The Scrum Alliance is working on
A program that will help Scrum teams be more successful by
improving developer practices,
which will no doubt become a product with some value returned to
the Alliance.
Some subset of us was invited directly by SA to help formulate this,
and that subset invited the rest, and now everyone.
Our goal, in some mixture is to
Come up with a clear statement of what a cert is and is not;
Come up with a statement that improvement is a lifetime thing;
Delineate as well as possible the key things a person should know;
and, for at least some members of the group, probably come up with
courses and consulting programs relating to those key things,
since that is, after all, what some of us already do.
Most of what we have said together and agreed among ourselves about
is up here. People can help us understand what should be done by
contributing.
Ron Jeffries
www.XProgramming.com
www.xprogramming.com/blog
A lot of preconceptions can be dismissed when you actually
try something out. -- Bruce Eckel
If the Scrum Alliance is going to sell it anyway (and given the
success of CSM / CST, it seems likely), I'd like to help make it
palatable. As long as it's nutritious, I don't care much what it
tastes like.
Cheers,
Liz.
--
Elizabeth Keogh
l...@lunivore.com
http://lizkeogh.com
http://jbehave.org
=============
You can easily certify that someone knows the Java String API. The
knowledge is objective. The tests are unambiguous, quick, and easy to
administer. The proof is something everyone can agree with.
You can not easily certify skill. A quick and easy test demonstrates
nothing about skill. Skill is best measured through long-term
observation of results.
The certification of knowledge is a measurement of the number of facts
that the subject has committed to memory.
The certification of skill is a measurement of how the subject uses
those facts, other facts, and many other talents and abilities, to
solve _difficult_ problems.
As a developer I am not suspicious of a test of knowledge. Go ahead
and test it. I know what I know, and I know I'll do well. Indeed I
_want_ you to test my knowledge. I know I'll knock your socks off.
As a developer I am deeply suspicious of a test of my skill. How do I
know your test will actually test what I'm skilled at? Why should I
believe that the skills I have are the skills that you value? Why
should I believe that you are intelligent and perceptive enough to
understand my skills? How do I know you aren't a religious zealot who
confuses skill with doctrine? How do I know you won't fail me just
because we disagree?
So as a developer I would not submit to a test that claims to certify
my skill. If you want to know if I'm skilled, you can look at the
projects I've worked on in the past. You can talk to the teams I've
worked in. You can look at the trail of failures and successes that
I've left behind. (Sometimes failures say more about skill than
successes!) And then you can give me a chance to prove my skills to
you in a project.
But I'm not taking a test that you think measures my skill.
=============
Now I know that we aren't truly talking about a test that certifies
skill. But I also think that, at a much deeper level, we are (see
the name of this email list). Or at least it will be perceived as such.
> That seems like a low bar to me. I was once a post-graduate student so
> I know from personal experience that it is perfectly possible to live
> on brown rice and baked beans and vitamin pills--but I wouldn't wan to
> do that ever again.
Of course not! You forgot the ramen noodles!
Ron Jeffries
www.XProgramming.com
www.xprogramming.com/blog
Any errors you find in this are the work of Secret Villains,
whose mad schemes will soon be revealed. -- Wil McCarthy
>> The Scrum Alliance is working on
>> A program that will help Scrum teams be more successful by
>> improving developer practices,
> I understand the the Scrum Alliance has no interest (in any sense of
> the word) in helping non-Scrum teams be more successful, but do the
> members of this group?
I think our records suggest that we have that interest, yes.
> Which leads me to ask, re this:
>> and, for at least some members of the group, probably come up with
>> courses and consulting programs relating to those key things,
>> since that is, after all, what some of us already do.
> What about those of us who do indeed do those those things but who are
> not plugged in to the Scrum Alliance?
What about those of us? I'm not sure I undersand the question. If we
say something good, here or elsewhere, about what developers need to
know, do, learn ... I expect that would be available to everyone.
However ...
> This is not a flippant question, and has a bearing on how this group
> proceeds. For example, I've spoken to some of the Kanban folks about
> their intended certification and they were quite strongly intending
> that there should be multiple competing certification bodies. The
> Scrum Alliance, by contrast, is a closed shop and operates a monopoly
> on Scrum certification.
> Which way does the intention run here?
Speaking just for myself (and Boskone, of course, which goes without
saying except that I just said it), I have grave doubts about the
notion of skills certification. I am always on the cusp between
pulling out of this effort and repudiating it, and trying to guide
it to something "not too abysmal".
Ron Jeffries
www.XProgramming.com
www.xprogramming.com/blog
Thousands of years ago, the first man discovered how to make fire.
He was probably burned at the stake he had taught his brothers to
light - Howard Roark (The Fountainhead, Ayn Rand)
> I have no problem with some value being returned to the Scrum
> Alliance.
> My concern is what is the risk of my participating and then the Scrum
> Alliance sets up a certification program with their instructors and
> those of us not certified in teaching through the SA will have
> essentially set up a competitor for ourselves. I don't mind helping
> individual consultants or even other companies (e.g., Object Mentor -
> whom I've gotten a lot more value than I've given them) but if the
> Scrum Alliance takes and starts offering a developer certification
> program based on the work here I am not sure that is a positive thing
> overall for the industry.
Yes, I very much share the concern. And certainly it is the
intention of the Scrum Alliance to set up a program, and since they
asked us to help, I suppose they plan to use what we are doing.
I, too, am not sure this is a good thing. I hope we'll be able to
have a productive discussion about that, as well as the meat of how
developers should gain skill, and how we might assess it.
> I am clear I can choose not to participate, but I think the idea of
> developer certification is a good idea if done in certain ways and
> would therefore like to participate.
I'm not so sure even of this and would like to hear more support for
the idea.
> Is the correct attitude that we're just a bunch of concerned
> practitioners/consultants speaking to certification and anybody
> anywhere can take advantage of it?
That's my theory ... and I expect that the Scrum Alliance may well
take some or all of what's here, or some new thing, and run with it.
Ron Jeffries
www.XProgramming.com
www.xprogramming.com/blog
Here is Edward Bear, coming downstairs now, bump, bump, bump, on the back
of his head. It is, as far as he knows, the only way of coming downstairs,
but sometimes he feels that there really is another way, if only he could
stop bumping for a moment and think of it. And then he feels that perhaps
there isn't. -- A. A. Milne
> Now I know that we aren't truly talking about a test that certifies
> skill. But I also think that, at a much deeper level, we are (see
> the name of this email list). Or at least it will be perceived as such.
Very very good points. I have no doubt that it will be perceived as
that, in the same way that the CSM is (and in the same way that the
CSM isn't so perceived, i.e. by people with a clue, it won't be).
I remain uncertain ...
Ron Jeffries
www.XProgramming.com
www.xprogramming.com/blog
If another does not intend offense, it is wrong for me to seek it;
if another does indeed intend offense, it is foolish for me to permit it.
-- Kelly Easterley
> Alan,
>
> Possibly. The difference is that potential SMs were *eager* to be
> certified. Developers, I think, will reject the concept out of hand
> (the same way they tend to reject CSMs).
>
> CSMs come from a PM background. The CSM body of knowledge is small
> and easy to apply. Developer skill is an entirely different ball-
> game.
One hesitates to bring in the martial arts analogies again... but how
do aikido/karate/etc. certify things like green/brown/black belt or
whatever?
That would seem like it would have to be skill based?
Adrian
--
http://quietstars.com - twitter.com/adrianh - delicious.com/adrianh
> The United Kingdom Driver and Vehicle Licensing Authority certify that
> on a certain day I demonstrated the ability to ride a motorcycle to a
> certain degree of proficiency. That's an interesting one, since anyone
> else who's interested will assume that this means that I can still do
> so.
Yeah - that one scares me. Despite the fact that:
a) I've not driven a car in... erm... 16 years now
b) When I did drive - the vast majority was in the US (by probably 2
orders of magnitude.)
it's perfectly legal for me to drive in the UK...
To add another one to the list - my recollection of my younger
brothers engineering apprenticeship was that he spent four years with
a mixture of class room teaching and actually "doing the job" under a
variety of different professionals. Last time I talked about it with
him he still found this much more useful than the degree he did later.
Eating baked beans allowed you to survive, while learning, for long
enough to start returning on that investment of time and penury.
Eventually you were able to collect different recipes with the same
nutritious values, then you were able to invent and blend your own.
I'm only looking for Shu on the Shu-Ha-Ri scale. I would like to see
enough technical practices in place for Agile to survive the initial
adoption and allow learning to really kick in. If someone wants to
make money from that... well, this list is public.
I agree mostly with Matt below on the nature of testing systems. We can learn a lot about how pilots get certified.
But pilots get more from their industry than just testing systems. They get entire learning systems.
The FAA probably has pretty good data somewhere about how well pilots are likely to perform who have (A) flown for X * 1000 hours and (B) have been accompanied by certified pilots and (C) have answered the 100 out of the possible 5000 questions correctly and (D) get re-certified at some interval. And while their failure rates and level of professionalism are, I suggest, much higher than ours, planes do crash, and there are still idiots and nitwits piloting airplanes. In that sense, certification may be a useful marketing term, but it's a lousy guarantee.
What do we really want our community to accomplish? I'll speak for myself.
I want our community to tend to generate more genuinely skilled and deeply committed Software Craftsman than it currently does, expressed as a percentage of all of the folks in the world who call themselves programmers. I want it to be more straightforward and affordable than it is for individual programmers who want that level of skill and comittment to be able to acquire much of the necessary knowledge, and then perform, in some controlled way, many of the necessary skills, in a "performance" that can be assessed, partly objectively, and partly subjectively. I want learning systems and testing systems. I want something that scales better than Corey Haines driving around from shop to shop in his car, as courageous as Corey's choice is.
I have a metaphor to consider for "open-book" software learning systems. Think of a through-hike of the Appalachian Trail (described hilariously by Bill Bryson, BTW, in "A Walk in the Woods.") If you actually accomplish an AT thru hike, then you have, of necessity, learned many things about hiking, backpacking, a bit of wildlife survival, stamina, perseverance, problem solving, etc. You face a lot of nasty challenges thru-hiking the AT, ranging from bears and sudden storms to injuries and stolen food. You cope and you learn.
Speaking again for myself, I want a series of "agile programming ordeals" or thru-hikes, of increasing scale, complexity, and perverse/pathological challenge. Hard stuff with high failure rates.
I want students to have affordable access to these ordeals.
I want a subjective/objective means of assessing that someone more-or-less survived an ordeal. Basically, a pass/fail grade. We can use something like "you'll be getting a hundred questions out of the following 5000," as Matt describes below.
And I want to give each such survivor of each such ordeal nothing more than a piece of paper and a T-Shirt that say something like "The Bear Didn't Get Me."
And I want to send hundreds and dozens of such T-Shirts out into the industry, so that the industry can then start to tell us how challenging, useful, and predictive the survivial of these ordeals tend to be. Indeed, I content only the industry can tell us how much people who have survived any learning system tend to have learned, and tend to perform on the job.
So let's concentrate at least as much on the learning systems as we do on the testing mechanism. And let's not have our test "certify" anything more than this per learning experience: "This person survived this ordeal; we suggest this has some meaning." Then let's have the industry tell us what it means for 1000 programmers to have accomplished that, and refine our learning and testing systems and mechanisms accordingly for the next 1000, and so on.
We do not have Object Certified Programming, after all, much less Object Certified Programmers, nor Object Certified programming languages.
I suggest that what we really want, at the root of it, is better ways for us as a community to help programmers learn software craftsmanship, and better ways to encourage the entire industry to prefer such programmers. Testing is really a pretty small part of this larger problem, as critical and thorny as the testing part is.
--Patrick
> ------- patrick welsh twitter: patrickwelsh blog: patrickwilsonwelsh.com
> My wife is 1st degree black belt in Tae Kwon Do. If I recall
> correctly, she had to demonstrate a series of 'forms' movements and
> break a certain number of boards with her hands while an instructor
> watched.
Is there some kind of global organisation that authorises the
instructor to award the belt? A competing set of organisations? No
organisation at all?
(just curious)
> On Sep 11, 12:47 pm, mheusser <matt.heus...@gmail.com> wrote:
>>> One hesitates to bring in the martial arts analogies again... but
>>> how
>>> do aikido/karate/etc. certify things like green/brown/black belt or
>>> whatever?
>>
>> Performance-based assessments.
>>
>> My wife is 1st degree black belt in Tae Kwon Do. If I recall
>> correctly, she had to demonstrate a series of 'forms' movements and
>> break a certain number of boards with her hands while an instructor
>> watched.
Matt - I'm guessing that this depends on a pre-existing relationship
of some kind? Or can anybody walk off the street and demand to be
graded as black belt?
> I wonder how your wife would do in the UFC? Seriously, real fighters
> laugh at black belts unless they have a reputation of proven
> experience in the ring/street. Same problem here.
Adam - I don't think it is the same problem :-)
I'm not interested in whether somebody with a Tae Kwon Do black belt
can beat everybody up. Nobody I know who is into martial arts (I'm not
myself BTW) has that as a primary motivation.
I'm interested in how the Tae Kwon Do (or whatever) _skill_ is learned
and graded - and how that grading infrastructure works. The people I
know who are into martial arts seem to take the grading system
seriously, work hard at it, and respect the people have have achieved
certain levels. Features lacking in most certification programs in the
computing arena.
I'm just wondering if there is anything to learn there?
For example, I remember overhearing a conversation years ago when a
couple of folk at uni were arguing about the various merits of two
different schools of some martial art (I forget which). I'm not even
sure if the "schools" referred to physical institutions/teacher or a
more philosophical divide. Could there be a ScrumSchool / XPSchool /
RonNChetKickAssSchool / whatever?
Dunno.
The colored belts used in many modern styles were invented to promote
a sort of tiered study. The idea is precisely to increase the length
of time that a student will stick around (And keep paying) so that
they have enough time to actually learn something. In traditional
Asian cultures you showed up to study and they abused the hell out of
you, and if you didn't quit they eventually started teaching you stuff
(At which point you were unlikely to leave.)
>> I wonder how your wife would do in the UFC? Seriously, real fighters
>> laugh at black belts unless they have a reputation of proven
>> experience in the ring/street. Same problem here.
>
> Adam - I don't think it is the same problem :-)
>
> I'm not interested in whether somebody with a Tae Kwon Do black belt
> can beat everybody up. Nobody I know who is into martial arts (I'm not
> myself BTW) has that as a primary motivation.
>
The point I am trying to make is that there is a clear disconnect
between achieving some belt level and achieving competence in a real
world skill. Most black belts cannot apply what they have learned
outside of the controlled environment of the dojo, and many white
belts could mop the floor with them based on natural ability and
competitive drive.
So, if you want a model that guarantees that people stick around for a
long time, pay some money, and can't demonstrate any useful skills
from it, then belts are probably a good thing to mimic. Incidentally
most certification systems do fit this model as well.
> I'm interested in how the Tae Kwon Do (or whatever) _skill_ is learned
> and graded - and how that grading infrastructure works. The people I
> know who are into martial arts seem to take the grading system
> seriously, work hard at it, and respect the people have have achieved
> certain levels. Features lacking in most certification programs in the
> computing arena.
>
Because it's a lot of fun, it's interesting, and it takes a lot of
work. There is a real sense that you are progressing and getting
somewhere and that you are working hard to achieve that. It can also
be a competitive sport.
> I'm just wondering if there is anything to learn there?
>
I doubt it. I like the mixed martial arts model better which is why I
brought it up. Your level/belt/rank/whatever is mostly determined by
your ability to compete successfully against people at that same
level. You only get a black belt in BJJ when you can roll with other
black belts.
I'm not sure that applies to software either, but, at least insofar as
we should be examining applied skills and not just the ability to
follow directions and stick around (In which case we can certify my
dog.)
> For example, I remember overhearing a conversation years ago when a
> couple of folk at uni were arguing about the various merits of two
> different schools of some martial art (I forget which). I'm not even
> sure if the "schools" referred to physical institutions/teacher or a
> more philosophical divide. Could there be a ScrumSchool / XPSchool /
> RonNChetKickAssSchool / whatever?
>
I hope not. That just leads to politics and infighting. The last thing
I want to hear is a "my school/teacher is better than yours." We
already have some of that with the whole Scrum vs. XP vs. Lean vs.
whatever is the best method nonsense.
I doubt it. I like the mixed martial arts model better which is why Ibrought it up. Your level/belt/rank/whatever is mostly determined by
your ability to compete successfully against people at that same
level. You only get a black belt in BJJ when you can roll with other
black belts.
I'm not sure that applies to software either, but, at least insofar as
we should be examining applied skills and not just the ability to
follow directions and stick around (In which case we can certify my
dog.)
MMA isn't about beating people up either. It is a competitive (And
lucrative, at least for the top few percent) sport. Tae Kwon Do is
also a competitive sport (Though no one makes any money at it.) The
difference is that MMA fighters are judged solely by their ability to
compete or to coach others to compete. Whereas, TKD belt levels have
little or nothing to do with competition. TKD belt levels have to do
with being able to reproduce movements of dubious actual usefulness in
a way that displays some coordination and lots of rote memorization.
In any case, I dispute the usefulness of this metaphor entirely.
>>
>> I am more interested in the later parts of that same certification.
>> The parts where the student has to fly a plane and land it - first
>> with a teacher, then by himself, then on instruments.
>>
>
> Hey; that's be great ... and it's a non-trivial problem to evaluate.
> If we had more open source development, or more apprenticeships, I
> could see it working. If we could have apprenticeships be standard,
> like internships/residence for medical doctors, I can see us coming
> close to this ideal.
>
I like the idea of apprenticeship, internship, etc. I think it might
matter more that someone learned from the right teacher than that they
can pass a simple test.
> But I doubt that is in scope for this scrum alliance proposal.
>
Certainly not, unfortunately.
>> Given a problem, can you work with a team to decompose the problem
>> into stories, write acceptance tests, test-drive the solution, and
>> integrate and deploy it. Supposing you could have a mini-iteration and
>> do all that as part of a course, then you have the problem of how to
>> accommodate all of the different languages and tool sets (Like trying
>> to train airline pilots and cessna pilots and helicopter pilots and
>> ship pilots and astronauts...)
>
> I like this in principle, and I suppose it could be done in a week.
> You might even combine it with something like 'givecamp' to create an
> environment with real problems, real customers, software that can be
> published and evaluated in public, etc.
>
Yes. Were you at the conference last month? Are you familiar with the
Agile Philanthropy (Live Aid Stage) stuff that Bob Payne and
Lithespeed are doing? http://www.codegreenlabs.com/
> Do you think you could convince industry to go for it? And if yes,
> what would it take?
>
I think that the industry would go for it if a couple big players got
behind it. I also think that, unfortunately, the Scrum Alliance is a
behemoth and whatever they chose to do will effect us for a long time
for better or for worse.
I think the model is inherently broken, and I don't think it suits our
purposes well enough to deserve the attention required to fix it. BTW,
the thing about dirty belts is just an urban myth. The Japanese have
been washing their clothes for several hundred years just like
everyone else. In fact, it works the other way around. The dyes that
were used to make black belts eventually fade to a mottled gray. This
is less true today since the technology to dye the belts has improved,
but some people still pride themselves on having old faded belts
(Means they have been a black belt for a long time and they are old
;-)
>> I doubt it. I like the mixed martial arts model better which is why I
>> brought it up. Your level/belt/rank/whatever is mostly determined by
>> your ability to compete successfully against people at that same
>> level. You only get a black belt in BJJ when you can roll with other
>> black belts.
>>
>> I'm not sure that applies to software either, but, at least insofar as
>> we should be examining applied skills and not just the ability to
>> follow directions and stick around (In which case we can certify my
>> dog.)
>
> Yeah, this is what I mean. So maybe you're sticking too much to the broken
> version of the model when thinking about the belt system.
Again, I think there is only the "broken version" and I don't think we
are likely to, nor should we attempt to, fix it.
> In any case, I dispute the usefulness of this metaphor entirely.
Yes. I think when we start talking about beating up people's wives,
we have gone rather far astray.
Ron Jeffries
www.XProgramming.com
www.xprogramming.com/blog
Ron gave me a good suggestion once. -- Carlton (banshee858)
LOL. For the record, what I was asking was whether the certification
Matt's wife had would be of any use in predicting her ability to apply
those skills against highly skilled opponents in a competitive arena.
I wish no harm to either Matt or his wife. :-)
> And I love the idea that we could receive some type of "merit badges"
> for such that could generate excitement and encourage others to follow
> this behavior, which would in turn help raise the level of
> craftsmanship in software as more people got involved.
> But the question is - is this even within the scope of this
> discussion? :)
A really good outcome might well be something that would work so
much better than whatever the Scrum Alliance has in mind, and have
so much more impact on Scrum success, that they would prefer it.
Ron Jeffries
www.XProgramming.com
www.xprogramming.com/blog
In programming, do, or undo. There is always try. --Yoda
> Like a few hundred others, I signed up here just now because Ron
> said it was a Good Thing. I saw this thread and figured it was
> one I should read first.
> So I have a question: What *is* the purpose of this list *now*?
> Is it to develop a certification program?
I think it is to understand and say what can be said about what a
good program would be to help Agile teams be successful, and if
possibly, to certify individuals and/or teams who are going to be
good at doing that.
> Is it do decide if such a program is a good idea?
That is certainly on my mind. I do not think it is possible to
/decide/, perhaps not even to come to consensus. I hope to come to
some clarity in my own mind.
> Is it to feed into an SA program?
The SA currently thinks they want a developer cert. They asked us to
work on it. Our work is probably not shaping up exactly as they
expected.
I am also aware of two or three other groups in the world who also
think they have been asked by the Scrum Alliance to do this. I do
not understand this fact, and hope to understand it better after
meeting with Jim next week.
> Is it to possibly lead to an independent program?
I can imagine that. Might even prefer it.
> Could it lead to a non-certification-based skill set?
We're definitely working on that. See the documents in the group.
> The name of the group suggests it's much broader than just
> helping the SA folks come up with a program that will be
> accepted. I'm interested in that broader context, so long
> as it's real and not just a label.
I'm as real as I know how to be. I think it possible that the SA
will take what we have and run with it. I think they are predisposed
to do so.
> I'm personally not interested in helping SA market a program,
> although I will defend your right to do so if that's what
> you're doing.
Our CSM course kicks ass whether it is a CSM course or how to do
Agile course. If we put together a development-focused course, asses
will run in terror from even more kicking.
Ron Jeffries
www.XProgramming.com
www.xprogramming.com/blog
Steering is more important than speed,
in driving and in software development.