinwiddie * http://blog.gdinwiddie.com
Software Development http://www.idiacomputing.com
Consultant and Coach http://www.agilemaryland.org
----------------------------------------------------------------------
I would agree to not creating branches as much as possible. In my
experience usually leads to non value added activities.
I think one factor to consider would be dependencies (functional or
otherwise) between stories that are being worked on. If 10 people are
working on 10 stories and all are independent then each story could be
released one at a time, without too much overhead. However if there are
functional dependencies then there is an added cost of the exploratory
testing being done on a code base comprising of incomplete (not fully
done done stories) stories. There may be additional costs due to
repetition of the exploratory testing per story released. Typically a
time boxed iteration where a certain set of stories are ascertained for
regression together seems to add value. This is particularly true,
where part of the regression testing is manual.
Another factor to consider would be to look at people assigned/story as
a factor. IMHO, more people per story seems to induce more need for
planning and hence potential waste due to wait time, unplanned
dependencies etc.,
The release per story approach might be more valid in cases, where the
cost of additional/missed exploratory testing is compensated by the
business risk of the story ( i.e. I expect to find lots of issues after
usage) and/or value due to early delivery. If the business risk is the
primary reason, then alternative approaches could be used to reduce
that risk - demo environment, end user hands down usage on test
environments etc., could be options.
Hope it helps.
Arun
D. André Dhondt wrote:
We're an XP team (of 10 people) trying to go a bit more lean, and
I don't know what to do with multiple work items in progress. With XP,
the end of the iteration was the time to make sure everything was
done-done, to cut a release, and deploy (I do oversimplify, I have
worked on teams where we deployed daily, but not my current team). Yet
with a kanban system, I get the impression I should be ready to deploy
as each card is finished... but there are other story cards underway,
that may be half finished--so I don't think I'm in a spot I can really
release after every card... and since we're talking about eliminating
iterations altogether, I don't know when we'll get to a fully-tested
done-done code base. What is your lean team doing to be confident the
code is ready to release? I don't like using branching, or flags in
the code, because it adds overhead. I can't see how limiting the
work-in-progress slots to 1 would be effective on a team of 10. Still,
maybe that's what you're doing--tell me what advantages/disadvantages
you see...
The inaugural Lean Software & Systems Conference is now open for registration
and paper submissions.
Register or submit now at http://atlanta2010.leanssc.org/
Conference Chair: David Anderson
Track Chairs: Alan Shalloway, Joshua Kerievsky, James Sutton, Richard Turner,
Eric Willeke, Chris Shinkle and David Anderson
Event Planner: Kelly Wilson (SEP)
Event Team: Janice Linden-Reed, Dennis Stevens, Aaron Sanders, Eric Landes
Registration
------------
Registration is limited by venue capacity to 250 in addition to our 40 speakers
and event planning team.
The first 50 registrants receive a special discount price of $800, plus an
exclusive invite to the speaker luncheon on Thursday April 22nd and a special
Ltd WIP Society t-shirt designed by David Anderson and exclusive to the event.
Regular registration will cost $995.
Late comer registration after March 31st 2010 will cost $1250.
There are currently no plans to offer single day registration.
Venue
-----
JW Marriott, Buckhead, Atlanta. A special room block of 180 rooms has been
reserved. Please use the link on the conference site to make your reservation.
Dates
-----
Wednesday April 21 to Friday April 23, 2010
Call For Papers
---------------
This year's conference features 2 key note speakers, 21 invited speakers and 17
presentations selected from the call for papers. Details of all invited speakers
are already on the web site.
The 2010 event features 3 tracks per day and either 1 hour or 45 minute time
slots depending on track and day.
Sponsorship
-----------
Software Engineering Professionals (SEP) is the organizing sponsor. We are
seeking other sponsors for the event and the sponsorship proposal package will
be available in December. If you wish to receive a copy please email
info@...
Register or submit now at http://atlanta2010.leanssc.org/
Hi all,
I was wondering if any of you guys had any insightful readings in regards to
implementing Lean/Agile in an unionized environement. I am based in the province
of Quebec and we are seeing government RFPs who are asking for Agile/Lean
content in IT.
I know Lean has gotten a lot of traction in healthcare, another sector packed
with unions. In my opinion, it is mainly due to the fact that doctors, who are
central characters in this field, are the ones pushing for Lean in healthcare.
But in IT, how would one help an organization who has union workers? I do agree
when authors like Demings or Seddon say that we should change the system. But
with a union to work with, it doesn't really work that way. Things are more
crystallized. Any thoughs on change management toward Agile/Lean in this
particular environment?
Thanks,
Louis-Philippe Carignan
Given that Lean has been applied in the most unionized business of
all, automobile manufacturing, I think it is possible. In Ohno's book
on TPS he mentions the labor strife they had at Toyota in the 50's and
eventually the unions capitulated and allowed their members to do more
than one job.
What specifically is the problem with the union in your case? Job
descriptions? Have you talked with any of the staff or union reps? If
the govt is mandating Agile, why would the unions be an issue?
Regards,
Robin Dymond.
On 11/15/09, Louis-Philippe Carignan <lpcarignan@...> wrote:
> Hi all,
>
> I was wondering if any of you guys had any insightful readings in regards to
> implementing Lean/Agile in an unionized environement. I am based in the
> province of Quebec and we are seeing government RFPs who are asking for
> Agile/Lean content in IT.
>
> I know Lean has gotten a lot of traction in healthcare, another sector
> packed with unions. In my opinion, it is mainly due to the fact that
> doctors, who are central characters in this field, are the ones pushing for
> Lean in healthcare.
>
> But in IT, how would one help an organization who has union workers? I do
> agree when authors like Demings or Seddon say that we should change the
> system. But with a union to work with, it doesn't really work that way.
> Things are more crystallized. Any thoughs on change management toward
> Agile/Lean in this particular environment?
>
> Thanks,
>
> Louis-Philippe Carignan
>
>
>
--
Robin Dymond, CST
Managing Partner, Innovel, LLC.
www.innovel.net
www.scrumtraining.com
Americas: (804) 239-4329
Europe: +32 489 674 366
Hi,
A medical unionized environment. Doctors are sensitive to cutbacks in
their pockets
so they will lead/push for overall trimming.
Doctors are also going strong in 'crew resource management' (CRM)
following air crew ideas.
Going from you having a 1 in 7 chance of getting the wrong medicine while
in a North American hospital now, to anything less will often be linked
to CRM.
A key is reducing the number of job classifications and improving the
adaptability/flexibility.
In Western Canada and other industries contractors become more common,
and more
activities are contracted outside the centralized union facilities
(hospitals) into private settings
(which is what has happened all along for politicians and police and
other special people).
Louis-Philippe Carignan wrote:
> . . .
> I know Lean has gotten a lot of traction in healthcare, another sector packed
with unions. In my opinion, it is mainly due to the fact that doctors, who are
central characters in this field, are the ones pushing for Lean in healthcare.
>
> But in IT, how would one help an organization who has union workers? I do
agree when authors like Demings or Seddon say that we should change the system.
But with a union to work with, it doesn't really work that way. Things are more
crystallized. Any thoughs on change management toward Agile/Lean in this
particular environment?
>
>
>
bye,
vic
--
"Conventional MBA programs train the wrong people in the
wrong ways with the wrong consequences."
-- Henry Mintzberg, McGill University
"It's not my problem. The hole is in their side of the boat."
- John Brewer - Vic Williams 13066824980
Hi Robin,
One of my worries was about a collocated team. As stupid as it may sounds, some
people can have a cubicle next to a window because they have been there for 10
years. They deserve a window spot, even if they aren't close to the rest of the
team. Another issue is how code is build. Believe or not but not install an
application in a QA environment, one person will compile the code and another
person will install it. I'm wondering how will I present them with ideas like
continuous integration where this could all be done automatically.
My best selling point is scarce ressource. As baby boomers are retiring, there's
less people
to replace the current people.
Interesting to hear that unions capitualed at Toyota in the 50s. Problem is, my
project will be for a public sector agency. There jobs aren't on the line.
One of my approach was to use the fun theory
(http://analytical-mind.com/2009/11/09/have-you-heard-of-the-fun-theory/). By
making Agile a better alternative than there current way of working, I should be
able to bring them on board. I was planning on identify the leader(s) and make
them early adopters so the rest of the gang can jump in.
I was just wondering if anybody out here had such an experience. I've found this
group very helpful in the past so I was wondering if anybody had pointers to
give me.
Louis-Philippe
--- In leandevelopment@yahoogroups.com, Robin Dymond <robin.dymond@...> wrote:
>
> Given that Lean has been applied in the most unionized business of
> all, automobile manufacturing, I think it is possible. In Ohno's book
> on TPS he mentions the labor strife they had at Toyota in the 50's and
> eventually the unions capitulated and allowed their members to do more
> than one job.
>
> What specifically is the problem with the union in your case? Job
> descriptions? Have you talked with any of the staff or union reps? If
> the govt is mandating Agile, why would the unions be an issue?
>
> Regards,
> Robin Dymond.
>
>
>
>
> On 11/15/09, Louis-Philippe Carignan <lpcarignan@...> wrote:
> > Hi all,
> >
> > I was wondering if any of you guys had any insightful readings in regards to
> > implementing Lean/Agile in an unionized environement. I am based in the
> > province of Quebec and we are seeing government RFPs who are asking for
> > Agile/Lean content in IT.
> >
> > I know Lean has gotten a lot of traction in healthcare, another sector
> > packed with unions. In my opinion, it is mainly due to the fact that
> > doctors, who are central characters in this field, are the ones pushing for
> > Lean in healthcare.
> >
> > But in IT, how would one help an organization who has union workers? I do
> > agree when authors like Demings or Seddon say that we should change the
> > system. But with a union to work with, it doesn't really work that way.
> > Things are more crystallized. Any thoughs on change management toward
> > Agile/Lean in this particular environment?
> >
> > Thanks,
> >
> > Louis-Philippe Carignan
> >
> >
> >
>
>
> --
> Robin Dymond, CST
> Managing Partner, Innovel, LLC.
> www.innovel.net
> www.scrumtraining.com
> Americas: (804) 239-4329
> Europe: +32 489 674 366
>
Louis-Philippe Carignan wrote:
> Hi Robin,
>
> One of my worries was about a collocated team. As stupid as it may sounds,
some people can have a cubicle next to a window because they have been there
for 10 years. They deserve a window spot, even if they aren't close to the rest
of the team. Another issue is how code is build. Believe or not but not install
an application in a QA environment, one person will compile the code and another
person will install it. I'm wondering how will I present them with ideas like
continuous integration where this could all be done automatically.
>
I've "done agile" in the public sector, and have yet to come across a
scenario where people had such narrow job descriptions. I have seen
situations where a specific group had to install applications on servers
in the QA/Integration (with other apps) environment, but that was only
mildly annoying and the group that performed the work did their best to
keep up with the demand for installations.
I guess what I would question is who is pushing the use of an agile
process for this client? Is their management doing it? If so, why,
i.e. what benefits are they trying to achieve? What pain are they
feeling now? If they can state that, then you have the leverage you
need to introduce the practices that will help them. If they don't
think they need help, you will only be hurting yourself.
As for co-location, I've coached in France before. The labour laws
there, I was told, stipulate that contractors/consultants must have
their desks in a physically separate work area than the full-time
employees. The group did plan together and quite often paired in the
same location, but there was no way to avoid that regulation 100% of the
time.
> Interesting to hear that unions capitualed at Toyota in the 50s. Problem is,
my project will be for a public sector agency. There jobs aren't on the line.
>
I hear this a lot about the public sector. Fear of losing a job is no
way to motivate people. When people actually enjoy their work, they
will be motivated. I've watched teams in both sectors transition to
Agile and actually start having fun again.
--
Dave Rooney
Co-founder and Consultant, The Agile Consortium
"Maximizing the value of your IT investments!"
E-mail: dave@...
Twitter: daverooneyca
http://www.theagileconsortium.comhttp://practicalagility.blogspot.com
Let me introduce myself as a Software consultant , mainly in Microsoft.NET platform, in Bangalore, India. I was recently asked to guide the software team of 10 people in a startup company in their processes. Since they are mostly junior people and do not have the time to spend time discussing or implementing heavy processes, i decided to put together a checklist for them. This checklist has become much bigger than i thought (almost 300 items) and seems a bit paranoid :) I do not like ISO like documentation myself, but it has crept in a lot in the list since i thought this company needed a way to retain it's ideas. I hope this list is influenced by a light weight approach like Agile, XP and others.
In any case, i hope some of you can take a look at it and tell me if there is some method to the madness. How can it be improved ?
Ricardo semler talks about unions in his book Maverick which is well worth reading.
Sent from my iPhone
On Nov 15, 2009, at 20:44, Robin Dymond <robin.dymond@...> wrote:
Given that Lean has been applied in the most unionized business of
all, automobile manufacturing, I think it is possible. In Ohno's book
on TPS he mentions the labor strife they had at Toyota in the 50's and
eventually the unions capitulated and allowed their members to do more
than one job.
What specifically is the problem with the union in your case? Job
descriptions? Have you talked with any of the staff or union reps? If
the govt is mandating Agile, why would the unions be an issue?
Regards,
Robin Dymond.
On 11/15/09, Louis-Philippe Carignan <lpcarignan@yahoo.com> wrote:
> Hi all,
>
> I was wondering if any of you guys had any insightful readings in regards to
> implementing Lean/Agile in an unionized environement. I am based in the
> province of Quebec and we are seeing government RFPs who are asking for
> Agile/Lean content in IT.
>
> I know Lean has gotten a lot of traction in healthcare, another sector
> packed with unions. In my opinion, it is mainly due to the fact that
> doctors, who are central characters in this field, are the ones pushing for
> Lean in healthcare.
>
> But in IT, how would one help an organization who has union workers? I do
> agree when authors like Demings or Seddon say that we should change the
> system. But with a union to work with, it doesn't really work that way.
> Things are more crystallized. Any thoughs on change management toward
> Agile/Lean in this particular environment?
>
> Thanks,
>
> Louis-Philippe Carignan
>
>
>
It's actually a check list of 30 items (not 300), but it's still pretty intimidating. I'd be afraid it could paralyze development, or scare people into adding much more complexity than they need right now. Is this a new code base? I think that if you set up the basics of version control, good TDD practices, weekly releases+retrospectives, you can let the team make their own processes, especially if you get them all reading something together like the Art of Agile Development or XP explained or Agile Principles, Patterns, and Practices...
Let me introduce myself as a Software consultant , mainly in Microsoft.NET platform, in Bangalore, India. I was recently asked to guide the software team of 10 people in a startup company in their processes. Since they are mostly junior people and do not have the time to spend time discussing or implementing heavy processes, i decided to put together a checklist for them. This checklist has become much bigger than i thought (almost 300 items) and seems a bit paranoid :) I do not like ISO like documentation myself, but it has crept in a lot in the list since i thought this company needed a way to retain it's ideas. I hope this list is influenced by a light weight approach like Agile, XP and others.
In any case, i hope some of you can take a look at it and tell me if there is some method to the madness. How can it be improved ?
Let me introduce myself as a Software consultant , mainly in Microsoft.NET platform, in Bangalore, India. I was recently asked to guide the software team of 10 people in a startup company in their processes. Since they are mostly junior people and do not have the time to spend time discussing or implementing heavy processes, i decided to put together a checklist for them. This checklist has become much bigger than i thought (almost 300 items) and seems a bit paranoid :) I do not like ISO like documentation myself, but it has crept in a lot in the list since i thought this company needed a way to retain it's ideas. I hope this list is influenced by a light weight approach like Agile, XP and others.
In any case, i hope some of you can take a look at it and tell me if there is some method to the madness. How can it be improved ?
If the team, which has already completed one SDLC life cycle of their product
for a local customer, shows more interest in 'process', i wanted them to fill in
their answers to each of these questions. This would have been the baseline of
the current process.
--- In leandevelopment@yahoogroups.com, Ilja Preuß <iljapreuss@...> wrote:
>
> How do you intend the list to be used?
>
> Curious, Ilja
>
> 2009/11/16 Chakravarthy R <chakravarthy.rajagopalan@...>
>
> >
> >
> > Hi,
> >
> > Let me introduce myself as a Software consultant , mainly in Microsoft.NET
> > platform, in Bangalore, India. I was recently asked to guide the software
> > team of 10 people in a startup company in their processes. Since they are
> > mostly junior people and do not have the time to spend time discussing or
> > implementing heavy processes, i decided to put together a checklist for
> > them. This checklist has become much bigger than i thought (almost 300
> > items) and seems a bit paranoid :) I do not like ISO like documentation
> > myself, but it has crept in a lot in the list since i thought this company
> > needed a way to retain it's ideas. I hope this list is influenced by a
> > light weight approach like Agile, XP and others.
> >
> > In any case, i hope some of you can take a look at it and tell me if there
> > is some method to the madness. How can it be improved ?
> >
> > Thanks,
> > Chak.
> >
> > http://moko59.wordpress.com/2009/11/15/self-check-list-for-a-software-team/
> >
> >
> >
>
Your example seems like more of a template for overall assessment than a checklist in the Toyota Product Development System sense.
I think of a checklist as something we apply to an task we regularly perform in the course of delivering software. A checklist, when completed, let's us know when we can place an order in the next pull buffer. A couple examples:
- Acceptance Tests on a story. - A deployment procedure. - A series of common steps we use to invest in defining an MMF (story map, conceptual map, ubiquitous language, story breakdown)
Let me introduce myself as a Software consultant , mainly in Microsoft.NET platform, in Bangalore, India. I was recently asked to guide the software team of 10 people in a startup company in their processes. Since they are mostly junior people and do not have the time to spend time discussing or implementing heavy processes, i decided to put together a checklist for them. This checklist has become much bigger than i thought (almost 300 items) and seems a bit paranoid :) I do not like ISO like documentation myself, but it has crept in a lot in the list since i thought this company needed a way to retain it's ideas. I hope this list is influenced by a light weight approach like Agile, XP and others.
In any case, i hope some of you can take a look at it and tell me if there is some method to the madness. How can it be improved ?
Thanks Dave. Yes, i could call it an assessment list or template. It is just to
know the current state so that the team can then plan their future processes
based on their level of interest and needs.
Your checklist example is good.
Regards,
Chak.
--- In leandevelopment@yahoogroups.com, David Laribee <david@...> wrote:
>
> Hi Chak -
>
> Your example seems like more of a template for overall assessment than a
> checklist in the Toyota Product Development System sense.
>
> I think of a checklist as something we apply to an task we regularly perform
> in the course of delivering software. A checklist, when completed, let's us
> know when we can place an order in the next pull buffer. A couple examples:
>
> - Acceptance Tests on a story.
> - A deployment procedure.
> - A series of common steps we use to invest in defining an MMF (story map,
> conceptual map, ubiquitous language, story breakdown)
>
> / Dave
>
> David Laribee
> http://thebeelog.com
> 1 (646) 462-4096
>
>
> On Mon, Nov 16, 2009 at 2:00 AM, Chakravarthy R <
> chakravarthy.rajagopalan@...> wrote:
>
> >
> >
> > Hi,
> >
> > Let me introduce myself as a Software consultant , mainly in Microsoft.NET
> > platform, in Bangalore, India. I was recently asked to guide the software
> > team of 10 people in a startup company in their processes. Since they are
> > mostly junior people and do not have the time to spend time discussing or
> > implementing heavy processes, i decided to put together a checklist for
> > them. This checklist has become much bigger than i thought (almost 300
> > items) and seems a bit paranoid :) I do not like ISO like documentation
> > myself, but it has crept in a lot in the list since i thought this company
> > needed a way to retain it's ideas. I hope this list is influenced by a
> > light weight approach like Agile, XP and others.
> >
> > In any case, i hope some of you can take a look at it and tell me if there
> > is some method to the madness. How can it be improved ?
> >
> > Thanks,
> > Chak.
> >
> > http://moko59.wordpress.com/2009/11/15/self-check-list-for-a-software-team/
> >
> >
> >
>
In sunny Chicago, we will hold the next Leading Lean Software Development
course.
This is a new 2-day course / workshop led by Mary & Tom Poppendieck based on
their new book (Nov 2009).
See http://bit.ly/bjuFD for more details. Or contact me with any questions.
Regards,
Julia Hill
Hi Chak
That is a good list. My suggestion is to organize them to logical groups, some
on design, some on risk management, some on architecture and so on. Then the
list will not look intimidating.
Also, what is the purpose of the checklist,do we want to prove something is fine
or something is not fine ? or is the checklist a generic memory jogger ?
Can the checklist be brought to a state that it could be either a Yes or a No to
it ? Some of the questions can run into lenghty descriptive response.
Some thoughts that came to my mind.....
PS:Thanks for hilighting the presence of this group in lean.org forum :)
Rgds
Siva
--- In leandevelopment@yahoogroups.com, "r_chakra"
<chakravarthy.rajagopalan@...> wrote:
>
> Thanks Dave. Yes, i could call it an assessment list or template. It is just
to know the current state so that the team can then plan their future processes
based on their level of interest and needs.
>
> Your checklist example is good.
>
> Regards,
> Chak.
>
> --- In leandevelopment@yahoogroups.com, David Laribee <david@> wrote:
> >
> > Hi Chak -
> >
> > Your example seems like more of a template for overall assessment than a
> > checklist in the Toyota Product Development System sense.
> >
> > I think of a checklist as something we apply to an task we regularly perform
> > in the course of delivering software. A checklist, when completed, let's us
> > know when we can place an order in the next pull buffer. A couple examples:
> >
> > - Acceptance Tests on a story.
> > - A deployment procedure.
> > - A series of common steps we use to invest in defining an MMF (story map,
> > conceptual map, ubiquitous language, story breakdown)
> >
> > / Dave
> >
> > David Laribee
> > http://thebeelog.com
> > 1 (646) 462-4096
> >
> >
> > On Mon, Nov 16, 2009 at 2:00 AM, Chakravarthy R <
> > chakravarthy.rajagopalan@> wrote:
> >
> > >
> > >
> > > Hi,
> > >
> > > Let me introduce myself as a Software consultant , mainly in Microsoft.NET
> > > platform, in Bangalore, India. I was recently asked to guide the software
> > > team of 10 people in a startup company in their processes. Since they are
> > > mostly junior people and do not have the time to spend time discussing or
> > > implementing heavy processes, i decided to put together a checklist for
> > > them. This checklist has become much bigger than i thought (almost 300
> > > items) and seems a bit paranoid :) I do not like ISO like documentation
> > > myself, but it has crept in a lot in the list since i thought this company
> > > needed a way to retain it's ideas. I hope this list is influenced by a
> > > light weight approach like Agile, XP and others.
> > >
> > > In any case, i hope some of you can take a look at it and tell me if there
> > > is some method to the madness. How can it be improved ?
> > >
> > > Thanks,
> > > Chak.
> > >
> > >
http://moko59.wordpress.com/2009/11/15/self-check-list-for-a-software-team/
> > >
> > >
> > >
> >
>
Hi Siva,
I made a mistake in calling this a Checklist, and you and Andre rightly pointed
out that it is a bit intimidating.
However, as Dave said, if i use this as an 'assessment template' , we get to
know why they made some decisions and whether they still apply in current
conditions, why some issues are not adressed yet, and so on. It then becomes
easier to take things forward. Besides, if they document most of their standard
work steps, then it can help a new team to take over.
The questions are open ended since it may otherwise focus towards predefined
solutions.
That was the idea.
Thanks to everyone for their feedback.
Regards,
Chak.
--- In leandevelopment@yahoogroups.com, "sub_mails" <sub_mails@...> wrote:
>
> Hi Chak
>
> That is a good list. My suggestion is to organize them to logical groups, some
on design, some on risk management, some on architecture and so on. Then the
list will not look intimidating.
>
> Also, what is the purpose of the checklist,do we want to prove something is
fine or something is not fine ? or is the checklist a generic memory jogger ?
>
> Can the checklist be brought to a state that it could be either a Yes or a No
to it ? Some of the questions can run into lenghty descriptive response.
>
> Some thoughts that came to my mind.....
>
> PS:Thanks for hilighting the presence of this group in lean.org forum :)
>
> Rgds
> Siva
>
>
>
> --- In leandevelopment@yahoogroups.com, "r_chakra" <chakravarthy.rajagopalan@>
wrote:
> >
> > Thanks Dave. Yes, i could call it an assessment list or template. It is
just to know the current state so that the team can then plan their future
processes based on their level of interest and needs.
> >
> > Your checklist example is good.
> >
> > Regards,
> > Chak.
> >
> > --- In leandevelopment@yahoogroups.com, David Laribee <david@> wrote:
> > >
> > > Hi Chak -
> > >
> > > Your example seems like more of a template for overall assessment than a
> > > checklist in the Toyota Product Development System sense.
> > >
> > > I think of a checklist as something we apply to an task we regularly
perform
> > > in the course of delivering software. A checklist, when completed, let's
us
> > > know when we can place an order in the next pull buffer. A couple
examples:
> > >
> > > - Acceptance Tests on a story.
> > > - A deployment procedure.
> > > - A series of common steps we use to invest in defining an MMF (story map,
> > > conceptual map, ubiquitous language, story breakdown)
> > >
> > > / Dave
> > >
> > > David Laribee
> > > http://thebeelog.com
> > > 1 (646) 462-4096
> > >
> > >
> > > On Mon, Nov 16, 2009 at 2:00 AM, Chakravarthy R <
> > > chakravarthy.rajagopalan@> wrote:
> > >
> > > >
> > > >
> > > > Hi,
> > > >
> > > > Let me introduce myself as a Software consultant , mainly in
Microsoft.NET
> > > > platform, in Bangalore, India. I was recently asked to guide the
software
> > > > team of 10 people in a startup company in their processes. Since they
are
> > > > mostly junior people and do not have the time to spend time discussing
or
> > > > implementing heavy processes, i decided to put together a checklist for
> > > > them. This checklist has become much bigger than i thought (almost 300
> > > > items) and seems a bit paranoid :) I do not like ISO like
documentation
> > > > myself, but it has crept in a lot in the list since i thought this
company
> > > > needed a way to retain it's ideas. I hope this list is influenced by a
> > > > light weight approach like Agile, XP and others.
> > > >
> > > > In any case, i hope some of you can take a look at it and tell me if
there
> > > > is some method to the madness. How can it be improved ?
> > > >
> > > > Thanks,
> > > > Chak.
> > > >
> > > >
http://moko59.wordpress.com/2009/11/15/self-check-list-for-a-software-team/
> > > >
> > > >
> > > >
> > >
> >
>