Home » Archived » EPF » OpenUP solution looking for a problem
OpenUP solution looking for a problem [message #38827] |
Thu, 30 August 2007 07:10 |
Eclipse User |
|
|
|
Originally posted by: tony.tatil.gmail.com
It may sound naive, but to me OpenUP looks like a solution looking for a
problem.
There are already many excellent agile and formal methodologies out there.
What development teams need is integration of these methodologies into
tools they use daily, not another new methodology created by methodology
high-priests. Jazz (Team Concert) is a great example of such a tool.
It would be even better if IBM just open sources RUP, so we don't spend
any more time coming up with a new open source methodology.
We may just have to accept the bitter reality that after two years (of
marketing and promotion) there is not enough interest in OpenUp for a
genuine community to form around it.
|
|
| |
Re: OpenUP solution looking for a problem [message #38961 is a reply to message #38827] |
Fri, 31 August 2007 15:56 |
Ricardo Balduino Messages: 191 Registered: July 2009 |
Senior Member |
|
|
Tony,
This is a great community and the EPF team is proud of that. We have got
tons of positive and genuine feedback so far.
For example: today, after one month of releasing OpenUP 1.0, we have people
in other countries translating OpenUP to 3 languages (see
http://www.epfwiki.net/). We've seen through this list and the developers
mail list the testimonial of project teams and organizations adopting OpenUP
and making use of EPF Composer to capture their processes. Various companies
and consultants are participating in the project and using EPF in their
customer engagements or creating solutions using EPF (just to name a few,
see examples on Eclipse Plug-in Central:
http://www.eclipseplugincentral.com/Web_Links-index-req-view catlink-cid-878.html).
An interesting fact that was recently pointed to me is that only one in
2,000 open source projects has more than 20 active developers. We have had
more than 20 people contributing to OpenUP, not bad, huh? :)
<Source: The Ecology of Open-Source Software Development, Kieran Healy &
Alan Schussman, University of Arizona, January 14, 2003. A study of 45,000
open source projects.>"
Another fact is that many people are trying to combine agile processes out
there, since they are not complete. Mixing and matching processes is a
capability offered by EPF that allows us to address this legitimate concern
projects and organizations have: one size does not fit all. For example, see
this discussion about ideas on extending Scrum, by mixing and matching it
with other practices:
http://tech.groups.yahoo.com/group/leanagilescrum/message/28 7.
I hope these few facts help to address some of your assertions.
Regards,
Ricardo Balduino
IBM | EPF Committer
"Tony" <tony.tatil@gmail.com> wrote in message
news:20071b69f7a74bd1dba697d1ab279a26$1@www.eclipse.org...
> It may sound naive, but to me OpenUP looks like a solution looking for a
> problem.
> There are already many excellent agile and formal methodologies out there.
> What development teams need is integration of these methodologies into
> tools they use daily, not another new methodology created by methodology
> high-priests. Jazz (Team Concert) is a great example of such a tool. It
> would be even better if IBM just open sources RUP, so we don't spend any
> more time coming up with a new open source methodology.
> We may just have to accept the bitter reality that after two years (of
> marketing and promotion) there is not enough interest in OpenUp for a
> genuine community to form around it.
>
|
|
|
Re: OpenUP solution looking for a problem [message #39180 is a reply to message #38889] |
Tue, 04 September 2007 10:01 |
Roman Smirak Messages: 136 Registered: July 2009 |
Senior Member |
|
|
In our case OpenUP perfectly matches to our current needs (and trend) within
the company: get small, highly extensible, agile process merging best of
RUP, Scrum, XP and Lean.
Thanks to the team for great job!
However, Tony is right about the tools - we have to configure and integrate
several tools (Jira, SVN, MPP, MOSS, etc.); unfortunately Jazz is still
quite immature - 1.0 release is still far far away.
Roman
"Jan Fincher" <myjunk68@yahoo.com> wrote in message
news:fb94fb$a1o$1@build.eclipse.org...
> There are a number of assertions made here which I am not qualified to
> address.
>
> I can address one though. For me and my team, OpenUp coupled with EPFC is
> a very good solution to a problem we have and I'm glad I found it.
>
> Jan Fincher, Business Applications Analyst
> Florida Div. of Rehab & Liq
> jan.fincher@fldfs.com
>
> "Tony" <tony.tatil@gmail.com> wrote in message
> news:20071b69f7a74bd1dba697d1ab279a26$1@www.eclipse.org...
>> It may sound naive, but to me OpenUP looks like a solution looking for a
>> problem.
>> There are already many excellent agile and formal methodologies out
>> there. What development teams need is integration of these methodologies
>> into tools they use daily, not another new methodology created by
>> methodology high-priests. Jazz (Team Concert) is a great example of such
>> a tool. It would be even better if IBM just open sources RUP, so we
>> don't spend any more time coming up with a new open source methodology.
>> We may just have to accept the bitter reality that after two years (of
>> marketing and promotion) there is not enough interest in OpenUp for a
>> genuine community to form around it.
>>
>
>
|
|
|
Re: OpenUP solution looking for a problem [message #39210 is a reply to message #38961] |
Tue, 04 September 2007 14:02 |
Eclipse User |
|
|
|
Originally posted by: tony.tatil.gmail.com
Ricardo,
In their recent book, Wikinomics, D. Tapscott and A. Williams write about
peer production:
In its purest form, it is a way of producing goods and services that
relies entirely on self-organizing, egalitarian communities of individuals
who come together voluntarily to produce a shared outcome. In reality,
peer production mixes elements of hierarchy and self-organization i.e.,
the most skilled and experienced members of the community provide
leadership and help integrate contributions from the community.
In many peer production communities, productive activities are voluntary
and nonmonetary (page 67)
The above definition maybe the gold standard of open collaborative
production community, achieved by elite few initiatives (likes of LUNIX
and Apache). At the same time, having over 90% of contributors from IBM
and handful of its business partners is not open peer production (i.e. you
dont have 20 independent contributors) by any standard. After two years,
maybe if another independent vendor (e.g. Borland, Sun, HP, Compuware) or
hundreds of independent developers contributed some code, I would have
agreed with you. But we have very different standard for qualification of
a community as truly open and collaborative.
What motivated me in posting my previous note were the last few weeks
discussions on community outreach (which is nothing more than
promotion). You are even putting community outreach at a higher priority
than fixing bugs (wow!). Genuine open peer production projects dont need
so much recruitment activities.
I totally disagree with your statement that agile processes are not
complete. I personally have been involved in many development projects
which successfully utilized different agile processes. Also I have come
across many developers with the same experience. Eclipse (the platform
that you are based on) was developed using Eclipse Way. This is a
methodology developed by practitioners in the trenches out of real needs
and is battle tested, Rather than, conceived by generals faraway from the
combat zone. There are many other examples of amazing methodologies
developed and utilized by development teams. We dont need another new
methodology create by high-priests of methodology to prove their divinity
(or for content of their next book).
I find your statement that, EPF is addressing the mixing and matching
different agile methodologies, arrogant. There are plenty of more
qualified grassroots agile communities who can address this if there is
truly a need for it. The idea of using an arcane meta-model created by an
out-dated organization (OMG) to mix and match agile methodologies is
bizarre.
EPF looks to me like a creative marketing strategy by IBM (possibly
version 2.0 of which is called Commercial Open Source), not a genuine
peer production community. There is nothing wrong with that. But lets not
fool ourselves. Also such miss representation raises some ethical concerns.
Finally, my request to you (as IBM representative) is to open source RUP
and make it available in a wiki form. I bet my Playstation 3 that, it
would be more useful to the community than all of the last 2 years
effort. Let self-organizing communities form around this base content.
Each community can then trim and mash up this content in ways we never
imagined.
The ability to continue to produce art without permission from the
latter-day aristocracy of creativity is central to both cultural and
economic progress
. So much of what makes a free society and economy
healthy and vibrant is that we have limited the control points in a way
that permits creation and experimentation in a largely anarchistic
fashion.
(Wikinomics, page 141)
Tony
|
|
|
Re: OpenUP solution looking for a problem [message #39303 is a reply to message #39210] |
Tue, 04 September 2007 20:31 |
Ricardo Balduino Messages: 191 Registered: July 2009 |
Senior Member |
|
|
I just thought I'd write a few paragraphs to comment on some of your
assertions that might be of interest to the subscribers of this newsgroup.
It's an interesting book the one you referenced. The definition you pointed
out in page 67 of this book describes the way open source projects work,
which should as well include the Eclipse and the EPF projects to the list.
You certainly know that the EPF project consists of various components, like
the EPF Composer tool, and processes like OpenUP , Scrum, XP, DSDM and
others coming up soon that are captured using the EPF framework. All these
efforts under the EPF project have participation from volunteers from
various organizations, thus the 90% number of IBMers you mention does not
correspond to the reality. Please refer to the list of committers and
contributors to the project: http://www.eclipse.org/epf/team/team.php.
Perhaps you misunderstood the statement about agile processes completeness
First of all, those are not my words. My comment refers to a current
discussion in a agile forum that talks about gaps some of these agile
processes have and how to fill in the gaps. See this particular posting:
http://tech.groups.yahoo.com/group/leanagilescrum/message/28 5.
Second, the fact that some processes have gaps is not necessarily a bad
thing, nor one that states that a process has gaps is necessarily
criticizing the quality of that process. Some processes such as Scrum -
object of the discussion on the list above - provides guidance on project
and requirements management, so engineering practices complement what Scrum
intentionally does not provide.
It is part of EPF project vision to be a framework that helps in capturing
combining and sharing software development practices. This vision has been
supported by multiple individuals from various organizations. As an
exemplary process, OpenUP borrows (with recognition) and combines practices
from all these processes you mentioned as genuinely being developed from the
trenches, so there's nothing of arrogant or written-from-the-ivory-tower in
that perspective, as you would assume.
Your comment about community outreach is inaccurate. In that context, we
were not referring to recruitment (now, if more people desire to join the
project, they may do so at any time as usual, and the EPF team will welcome
their participation ;-)).
The context in the discussion was OpenUP, and how to address the current
Bugzilla items that were postponed for after release 1.0. The alternative to
that would be to invest the volunteers' time in creating collaterals on
OpenUP usage and deployment. As part of the discussion, we understood it
would be acceptable - after the appropriate bug triage was done - that
development of collaterals would take precedence over the remaining Bugzilla
items for OpenUP component, but again, not before appropriate bug triage.
The other components (tool and other processes) follow their own priorities
set in Bugzilla.
Ricardo Balduino.
"Tony" <tony.tatil@gmail.com> wrote in message
news:59a4e2b29c427fdafd337e1f4f734aa9$1@www.eclipse.org...
> Ricardo,
>
> In their recent book, Wikinomics, D. Tapscott and A. Williams write about
> peer production:
>
>
|
|
|
Re: OpenUP solution looking for a problem [message #39681 is a reply to message #39303] |
Thu, 06 September 2007 05:25 |
Eclipse User |
|
|
|
Originally posted by: tony.tatil.gmail.com
Ricardo,
Firstly I would like to tank you for your professional responses. Its
easier to respond to your points and clarify some of my previous points in
multiple postings.
In this post I would like to simply ask if IBM is open to open sourcing
RUP and make it available in a wiki format? As I mentioned at the end of
my previous note, this would result in formation of multiple
self-organizing communities producing content that we never imagined. The
possibilities of mixing and mashing done by nontraditional methodologies
are exhilarating. The idea is to have no preset directive, no restriction
on using EPF composer or a meta-model.
Tony
|
|
|
Re: OpenUP solution looking for a problem [message #40215 is a reply to message #39681] |
Fri, 07 September 2007 20:30 |
Ricardo Balduino Messages: 191 Registered: July 2009 |
Senior Member |
|
|
Tony, thanks for your comments.
Back in time two years ago, IBM contributed a portion of RUP known as RUP
for Small Projects and the embryo of what today we know as OpenUP (called
BUP at that time). OpenUP borrows from this RUP content, plus other agile
methods and from the practical software development experience of those who
have been contributing to the project thus far.
You have a good point about using Wiki to present process content. We are
currently making use of this technology to share and collect feedback on the
processes under the EPF project. You can see it on http://www.epfwiki.net/
where you can find a handful of processes we publish using EPF Composer. As
far as I can tell, it's not a conventional Wiki based on "unstructured
content", so to speak, because this Wiki "has the knowledge" of the
meta-model relationships that method elements need to have. The EPF Wiki has
been extensively used by the EPF community. Just to mention one example, the
Portuguese translation of OpenUP has provided around 120 pages of translated
content in a few months since it started.
Regards,
Ricardo Balduino.
"Tony" <tony.tatil@gmail.com> wrote in message
news:67fb28a9bd18e3c049f07c88d7c37154$1@www.eclipse.org...
> Ricardo,
>
> Firstly I would like to tank you for your professional responses. It
|
|
| | | |
Re: OpenUP solution looking for a problem [message #41557 is a reply to message #41533] |
Fri, 28 September 2007 15:29 |
Eclipse User |
|
|
|
Originally posted by: tony.tatil.gmail.com
In the most recent Rational Edge artical, Per says "The work was
contributed to Eclipse under the Eclipse Process Framework (EPF) 5
project, and we continued co-developing the process with roughly twenty
people from a dozen companies. Some of us had done agile implementations
of RUP, others were on the Dynamic System Development Method (DSDM) 6
board of advisors or had created Agile Model Driven Development (AMDD); 7
others practiced Scrum, 8 and some were key people behind the Eclipse
Way...".
I wonder who are the key people behind the Eclipse Way that made a major
contibution to OpenUp.
This to me is a good example of pure marketing (not enablement). Read and
decide for yourself.
http://www.ibm.com/developerworks/rational/library/sep07/kro ll/index.html
Tony
|
|
| | | |
Re: OpenUP solution looking for a problem [message #41832 is a reply to message #41771] |
Mon, 01 October 2007 14:19 |
Eclipse User |
|
|
|
Originally posted by: tony.tatil.gmail.com
Xavier,
I disagree with your statement that: Eclipse Way is only used for
developing Eclipse Plugins. In one of presentations on Jazz, the Jazz team
clearly stated that Jazz was developed using Eclipse Way (and Jazz is much
more than a plugin). To them, Eclipse Way is a methodology that is
appropriate for wide variety of projects.
On your last point, I am not convinced. Can you expand on what is the
problem OprenUp is trying solve? So far what I read from IBM (Pers team)
on the subject are mostly marketing content.
Tony
|
|
|
Re: OpenUP solution looking for a problem [message #41861 is a reply to message #41832] |
Mon, 01 October 2007 14:38 |
Xavier Méhaut Messages: 133 Registered: July 2009 |
Senior Member |
|
|
I don't want to start a method war :) Certainly EclipseWay may/can be
used for any other development than developping eclipse plugin, like
actually any other method. I just wanted to point out that EclipseWay
has been developped initially for delivering on time and in a
collaborative manner Eclipse artefacts. The noticeable fact is also that
EclipseWay has been developped/formalised in an iterative an incremental
manner too. Personnaly I like the process which is described and I use
it too.
But there is a point that EclipseWay has not well formalised (but I can
have wrong, correct me if so) is the way on how to communicate with
Customers , especially in specification phase (inception), because
Eclipse itself has not yet traditionnal customers like bank, insurances,
Aerospace companies, ... There is another point out of the scope is how
to use formalisms like UML and so on... to modelize your application. Of
course we don"t need UML to well develop an application, neither OMT,
Hood, or other formalisms. But there are cases where the customer wants
documentation with models inside,where the application is a large scale
application, or needs safety, compliance with international norms like
in aerospace, and so on... And then we need to offer a methodology which
encounter their needs. We must also recall in mind that agile modelling
needs a great agreement betwwen every parts od the project to well work...
It is why I say that OpenUP can fill in the bridge beetween a
traditional way of developping an application and the benefits we have
experimented with agile processes like EclipseWay or Scrum.
To finish, as for the language domain where there is no "universal"
language for every kind of applications, there is no "universal" method
for every kind of needs. There is only occurences of different kinds of
methods for specific needs.
So I find quite interesting to think to have like a Painter artist a
palette of methods at disposal, with different kinds of concepts, and to
create on demand the right process for the right problem. In this
vision OpenUp is a process among others, and can be valuable for certain
types of developements.
best regards
Xavier
> Xavier,
>
> I disagree with your statement that: Eclipse Way is only used for
> developing Eclipse Plugins. In one of presentations on Jazz, the Jazz
> team clearly stated that Jazz was developed using Eclipse Way (and Jazz
> is much more than a plugin). To them, Eclipse Way is a methodology that
> is appropriate for wide variety of projects.
> On your last point, I am not convinced. Can you expand on what is the
> problem OprenUp is trying solve? So far what I read from IBM (Per�s
> team) on the subject are mostly marketing content.
> Tony
>
>
|
|
|
Re: OpenUP solution looking for a problem [message #41893 is a reply to message #41861] |
Mon, 01 October 2007 15:07 |
Eclipse User |
|
|
|
Originally posted by: tony.tatil.gmail.com
Xavier,
Keep in mind, that processes are meant to be tailored before use. So in
many instances you take an agile, Unified Process (doesnt have to be RUP)
or other methodology and tailor it for your needs. For example, you can
tailor a process to better support customer intimacy. You dont take years
and develop a new methodology (reinventing the wheel) when your problem
can be solved by tailoring an existing process.
When IBM initiated the EPF project they never clearly stated the problem
that needs solving. Is it to fill the gap between RUP and Agile
methodology?
Tony
|
|
|
Re: OpenUP solution looking for a problem [message #41924 is a reply to message #41893] |
Mon, 01 October 2007 15:20 |
Xavier Méhaut Messages: 133 Registered: July 2009 |
Senior Member |
|
|
You haven't really read what I have written I think :-(
Furthermore, I'm not from IBM, and what IBM has in mind is not my
purpose. Up to now, we have 4 methods which are modelized with EPFC (and
I hope there will be more in future, and especially EclipseWay), and the
one I found (with the two others I knew - scrum and xp) is OpenUp. And
personnaly I find that this method has an interest for the reasons I've
told you in the previous mail.
Each day there are new methods which are developped. But here we have
one which is modelized with a great tool, and I find it is a good point
for this one too.
best regards
xavier
> Xavier,
>
> Keep in mind, that processes are meant to be tailored before use. So in
> many instances you take an agile, Unified Process (doesn�t have to be
> RUP) or other methodology and tailor it for your needs. For example, you
> can tailor a process to better support customer intimacy. You don�t take
> years and develop a new methodology (reinventing the wheel) when your
> problem can be solved by tailoring an existing process.
>
> When IBM initiated the EPF project they never clearly stated the problem
> that needs solving. Is it to fill the gap between RUP and Agile
> methodology?
>
> Tony
>
>
>
>
>
|
|
|
Re: OpenUP solution looking for a problem [message #41955 is a reply to message #41893] |
Mon, 01 October 2007 15:30 |
Xavier Méhaut Messages: 133 Registered: July 2009 |
Senior Member |
|
|
Tony,
To complete my previous posts, I would like to make notice that the
"subjective" part in choosing a method is very important, especially
when we have to convince our management or the customers that we should
use this method instead of another one.
I like Scrum, XP, and so on , and use them when I can.
But the notions, the vocabulary, the philosophy , the learning curve for
teams which don't know them is quite important and sometine not
"sellable" to the hierarchy.
OpenUP has this advantage to make up an agile method with traditional
concepts and vocabulary... It is also important! It is the same way why
Java is more used than Smalltalk... People had impression that it was
simpler because the vocabulary was the one of C or C++ ... but for the
advanced developpers, Smalltalk was still the best language :)
regards
xavier
> Xavier,
>
> Keep in mind, that processes are meant to be tailored before use. So in
> many instances you take an agile, Unified Process (doesn�t have to be
> RUP) or other methodology and tailor it for your needs. For example, you
> can tailor a process to better support customer intimacy. You don�t take
> years and develop a new methodology (reinventing the wheel) when your
> problem can be solved by tailoring an existing process.
>
> When IBM initiated the EPF project they never clearly stated the problem
> that needs solving. Is it to fill the gap between RUP and Agile
> methodology?
>
> Tony
>
>
>
>
>
|
|
|
Re: OpenUP solution looking for a problem [message #41985 is a reply to message #41955] |
Mon, 01 October 2007 17:53 |
Eclipse User |
|
|
|
Originally posted by: kamal.osellus.com
This is a multi-part message in MIME format.
------=_NextPart_000_0059_01C80432.666BB5F0
Content-Type: text/plain;
charset="UTF-8"
Content-Transfer-Encoding: quoted-printable
Xavier, Tony and others
=20
I see this as a great discussion in trying to identify the root problem =
domain that can be solved with something like OpenUP. Moreover by =
keeping this discussion on a technical investigative level we can make =
good progress.
=20
Xavier, I am having some trouble getting clarity on the problems you =
mentioned that make OpenUP attractive to you. Does the problem you are =
trying to solve relate to the management not being open in trying =
existing methods such as XP, Scrum etc because they don=E2=80=99t =
understand them?=20
=20
Perhaps if you share your experience in terms of a before-after =
situation it can throw some more light on this discussion and bring out =
the real value of introducing OpenUP. After going through this thread I =
have tried to summarize a few scenarios below. The first two points =
relate to OpenUP (the methodology) the third to EPFC (the tool):
1. Before being aware of OpenUP management did not try the existing =
agile-family methodologies because they did not understand them due to =
unfamiliar terms. After you showed them OpenUP they are willing to try =
it in projects because they are able to understand the methodology.=20
2. Before using OpenUP management is unconvinced that any of the =
existing agile-family methodologies will result in projects that are =
delivered on time and within budget. After using OpenUP in projects they =
see (with real examples) projects are delivered on time and within =
budget.
3. Before OpenUP you did not have a way to mix and match the applicable =
parts/modules of any available agile (or other) methodology to customize =
it for your project eco-system. After using EPFC you have a way to blend =
and tailor one or more methodologies.
=20
I am hoping if you can indicate which (one or more) of these scenarios =
reflect your situation this discussion can go deeper in defining and =
understanding that problem domain.=20
=20
Kamal
=20
--------------------------------------------
Kamal Ahluwalia
Lead Solution Specialist
Osellus Inc.
750-144 Front Street West
Toronto, Ontario M5J 3L7 CANADA
Tel: +1 (416) 603-6667 Ext 5063
Email: kamal@osellus.com
Blog: http://www.osellus.com/blogs/author/kamal
----------------------------------------------
=20
=20
"XavierMehaut" <xavier.mehaut@free.fr> wrote in message =
news:fdr3sn$rd4$1@build.eclipse.org...
Tony,
To complete my previous posts, I would like to make notice that the=20
"subjective" part in choosing a method is very important, especially=20
when we have to convince our management or the customers that we =
should=20
use this method instead of another one.
I like Scrum, XP, and so on , and use them when I can.
But the notions, the vocabulary, the philosophy , the learning curve =
for=20
teams which don't know them is quite important and sometine not=20
"sellable" to the hierarchy.
OpenUP has this advantage to make up an agile method with traditional=20
concepts and vocabulary... It is also important! It is the same way =
why=20
Java is more used than Smalltalk... People had impression that it was=20
simpler because the vocabulary was the one of C or C++ ... but for the =
advanced developpers, Smalltalk was still the best language :)
regards
xavier
> Xavier,
>=20
> Keep in mind, that processes are meant to be tailored before use. So =
in=20
> many instances you take an agile, Unified Process (doesn=EF=BF=BDt =
have to be=20
> RUP) or other methodology and tailor it for your needs. For example, =
you=20
> can tailor a process to better support customer intimacy. You =
don=EF=BF=BDt take=20
> years and develop a new methodology (reinventing the wheel) when =
your=20
> problem can be solved by tailoring an existing process.
>=20
> When IBM initiated the EPF project they never clearly stated the =
problem=20
> that needs solving. Is it to fill the gap between RUP and Agile=20
> methodology?
>=20
> Tony
>=20
>=20
>=20
>=20
>
------=_NextPart_000_0059_01C80432.666BB5F0
Content-Type: text/html;
charset="UTF-8"
Content-Transfer-Encoding: quoted-printable
=EF=BB=BF<!DOCTYPE HTML PUBLIC "-//W3C//DTD HTML 4.0 Transitional//EN">
<HTML><HEAD>
<META http-equiv=3DContent-Type content=3D"text/html; charset=3Dutf-8">
<META content=3D"MSHTML 6.00.6000.16527" name=3DGENERATOR>
<STYLE></STYLE>
</HEAD>
<BODY bgColor=3D#ffffff>
<DIV><FONT face=3DArial size=3D2>
<P class=3DMsoNormal style=3D"MARGIN: 0cm 0cm 0pt">Xavier, Tony and=20
others<?xml:namespace prefix =3D o ns =3D =
"urn:schemas-microsoft-com:office:office"=20
/><o:p></o:p></P>
<P class=3DMsoNormal style=3D"MARGIN: 0cm 0cm 0pt"><o:p> </o:p></P>
<P class=3DMsoNormal style=3D"MARGIN: 0cm 0cm 0pt">I see this as a great =
discussion=20
in trying to identify the root problem domain that can be solved with =
something=20
like OpenUP. Moreover by keeping this discussion on a technical =
investigative=20
level we can make good progress.<o:p></o:p></P>
<P class=3DMsoNormal style=3D"MARGIN: 0cm 0cm 0pt"><o:p> </o:p></P>
<P class=3DMsoNormal style=3D"MARGIN: 0cm 0cm 0pt">Xavier, I am having =
some=20
trouble getting clarity on the problems you mentioned that =
make OpenUP=20
attractive to you. Does the problem you are trying to solve relate to =
the=20
management not being open in trying existing methods such as XP, =
Scrum etc=20
because they don=E2=80=99t understand them? <o:p></o:p></P>
<P class=3DMsoNormal style=3D"MARGIN: 0cm 0cm 0pt"><o:p> </o:p></P>
<P class=3DMsoNormal style=3D"MARGIN: 0cm 0cm 0pt">Perhaps if you share =
your=20
experience in terms of a before-after situation it can throw some more =
light on=20
this discussion and bring out the real value of introducing OpenUP. =
After=20
going through this thread I have tried to summarize a few scenarios =
below.=20
The first two points relate to OpenUP (the methodology) the third =
to EPFC=20
(the tool):</P>
<P class=3DMsoNormal style=3D"MARGIN: 0cm 0cm 0pt"><o:p></o:p> </P>
<P class=3DMsoNormal style=3D"MARGIN: 0cm 0cm 0pt"><SPAN=20
style=3D"mso-fareast-font-family: Calibri; mso-bidi-font-family: =
Calibri"><SPAN=20
style=3D"mso-list: Ignore">1.<SPAN style=3D"FONT: 7pt 'Times New =
Roman'"><FONT=20
face=3DArial size=3D2> </FONT></SPAN></SPAN></SPAN>Before being =
aware of OpenUP=20
management did not try the existing agile-family methodologies =
because they=20
did not understand them due to unfamiliar terms. After you showed them =
OpenUP=20
they are willing to try it in projects because they are able to =
understand the=20
methodology. </P>
<P class=3DMsoNormal style=3D"MARGIN: 0cm 0cm 0pt"><SPAN=20
style=3D"mso-fareast-font-family: Calibri; mso-bidi-font-family: =
Calibri"><SPAN=20
style=3D"mso-list: Ignore">2. </SPAN></SPAN>Before using OpenUP =
management is=20
unconvinced that any of the existing agile-family methodologies will =
result in=20
projects that are delivered on time and within budget. After using =
OpenUP in=20
projects they see (with real examples) projects are delivered on time =
and within=20
budget.</P>
<P class=3DMsoNormal style=3D"MARGIN: 0cm 0cm 0pt"><SPAN=20
style=3D"mso-fareast-font-family: Calibri; mso-bidi-font-family: =
Calibri"><SPAN=20
style=3D"mso-list: Ignore">3. </SPAN></SPAN>Before OpenUP you did not =
have a way=20
to mix and match the applicable parts/modules of any available =
agile (or=20
other) methodology to customize it for your project eco-system. After =
using EPFC=20
you have a way to blend and tailor one or more =
methodologies.<o:p></o:p></P>
<P class=3DMsoNormal style=3D"MARGIN: 0cm 0cm 0pt"><o:p> </o:p></P>
<P class=3DMsoNormal style=3D"MARGIN: 0cm 0cm 0pt">I am hoping if you =
can indicate=20
which (one or more) of these scenarios reflect your situation this =
discussion=20
can go deeper in defining and understanding that problem domain. =
<o:p></o:p></P>
<P class=3DMsoNormal style=3D"MARGIN: 0cm 0cm 0pt"><o:p> </o:p></P>
<P class=3DMsoNormal style=3D"MARGIN: 0cm 0cm 0pt">Kamal<o:p></o:p></P>
<P class=3DMsoNormal style=3D"MARGIN: 0cm 0cm 0pt"><o:p> </o:p></P>
<P class=3DMsoNormal style=3D"MARGIN: 0cm 0cm 0pt"><SPAN lang=3DEN-US=20
style=3D"FONT-SIZE: 10pt; COLOR: navy; FONT-FAMILY: =
'Arial','sans-serif'; mso-ansi-language: =
EN-US">--------------------------------------------</SPAN> <SPAN=20
lang=3DEN-US style=3D"mso-ansi-language: EN-US"><o:p></o:p></SPAN></P>
<P class=3DMsoNormal style=3D"MARGIN: 0cm 0cm 0pt"><SPAN lang=3DEN-US=20
style=3D"FONT-SIZE: 10pt; COLOR: black; FONT-FAMILY: =
'Arial','sans-serif'; mso-ansi-language: EN-US">Kamal=20
Ahluwalia</SPAN><SPAN lang=3DEN-US=20
style=3D"mso-ansi-language: EN-US"><o:p></o:p></SPAN></P>
<P class=3DMsoNormal style=3D"MARGIN: 0cm 0cm 0pt"><SPAN lang=3DEN-US=20
style=3D"FONT-SIZE: 10pt; COLOR: black; FONT-FAMILY: =
'Arial','sans-serif'; mso-ansi-language: EN-US">Lead=20
Solution Specialist<o:p></o:p></SPAN></P>
<P class=3DMsoNormal style=3D"MARGIN: 0cm 0cm 0pt"><SPAN lang=3DEN-US=20
style=3D"FONT-SIZE: 10pt; COLOR: black; FONT-FAMILY: =
'Arial','sans-serif'; mso-ansi-language: EN-US">Osellus=20
Inc.<o:p></o:p></SPAN></P>
<P class=3DMsoNormal style=3D"MARGIN: 0cm 0cm 0pt"><SPAN lang=3DEN-US=20
style=3D"FONT-SIZE: 10pt; FONT-FAMILY: 'Arial','sans-serif'; =
mso-ansi-language: EN-US">750-144=20
Front Street West</SPAN><SPAN lang=3DEN-US=20
style=3D"FONT-SIZE: 12pt; FONT-FAMILY: 'Times New Roman','serif'; =
mso-ansi-language: EN-US"><o:p></o:p></SPAN></P>
<P class=3DMsoNormal style=3D"MARGIN: 0cm 0cm 0pt"><SPAN lang=3DEN-US=20
style=3D"FONT-SIZE: 10pt; FONT-FAMILY: 'Arial','sans-serif'; =
mso-ansi-language: EN-US">Toronto,=20
Ontario M5J 3L7 CANADA</SPAN><SPAN lang=3DEN-US=20
style=3D"mso-ansi-language: EN-US"><o:p></o:p></SPAN></P>
<P class=3DMsoNormal style=3D"MARGIN: 0cm 0cm 0pt"><SPAN lang=3DEN-US=20
style=3D"FONT-SIZE: 10pt; COLOR: black; FONT-FAMILY: =
'Arial','sans-serif'; mso-ansi-language: EN-US">Tel:=20
+1 (416) 603-6667 Ext 5063</SPAN><SPAN lang=3DEN-US=20
style=3D"mso-ansi-language: EN-US"><o:p></o:p></SPAN></P>
<P class=3DMsoNormal style=3D"MARGIN: 0cm 0cm 0pt"><SPAN lang=3DEN-US=20
style=3D"FONT-SIZE: 10pt; FONT-FAMILY: 'Arial','sans-serif'; =
mso-ansi-language: EN-US">Email:=20
<A =
title=3D"blocked::mailto:kamal@osellus.com mailto:kamal@osellus.com" =
href=3D"mailto:kamal@osellus.com"><FONT=20
color=3D#0000ff>kamal@osellus.com</FONT></A></SPAN><SPAN lang=3DEN-US=20
style=3D"mso-ansi-language: EN-US"><o:p></o:p></SPAN></P>
<P class=3DMsoNormal style=3D"MARGIN: 0cm 0cm 0pt"><SPAN lang=3DEN-US=20
style=3D"FONT-SIZE: 10pt; FONT-FAMILY: 'Arial','sans-serif'; =
mso-ansi-language: EN-US">Blog:=20
<A title=3Dblocked::http://www.osellus.com/blogs/author/kamal=20
href=3D"http://www.osellus.com/blogs/author/kamal">http://www.osellus.com=
/blogs/author/kamal</A><o:p></o:p></SPAN></P>
<P class=3DMsoNormal style=3D"MARGIN: 0cm 0cm 0pt"><SPAN lang=3DEN-US=20
style=3D"FONT-SIZE: 10pt; COLOR: navy; FONT-FAMILY: =
'Arial','sans-serif'; mso-ansi-language: =
EN-US">----------------------------------------------</SPAN ><SPAN=20
lang=3DEN-US=20
style=3D"FONT-SIZE: 12pt; FONT-FAMILY: 'Times New Roman','serif'; =
mso-ansi-language: EN-US"><o:p></o:p></SPAN></P>
<P class=3DMsoNormal style=3D"MARGIN: 0cm 0cm 0pt"><o:p><FONT =
face=3DCalibri=20
size=3D3> </FONT></o:p></P>
<P class=3DMsoNormal style=3D"MARGIN: 0cm 0cm 0pt"><SPAN=20
style=3D"mso-bidi-font-family: 'Times New Roman'; mso-ascii-font-family: =
Calibri; mso-ascii-theme-font: minor-latin; mso-hansi-font-family: =
Calibri; mso-hansi-theme-font: minor-latin; mso-bidi-theme-font: =
minor-bidi"><o:p><FONT=20
face=3DCalibri size=3D3> </FONT></o:p></SPAN></P></FONT></DIV >
<BLOCKQUOTE=20
style=3D"PADDING-RIGHT: 0px; PADDING-LEFT: 5px; MARGIN-LEFT: 5px; =
BORDER-LEFT: #000000 2px solid; MARGIN-RIGHT: 0px">
<DIV>"XavierMehaut" <<A=20
href=3D"mailto:xavier.mehaut@free.fr">xavier.mehaut@free.fr</A>> =
wrote in=20
message <A=20
=
href=3D"news:fdr3sn$rd4$1@build.eclipse.org">news:fdr3sn$rd4$1@build.ecli=
pse.org</A>...</DIV>Tony,<BR>To=20
complete my previous posts, I would like to make notice that the=20
<BR>"subjective" part in choosing a method is very important, =
especially=20
<BR>when we have to convince our management or the customers that we =
should=20
<BR>use this method instead of another one.<BR>I like Scrum, XP, and =
so on ,=20
and use them when I can.<BR>But the notions, the vocabulary, the =
philosophy , the learning curve for <BR>teams which don't know them is =
quite=20
important and sometine not <BR>"sellable" to the hierarchy.<BR>OpenUP =
has this=20
advantage to make up an agile method with traditional <BR>concepts and =
vocabulary... It is also important! It is the same way why <BR>Java is =
more=20
used than Smalltalk... People had impression that it was <BR>simpler =
because=20
the vocabulary was the one of C or C++ ... but for the <BR>advanced=20
developpers, Smalltalk was still the best language=20
:)<BR><BR>regards<BR>xavier<BR><BR>> Xavier,<BR>> <BR>> Keep =
in mind,=20
that processes are meant to be tailored before use. So in <BR>> =
many=20
instances you take an agile, Unified Process (doesn=EF=BF=BDt have to =
be <BR>> RUP)=20
or other methodology and tailor it for your needs. For example, you =
<BR>>=20
can tailor a process to better support customer intimacy. You =
don=EF=BF=BDt take=20
<BR>> years and develop a new methodology (reinventing the wheel) =
when your=20
<BR>> problem can be solved by tailoring an existing =
process.<BR>>=20
<BR>> When IBM initiated the EPF project they never clearly stated =
the=20
problem <BR>> that needs solving. Is it to fill the gap between RUP =
and=20
Agile <BR>> methodology?<BR>> <BR>> Tony<BR>> <BR>> =
<BR>>=20
<BR>> <BR>></BLOCKQUOTE></BODY></HTML>
------=_NextPart_000_0059_01C80432.666BB5F0--
|
|
|
Re: OpenUP solution looking for a problem [message #42018 is a reply to message #41557] |
Mon, 01 October 2007 19:05 |
Peter Haumer Messages: 228 Registered: July 2009 |
Senior Member |
|
|
Hello.
Per and I worked very closely with the Jazz team (in particular Erich Gamma
and Kai-Uwe Maetzel) on formalizing Eclipse Way for EPF Composer as well as
defining integration points with Jazz for content as well as tools (e.g. the
ones we previewed at the RSDC in Florida this year). During this
collaboration a lot of feedback went both ways. However, principally
Eclipse Way and OpenUP have two different goals. Eclipse Way aims at
'precisely' describing how the Eclipse teams work. OpenUP is a generic
process framework that provides an essential and consistent collection of
best practices for development teams to use as a starting point to develop
and tailor their own processes. You can use OpenUP as is, but probably most
teams want to tailor it to their needs. Eclipse Way aims at describing only
the practices that are used in the development of Eclipse instead. In that
sense it is already the tailored 'end product', the process in use by the
Eclipse teams. You can see the clear influence the Eclipse practices had on
OpenUP looking at work items, micro iterations, stable builds, builds,
iteration reviews and feedback etc. The other way round the Eclipse and
Jazz teams started looking at Risk management after reviewing OpenUP with us
as well as Jazz' review workflows are close to RUP's compliance content. In
respect to AJ's comment it is correct that the EMF or GMF teams would not
say that they use OpenUP, but certainly OpenUP represents a lot of their
practices. Hence, for people interested in learning about such practices
OpenUP is a good way to start looking plus EPF Composer allows tailoring and
enriching that content as well.
Peter.
"Tony" <tony.tatil@gmail.com> wrote in message
news:61813296b370f62fcb59821f579dd20e$1@www.eclipse.org...
> In the most recent Rational Edge artical, Per says "The work was
> contributed to Eclipse under the Eclipse Process Framework (EPF) 5
> project, and we continued co-developing the process with roughly twenty
> people from a dozen companies. Some of us had done agile implementations
> of RUP, others were on the Dynamic System Development Method (DSDM) 6
> board of advisors or had created Agile Model Driven Development (AMDD); 7
> others practiced Scrum, 8 and some were key people behind the Eclipse
> Way...".
> I wonder who are the key people behind the Eclipse Way that made a major
> contibution to OpenUp.
>
> This to me is a good example of pure marketing (not enablement). Read and
> decide for yourself.
>
> http://www.ibm.com/developerworks/rational/library/sep07/kro ll/index.html
>
>
> Tony
>
|
|
|
Re: OpenUP solution looking for a problem [message #42080 is a reply to message #41985] |
Mon, 01 October 2007 19:22 |
Xavier Méhaut Messages: 133 Registered: July 2009 |
Senior Member |
|
|
Hi Kamal,
I see that this thread goes further than I wanted to go :) and my
different answers were not a general answer built with a specific intent
but were mere brought by the flow of the discussion.
I repeat myself in other words, I'm not an evangelist of OpenUP and I
just try to give counter-arguments to Tony to show that OpenUp is not a
bad method because it is made by people from IBM/Rational, and because
some roots are from RUP.
I'm method agnostic, because I used to use many ones from SART, SADT,
SA, Schlaer and Mellor, Merise, Hood, Hoora, AonixWay, and so on... With
time, we relativise the use of methods, and we only know that we need
some to develop apllications. If we look closer to all thoses methods,
we can see that there are finally quite few concepts to handle. And the
interesting part of a methodologist work is to arrange these few
concepts to make a process the most efficient as possible in the context
in which we are.
So the arguments I gave for OpenUP are some arguments among others to
show that a method must be tailored as told by Tony to match the
customer and development team needs. I could also add as arguments for
OpenUp the fact that the method insists on the architecture, which is a
very important point in the case of a quite big project.
In conclusion, OpenUp is :
- very young
- not yet tooled
- not yet implemented in public projects
- not a rupture with other methods
but
- exists,
- has a team for working on it
- can be enhanced, improved, derived thnaks to EPF (like Scrum or XP or
DSDM up to now)
- is clear
- has no visible bottleneck for the moment (or I haven't yet seen them)
- is "sellable" to a hierarchy
It is enough to give a chance to this method, isn't it?
Best regards
Xavier
>
>
> Xavier, Tony and others
>
>
>
> I see this as a great discussion in trying to identify the root problem
> domain that can be solved with something like OpenUP. Moreover by
> keeping this discussion on a technical investigative level we can make
> good progress.
>
>
>
> Xavier, I am having some trouble getting clarity on the problems you
> mentioned that make OpenUP attractive to you. Does the problem you are
> trying to solve relate to the management not being open in trying
> existing methods such as XP, Scrum etc because they don’t understand them?
>
>
>
> Perhaps if you share your experience in terms of a before-after
> situation it can throw some more light on this discussion and bring out
> the real value of introducing OpenUP. After going through this thread I
> have tried to summarize a few scenarios below. The first two points
> relate to OpenUP (the methodology) the third to EPFC (the tool):
>
>
>
> 1. Before being aware of OpenUP management did not try the existing
> agile-family methodologies because they did not understand them due to
> unfamiliar terms. After you showed them OpenUP they are willing to try
> it in projects because they are able to understand the methodology.
>
> 2. Before using OpenUP management is unconvinced that any of the
> existing agile-family methodologies will result in projects that are
> delivered on time and within budget. After using OpenUP in projects they
> see (with real examples) projects are delivered on time and within budget.
>
> 3. Before OpenUP you did not have a way to mix and match the applicable
> parts/modules of any available agile (or other) methodology to customize
> it for your project eco-system. After using EPFC you have a way to blend
> and tailor one or more methodologies.
>
>
>
> I am hoping if you can indicate which (one or more) of these scenarios
> reflect your situation this discussion can go deeper in defining and
> understanding that problem domain.
>
>
>
> Kamal
>
>
>
> --------------------------------------------
>
> Kamal Ahluwalia
>
> Lead Solution Specialist
>
> Osellus Inc.
>
> 750-144 Front Street West
>
> Toronto, Ontario M5J 3L7 CANADA
>
> Tel: +1 (416) 603-6667 Ext 5063
>
> Email: kamal@osellus.com <mailto:kamal@osellus.com>
>
> Blog: http://www.osellus.com/blogs/author/kamal
>
> ----------------------------------------------
>
>
>
>
>
> "XavierMehaut" <xavier.mehaut@free.fr
> <mailto:xavier.mehaut@free.fr>> wrote in message
> news:fdr3sn$rd4$1@build.eclipse.org...
> Tony,
> To complete my previous posts, I would like to make notice that the
> "subjective" part in choosing a method is very important, especially
> when we have to convince our management or the customers that we should
> use this method instead of another one.
> I like Scrum, XP, and so on , and use them when I can.
> But the notions, the vocabulary, the philosophy , the learning curve
> for
> teams which don't know them is quite important and sometine not
> "sellable" to the hierarchy.
> OpenUP has this advantage to make up an agile method with traditional
> concepts and vocabulary... It is also important! It is the same way why
> Java is more used than Smalltalk... People had impression that it was
> simpler because the vocabulary was the one of C or C++ ... but for the
> advanced developpers, Smalltalk was still the best language :)
>
> regards
> xavier
>
> > Xavier,
> >
> > Keep in mind, that processes are meant to be tailored before use.
> So in
> > many instances you take an agile, Unified Process (doesn�t have
> to be
> > RUP) or other methodology and tailor it for your needs. For
> example, you
> > can tailor a process to better support customer intimacy. You
> don�t take
> > years and develop a new methodology (reinventing the wheel) when
> your
> > problem can be solved by tailoring an existing process.
> >
> > When IBM initiated the EPF project they never clearly stated the
> problem
> > that needs solving. Is it to fill the gap between RUP and Agile
> > methodology?
> >
> > Tony
> >
> >
> >
> >
> >
|
|
|
Re: OpenUP solution looking for a problem [message #42143 is a reply to message #42018] |
Tue, 02 October 2007 02:45 |
Eclipse User |
|
|
|
Originally posted by: tony.tatil.gmail.com
Hi Peter,
In my previous note I was referring to Pers article titled OpenUp In a
Nutshell where he says
some of key people behind Eclipse way
. What
he is claming by such a statement is that some of the key people who
developed Eclipse way are part of OpenUP development. It may be just me,
but thats how it reads.
So now you are saying that you and Per are the key people behind Eclipse
way. I wonder what Eric thinks about that.
Tony
|
|
| | |
Re: OpenUP solution looking for a problem [message #42312 is a reply to message #42208] |
Tue, 02 October 2007 13:48 |
Eclipse User |
|
|
|
Originally posted by: tony.tatil.gmail.com
Hi Peter,
What I am getting at is an extremely serious matter in both academic and
corporate worlds, which is knowingly giving the public false information.
The statement some of key people behind Eclipse way is false. I would
like IBM, Eclipse Board of Directors and Rational Edge editors to look
into this.
What is needed here is better governance.
Tony
|
|
|
Re: OpenUP solution looking for a problem [message #42438 is a reply to message #42312] |
Tue, 02 October 2007 18:49 |
Per Kroll Messages: 60 Registered: July 2009 |
Member |
|
|
Tony,
The intent with the sentence was to acknowledge a diverse background of
people, and influences from many sources. It is important to acknolwedge
that some content from Eclipse Way is in OpenUP. Specifically, the iteration
lifecycle, which is key to how to effectively manage an iteration so you end
up delivering a useful executable of high quality is from Eclipse Way, see
Figure 2 in
http://www.ibm.com/developerworks/rational/library/sep07/kro ll/index.html.
The statement "some were key people behind the Eclipse Way" is however
incorrect. We did work with some key people behind Eclipse Way, which
resulted in content making its way into OpenUP, but they did not themself
contribute directly to OpenUP. The rest of the sentence is correct. "Some of
us had done agile implementations of RUP, others were on the Dynamic System
Development Method (DSDM) 6 board of advisors or had created Agile Model
Driven Development (AMDD); 7 others practiced Scrum, 8 .... "
So, thanks for finding the mistake, I will have it corrected.
Cheers
/Per
"Tony" <tony.tatil@gmail.com> wrote in message
news:05877041f556bfc34d1c48c9cc13ba94$1@www.eclipse.org...
> Hi Peter,
>
> What I am getting at is an extremely serious matter in both academic and
> corporate worlds, which is knowingly giving the public false information.
> The statement
|
|
| | | |
Re: OpenUP solution looking for a problem [message #42558 is a reply to message #41579] |
Thu, 04 October 2007 03:33 |
Eclipse User |
|
|
|
Originally posted by: tony.tatil.gmail.com
It's extremely alarming that IBM has been the only contributor to EPF
Composer for the last two years. It has been and is a useful tool for
some users but it clearly doesn't deserve to be an Eclipse project like
many of the other multi-organization supported projects.
No matter how much RMC team muddies the water, there is only one big blue
fish in EPF Composer pond.
I call this "crony open source"
Tony
Shilpa Toraskar wrote:
> As far as I know, it's mostly IBM who contributed to EPF code. Of course
> for content (OpenUp) we have lot of non-IBM committers.
> Shilpa
|
|
|
Re: OpenUP solution looking for a problem [message #42619 is a reply to message #42558] |
Thu, 04 October 2007 17:51 |
AJ Bencomo Messages: 42 Registered: July 2009 |
Member |
|
|
Tony,
I am not sure what you are actually trying to say with "It's extremely
alarming that IBM has been the only contributor to EPF Composer". It
seems to me that you are trying to be either extremely
yellow/controversial or perhaps negative. So, I personally have to
disagree with you on this matter.
EPF is not the only tool/project supported by only IBMers. For instance,
the EMF project is remarkably well sustained by only 5 IBM developers, but
it has been embarrassed by multiple individuals and organizations.
It is obvious that you have a misconception/misunderstanding of why
projects/tools/code are donated to the Eclipse Open Source community.
The success of an open source project (at least in the Eclipse community)
is not determined or measured by the number of developer committing code.
Instead, it is determined by the benefits that it brings to the open
source community. So, I think EPF has clearly accomplish that.
Therefore, when you call it a "crony open source", basically you are
saying that the eclipse SDK is also crony because who do you think is
doing most the code development? Microsoft?
Ciao,
-Alfredo
BTW, I don't work for IBM ;-)
|
|
|
Re: OpenUP solution looking for a problem [message #42712 is a reply to message #42619] |
Fri, 05 October 2007 03:23 |
Eclipse User |
|
|
|
Originally posted by: tony.tatil.gmail.com
Alfredo,
You are right, I need to clarify some of my previous points. Actually I
am very glad that you responded intelligently to my note. Intelligent,
in-depth and open discussions are the best way to learn.
As you know IBM, like many other IT vendors, is a public for-profit
company. Therefore, its first obligation is maximization of shareholder
value. Since, IBM has very good governance, all its business units have to
make sure their projects are aligned with the overall business objectives.
So, its realistic that next quarter a decision is made to move the
resources assigned to EPF to another more critical or lucrative project.
Many of us who are, or have been, working for large companies know this is
very possible.
At such a point if there is no IBM involvement, who is going to answer
your technical questions, who is going to fix bugs and who is going
enhance the tool? There is nobody else right now beside IBM employees who
get paid to do this work. You have spent the time learning how to use a
complex tool (plug-ins, variability, configurations etc.) and entered all
your methods. Now you are stuck with a dead-end tool. This is a major
risk that users are taking today. So, it may be helpful if we acknowledge
this risk (users be aware), or IBM can commit to support EPF for the
next two years.
I dont know if you are aware of the debates on the topics of IBM Eclipse
support and single vendor Eclipse projects at the Eclipse board of
director level. The Eclipse community has been alarmed by IBM focus on
Jazz (a commercial product) and shifting resources from Eclipse, hence
slowing the Eclipse related development. Also in their very latest
meeting the topic of single vendor project was discussed:
http://www.eclipse.org/org/foundation/boardminutes/2007_06_2 0-21_Minutes.php
From the second last paragraph in this page:
Mike Milinkovich introduced the topic of vendor neutrality.
Essentially in some circumstances we are sourcing open source code
originating from a single vendor and that is something we should
specifically consider the implications of. The lack of a definition of
open source, what constitutes a neutral project, and the fact that
this state could change over time was discussed and the associated
problems of managing any such distinction.
.
As far as I am concerned any Eclipse project (not just EPF) that is
dependent on a single vendor has the same problem.
How did you come up with the number of IBMers supporting this effort? Are
there really only 5 developers writing all the Java code for EPF composer?
There are more issues with EPF Composer being labeled as open source which
Ill cover in future posts.
Sorry for the long post,
Tony
BTW its not important to me who you work for. Only what you have to say
is important.
|
|
| |
Re: OpenUP solution looking for a problem [message #42809 is a reply to message #42778] |
Fri, 05 October 2007 16:17 |
Eclipse User |
|
|
|
Originally posted by: tony.tatil.gmail.com
Ana,
Firstly, it is not my intension to be negative. As I stated multiple
times before, deep discussion helps everyone. We should be open to it.
I disagree with you on innovativeness of EPF Composer. Its a traditional
thick client application in an on-demand web application age. Just look at
what Adobe is doing with its Photoshop Lite. A rich media application like
Photoshop running in a browser. Now thats what I call innovative. I
would get very excited if someone works on web-based version of EPF
Composer. The other problem is that EPF Composer and the meta-model are
unnecessarily complex and not flexible. Some of the OpenUP concepts such
as work-items are not supported by the tool and the meta-model.
Also, it sounds like more and more wiki technology is being used for
OpenUP on going development.
In my opinion, OpenUP, or any methodology for that matter, is not a
suitable Eclipse community project. Keep in mind even Eclipse way is not
an Eclipse project. The best model for such a content-intensive
collaborative development is Wiki. Wikipidia is the best example of a very
successful and universally beneficial content-intensive project. Maybe
the RUP team selected Eclipse because its IBM friendly environment.
Hopefully IBM and other members of this project have open mind and
consider the above suggestions.
Ana, I have read carefully every one of the teams meeting minutes and
have my own idea how OpenUP 1.0 evolved from the initial IBM contribution.
You may remember since you were part of these discussions that IBM and
other members felt that it initial version was not innovative enough
(nothing new to offer) and not marketable so it was changed to look more
interesting (more window dressing). If you want I can provide you with
quotations from these meetings.
BTW, I came up with Crony Open Source from a very popular term Crony
Capitalism which was coined to describe South East Asian regimes. They
are skillfully made to look capitalist from outside but they are far from
it. Their economy is controlled by a single dictator, benefiting his
family member and close associates.
Tony
|
|
|
Re: OpenUP solution looking for a problem [message #42839 is a reply to message #42809] |
Sun, 07 October 2007 18:00 |
Ana Valente Pereira Messages: 18 Registered: July 2009 |
Junior Member |
|
|
Well ... it really sounded really negative to me ... maybe because I am
not a native English speaker and I may not had fully understood the
meaning of words like "crony"
.... also I was not referring to the innovativeness of EPF Composer ...
or OpenUP ... what I consider an innovation is the "EPF project" ...
a path to an open way of defining processes ... I find very interesting
and powerful the idea of having open source applied to processes and not
only to software products... even if this project has not created
something very "innovative" so far I really believe in its potential
Ana
Tony wrote:
> Ana,
>
> Firstly, it is not my intension to be negative. As I stated multiple
> times before, deep discussion helps everyone. We should be open to it.
> I disagree with you on innovativeness of EPF Composer. It�s a
> traditional thick client application in an on-demand web application
> age. Just look at what Adobe is doing with its Photoshop Lite. A rich
> media application like Photoshop running in a browser. Now that�s what
> I call innovative. I would get very excited if someone works on
> web-based version of EPF Composer. The other problem is that EPF
> Composer and the meta-model are unnecessarily complex and not flexible.
> Some of the OpenUP concepts such as work-items are not supported by the
> tool and the meta-model.
> Also, it sounds like more and more wiki technology is being used for
> OpenUP on going development.
>
> In my opinion, OpenUP, or any methodology for that matter, is not a
> suitable Eclipse community project. Keep in mind even Eclipse way is not
> an Eclipse project. The best model for such a content-intensive
> collaborative development is Wiki. Wikipidia is the best example of a
> very successful and universally beneficial content-intensive project.
> Maybe the RUP team selected Eclipse because it�s IBM friendly environment.
> Hopefully IBM and other members of this project have open mind and
> consider the above suggestions.
>
> Ana, I have read carefully every one of the team�s meeting minutes and
> have my own idea how OpenUP 1.0 evolved from the initial IBM
> contribution. You may remember since you were part of these discussions
> that IBM and other members felt that it initial version was not
> innovative enough (nothing new to offer) and not marketable so it was
> changed to look more interesting (more window dressing). If you want I
> can provide you with quotations from these meetings.
> BTW, I came up with �Crony Open Source� from a very popular term �Crony
> Capitalism� which was coined to describe South East Asian regimes. They
> are skillfully made to look capitalist from outside but they are far
> from it. Their economy is controlled by a single dictator, benefiting
> his family member and close associates.
>
> Tony
>
>
>
|
|
|
Re: EclipseWay in EPFC [message #43322 is a reply to message #42493] |
Tue, 09 October 2007 17:54 |
Eclipse User |
|
|
|
Originally posted by: david.osellus.com
Hello Xavier,
We were also interested in using The Eclipse Way process. After coming
across your post, I posted a question on the jazz.net newsgroup regarding
the licensing required to use the process and received a reply stating
that The Eclipse Way is currently only available from jazz.net and is
covered by the terms of the license required for jazz.net so it seems that
they may not be making The Eclipse Way publicly available.
Regards,
Dave
|
|
| |
Re: EclipseWay in EPFC [message #43587 is a reply to message #43497] |
Wed, 10 October 2007 18:22 |
Eclipse User |
|
|
|
Originally posted by: tony.tatil.gmail.com
It is indeed very strange that a methodology which has come about as a
result of open source development is protected by copyright itself. Maybe
it was just a slip on Jazz's team part.
It would be great if Per or Peter use their influence with the Eclipse Way
and Jazz teams and convince them to make Eclipse Way available under
Eclipse licensing agreement.
Tony
|
|
| |
Re: OpenUP solution looking for a problem [message #582421 is a reply to message #38827] |
Fri, 31 August 2007 15:56 |
Ricardo Balduino Messages: 191 Registered: July 2009 |
Senior Member |
|
|
Tony,
This is a great community and the EPF team is proud of that. We have got
tons of positive and genuine feedback so far.
For example: today, after one month of releasing OpenUP 1.0, we have people
in other countries translating OpenUP to 3 languages (see
http://www.epfwiki.net/). We've seen through this list and the developers
mail list the testimonial of project teams and organizations adopting OpenUP
and making use of EPF Composer to capture their processes. Various companies
and consultants are participating in the project and using EPF in their
customer engagements or creating solutions using EPF (just to name a few,
see examples on Eclipse Plug-in Central:
http://www.eclipseplugincentral.com/Web_Links-index-req-view catlink-cid-878.html).
An interesting fact that was recently pointed to me is that only one in
2,000 open source projects has more than 20 active developers. We have had
more than 20 people contributing to OpenUP, not bad, huh? :)
<Source: The Ecology of Open-Source Software Development, Kieran Healy &
Alan Schussman, University of Arizona, January 14, 2003. A study of 45,000
open source projects.>"
Another fact is that many people are trying to combine agile processes out
there, since they are not complete. Mixing and matching processes is a
capability offered by EPF that allows us to address this legitimate concern
projects and organizations have: one size does not fit all. For example, see
this discussion about ideas on extending Scrum, by mixing and matching it
with other practices:
http://tech.groups.yahoo.com/group/leanagilescrum/message/28 7
I hope these few facts help to address some of your assertions.
Regards,
Ricardo Balduino
IBM | EPF Committer
"Tony" <tony.tatil@gmail.com> wrote in message
news:20071b69f7a74bd1dba697d1ab279a26$1@www.eclipse.org...
> It may sound naive, but to me OpenUP looks like a solution looking for a
> problem.
> There are already many excellent agile and formal methodologies out there.
> What development teams need is integration of these methodologies into
> tools they use daily, not another new methodology created by methodology
> high-priests. Jazz (Team Concert) is a great example of such a tool. It
> would be even better if IBM just open sources RUP, so we don't spend any
> more time coming up with a new open source methodology.
> We may just have to accept the bitter reality that after two years (of
> marketing and promotion) there is not enough interest in OpenUp for a
> genuine community to form around it.
>
|
|
|
Re: OpenUP solution looking for a problem [message #582493 is a reply to message #38889] |
Tue, 04 September 2007 10:01 |
Roman Smirak Messages: 136 Registered: July 2009 |
Senior Member |
|
|
In our case OpenUP perfectly matches to our current needs (and trend) within
the company: get small, highly extensible, agile process merging best of
RUP, Scrum, XP and Lean.
Thanks to the team for great job!
However, Tony is right about the tools - we have to configure and integrate
several tools (Jira, SVN, MPP, MOSS, etc.); unfortunately Jazz is still
quite immature - 1.0 release is still far far away.
Roman
"Jan Fincher" <myjunk68@yahoo.com> wrote in message
news:fb94fb$a1o$1@build.eclipse.org...
> There are a number of assertions made here which I am not qualified to
> address.
>
> I can address one though. For me and my team, OpenUp coupled with EPFC is
> a very good solution to a problem we have and I'm glad I found it.
>
> Jan Fincher, Business Applications Analyst
> Florida Div. of Rehab & Liq
> jan.fincher@fldfs.com
>
> "Tony" <tony.tatil@gmail.com> wrote in message
> news:20071b69f7a74bd1dba697d1ab279a26$1@www.eclipse.org...
>> It may sound naive, but to me OpenUP looks like a solution looking for a
>> problem.
>> There are already many excellent agile and formal methodologies out
>> there. What development teams need is integration of these methodologies
>> into tools they use daily, not another new methodology created by
>> methodology high-priests. Jazz (Team Concert) is a great example of such
>> a tool. It would be even better if IBM just open sources RUP, so we
>> don't spend any more time coming up with a new open source methodology.
>> We may just have to accept the bitter reality that after two years (of
>> marketing and promotion) there is not enough interest in OpenUp for a
>> genuine community to form around it.
>>
>
>
|
|
|
Re: OpenUP solution looking for a problem [message #582509 is a reply to message #38961] |
Tue, 04 September 2007 14:02 |
Tony Messages: 52 Registered: July 2009 |
Member |
|
|
Ricardo,
In their recent book, Wikinomics, D. Tapscott and A. Williams write about
peer production:
In its purest form, it is a way of producing goods and services that
relies entirely on self-organizing, egalitarian communities of individuals
who come together voluntarily to produce a shared outcome. In reality,
peer production mixes elements of hierarchy and self-organization i.e.,
the most skilled and experienced members of the community provide
leadership and help integrate contributions from the community.
In many peer production communities, productive activities are voluntary
and nonmonetary (page 67)
The above definition maybe the gold standard of open collaborative
production community, achieved by elite few initiatives (likes of LUNIX
and Apache). At the same time, having over 90% of contributors from IBM
and handful of its business partners is not open peer production (i.e. you
dont have 20 independent contributors) by any standard. After two years,
maybe if another independent vendor (e.g. Borland, Sun, HP, Compuware) or
hundreds of independent developers contributed some code, I would have
agreed with you. But we have very different standard for qualification of
a community as truly open and collaborative.
What motivated me in posting my previous note were the last few weeks
discussions on community outreach (which is nothing more than
promotion). You are even putting community outreach at a higher priority
than fixing bugs (wow!). Genuine open peer production projects dont need
so much recruitment activities.
I totally disagree with your statement that agile processes are not
complete. I personally have been involved in many development projects
which successfully utilized different agile processes. Also I have come
across many developers with the same experience. Eclipse (the platform
that you are based on) was developed using Eclipse Way. This is a
methodology developed by practitioners in the trenches out of real needs
and is battle tested, Rather than, conceived by generals faraway from the
combat zone. There are many other examples of amazing methodologies
developed and utilized by development teams. We dont need another new
methodology create by high-priests of methodology to prove their divinity
(or for content of their next book).
I find your statement that, EPF is addressing the mixing and matching
different agile methodologies, arrogant. There are plenty of more
qualified grassroots agile communities who can address this if there is
truly a need for it. The idea of using an arcane meta-model created by an
out-dated organization (OMG) to mix and match agile methodologies is
bizarre.
EPF looks to me like a creative marketing strategy by IBM (possibly
version 2.0 of which is called Commercial Open Source), not a genuine
peer production community. There is nothing wrong with that. But lets not
fool ourselves. Also such miss representation raises some ethical concerns.
Finally, my request to you (as IBM representative) is to open source RUP
and make it available in a wiki form. I bet my Playstation 3 that, it
would be more useful to the community than all of the last 2 years
effort. Let self-organizing communities form around this base content.
Each community can then trim and mash up this content in ways we never
imagined.
The ability to continue to produce art without permission from the
latter-day aristocracy of creativity is central to both cultural and
economic progress
. So much of what makes a free society and economy
healthy and vibrant is that we have limited the control points in a way
that permits creation and experimentation in a largely anarchistic
fashion.
(Wikinomics, page 141)
Tony
|
|
|
Re: OpenUP solution looking for a problem [message #582556 is a reply to message #39210] |
Tue, 04 September 2007 20:31 |
Ricardo Balduino Messages: 191 Registered: July 2009 |
Senior Member |
|
|
I just thought I'd write a few paragraphs to comment on some of your
assertions that might be of interest to the subscribers of this newsgroup.
It's an interesting book the one you referenced. The definition you pointed
out in page 67 of this book describes the way open source projects work,
which should as well include the Eclipse and the EPF projects to the list.
You certainly know that the EPF project consists of various components, like
the EPF Composer tool, and processes like OpenUP , Scrum, XP, DSDM and
others coming up soon that are captured using the EPF framework. All these
efforts under the EPF project have participation from volunteers from
various organizations, thus the 90% number of IBMers you mention does not
correspond to the reality. Please refer to the list of committers and
contributors to the project: http://www.eclipse.org/epf/team/team.php
Perhaps you misunderstood the statement about agile processes completeness
First of all, those are not my words. My comment refers to a current
discussion in a agile forum that talks about gaps some of these agile
processes have and how to fill in the gaps. See this particular posting:
http://tech.groups.yahoo.com/group/leanagilescrum/message/28 5
Second, the fact that some processes have gaps is not necessarily a bad
thing, nor one that states that a process has gaps is necessarily
criticizing the quality of that process. Some processes such as Scrum -
object of the discussion on the list above - provides guidance on project
and requirements management, so engineering practices complement what Scrum
intentionally does not provide.
It is part of EPF project vision to be a framework that helps in capturing
combining and sharing software development practices. This vision has been
supported by multiple individuals from various organizations. As an
exemplary process, OpenUP borrows (with recognition) and combines practices
from all these processes you mentioned as genuinely being developed from the
trenches, so there's nothing of arrogant or written-from-the-ivory-tower in
that perspective, as you would assume.
Your comment about community outreach is inaccurate. In that context, we
were not referring to recruitment (now, if more people desire to join the
project, they may do so at any time as usual, and the EPF team will welcome
their participation ;-)).
The context in the discussion was OpenUP, and how to address the current
Bugzilla items that were postponed for after release 1.0. The alternative to
that would be to invest the volunteers' time in creating collaterals on
OpenUP usage and deployment. As part of the discussion, we understood it
would be acceptable - after the appropriate bug triage was done - that
development of collaterals would take precedence over the remaining Bugzilla
items for OpenUP component, but again, not before appropriate bug triage.
The other components (tool and other processes) follow their own priorities
set in Bugzilla.
Ricardo Balduino.
"Tony" <tony.tatil@gmail.com> wrote in message
news:59a4e2b29c427fdafd337e1f4f734aa9$1@www.eclipse.org...
> Ricardo,
>
> In their recent book, Wikinomics, D. Tapscott and A. Williams write about
> peer production:
>
>
|
|
| |
Re: OpenUP solution looking for a problem [message #583036 is a reply to message #39681] |
Fri, 07 September 2007 20:30 |
Ricardo Balduino Messages: 191 Registered: July 2009 |
Senior Member |
|
|
Tony, thanks for your comments.
Back in time two years ago, IBM contributed a portion of RUP known as RUP
for Small Projects and the embryo of what today we know as OpenUP (called
BUP at that time). OpenUP borrows from this RUP content, plus other agile
methods and from the practical software development experience of those who
have been contributing to the project thus far.
You have a good point about using Wiki to present process content. We are
currently making use of this technology to share and collect feedback on the
processes under the EPF project. You can see it on http://www.epfwiki.net/
where you can find a handful of processes we publish using EPF Composer. As
far as I can tell, it's not a conventional Wiki based on "unstructured
content", so to speak, because this Wiki "has the knowledge" of the
meta-model relationships that method elements need to have. The EPF Wiki has
been extensively used by the EPF community. Just to mention one example, the
Portuguese translation of OpenUP has provided around 120 pages of translated
content in a few months since it started.
Regards,
Ricardo Balduino.
"Tony" <tony.tatil@gmail.com> wrote in message
news:67fb28a9bd18e3c049f07c88d7c37154$1@www.eclipse.org...
> Ricardo,
>
> Firstly I would like to tank you for your professional responses. It
|
|
| | | | | | | | |
Re: OpenUP solution looking for a problem [message #583836 is a reply to message #41832] |
Mon, 01 October 2007 14:38 |
Xavier Méhaut Messages: 133 Registered: July 2009 |
Senior Member |
|
|
I don't want to start a method war :) Certainly EclipseWay may/can be
used for any other development than developping eclipse plugin, like
actually any other method. I just wanted to point out that EclipseWay
has been developped initially for delivering on time and in a
collaborative manner Eclipse artefacts. The noticeable fact is also that
EclipseWay has been developped/formalised in an iterative an incremental
manner too. Personnaly I like the process which is described and I use
it too.
But there is a point that EclipseWay has not well formalised (but I can
have wrong, correct me if so) is the way on how to communicate with
Customers , especially in specification phase (inception), because
Eclipse itself has not yet traditionnal customers like bank, insurances,
Aerospace companies, ... There is another point out of the scope is how
to use formalisms like UML and so on... to modelize your application. Of
course we don"t need UML to well develop an application, neither OMT,
Hood, or other formalisms. But there are cases where the customer wants
documentation with models inside,where the application is a large scale
application, or needs safety, compliance with international norms like
in aerospace, and so on... And then we need to offer a methodology which
encounter their needs. We must also recall in mind that agile modelling
needs a great agreement betwwen every parts od the project to well work...
It is why I say that OpenUP can fill in the bridge beetween a
traditional way of developping an application and the benefits we have
experimented with agile processes like EclipseWay or Scrum.
To finish, as for the language domain where there is no "universal"
language for every kind of applications, there is no "universal" method
for every kind of needs. There is only occurences of different kinds of
methods for specific needs.
So I find quite interesting to think to have like a Painter artist a
palette of methods at disposal, with different kinds of concepts, and to
create on demand the right process for the right problem. In this
vision OpenUp is a process among others, and can be valuable for certain
types of developements.
best regards
Xavier
> Xavier,
>
> I disagree with your statement that: Eclipse Way is only used for
> developing Eclipse Plugins. In one of presentations on Jazz, the Jazz
> team clearly stated that Jazz was developed using Eclipse Way (and Jazz
> is much more than a plugin). To them, Eclipse Way is a methodology that
> is appropriate for wide variety of projects.
> On your last point, I am not convinced. Can you expand on what is the
> problem OprenUp is trying solve? So far what I read from IBM (Per�s
> team) on the subject are mostly marketing content.
> Tony
>
>
|
|
| |
Re: OpenUP solution looking for a problem [message #583858 is a reply to message #41893] |
Mon, 01 October 2007 15:20 |
Xavier Méhaut Messages: 133 Registered: July 2009 |
Senior Member |
|
|
You haven't really read what I have written I think :-(
Furthermore, I'm not from IBM, and what IBM has in mind is not my
purpose. Up to now, we have 4 methods which are modelized with EPFC (and
I hope there will be more in future, and especially EclipseWay), and the
one I found (with the two others I knew - scrum and xp) is OpenUp. And
personnaly I find that this method has an interest for the reasons I've
told you in the previous mail.
Each day there are new methods which are developped. But here we have
one which is modelized with a great tool, and I find it is a good point
for this one too.
best regards
xavier
> Xavier,
>
> Keep in mind, that processes are meant to be tailored before use. So in
> many instances you take an agile, Unified Process (doesn�t have to be
> RUP) or other methodology and tailor it for your needs. For example, you
> can tailor a process to better support customer intimacy. You don�t take
> years and develop a new methodology (reinventing the wheel) when your
> problem can be solved by tailoring an existing process.
>
> When IBM initiated the EPF project they never clearly stated the problem
> that needs solving. Is it to fill the gap between RUP and Agile
> methodology?
>
> Tony
>
>
>
>
>
|
|
|
Re: OpenUP solution looking for a problem [message #583874 is a reply to message #41893] |
Mon, 01 October 2007 15:30 |
Xavier Méhaut Messages: 133 Registered: July 2009 |
Senior Member |
|
|
Tony,
To complete my previous posts, I would like to make notice that the
"subjective" part in choosing a method is very important, especially
when we have to convince our management or the customers that we should
use this method instead of another one.
I like Scrum, XP, and so on , and use them when I can.
But the notions, the vocabulary, the philosophy , the learning curve for
teams which don't know them is quite important and sometine not
"sellable" to the hierarchy.
OpenUP has this advantage to make up an agile method with traditional
concepts and vocabulary... It is also important! It is the same way why
Java is more used than Smalltalk... People had impression that it was
simpler because the vocabulary was the one of C or C++ ... but for the
advanced developpers, Smalltalk was still the best language :)
regards
xavier
> Xavier,
>
> Keep in mind, that processes are meant to be tailored before use. So in
> many instances you take an agile, Unified Process (doesn�t have to be
> RUP) or other methodology and tailor it for your needs. For example, you
> can tailor a process to better support customer intimacy. You don�t take
> years and develop a new methodology (reinventing the wheel) when your
> problem can be solved by tailoring an existing process.
>
> When IBM initiated the EPF project they never clearly stated the problem
> that needs solving. Is it to fill the gap between RUP and Agile
> methodology?
>
> Tony
>
>
>
>
>
|
|
|
Re: OpenUP solution looking for a problem [message #583887 is a reply to message #41955] |
Mon, 01 October 2007 17:53 |
Eclipse User |
|
|
|
Originally posted by: kamal.osellus.com
This is a multi-part message in MIME format.
------=_NextPart_000_0059_01C80432.666BB5F0
Content-Type: text/plain;
charset="UTF-8"
Content-Transfer-Encoding: quoted-printable
Xavier, Tony and others
=20
I see this as a great discussion in trying to identify the root problem =
domain that can be solved with something like OpenUP. Moreover by =
keeping this discussion on a technical investigative level we can make =
good progress.
=20
Xavier, I am having some trouble getting clarity on the problems you =
mentioned that make OpenUP attractive to you. Does the problem you are =
trying to solve relate to the management not being open in trying =
existing methods such as XP, Scrum etc because they don=E2=80=99t =
understand them?=20
=20
Perhaps if you share your experience in terms of a before-after =
situation it can throw some more light on this discussion and bring out =
the real value of introducing OpenUP. After going through this thread I =
have tried to summarize a few scenarios below. The first two points =
relate to OpenUP (the methodology) the third to EPFC (the tool):
1. Before being aware of OpenUP management did not try the existing =
agile-family methodologies because they did not understand them due to =
unfamiliar terms. After you showed them OpenUP they are willing to try =
it in projects because they are able to understand the methodology.=20
2. Before using OpenUP management is unconvinced that any of the =
existing agile-family methodologies will result in projects that are =
delivered on time and within budget. After using OpenUP in projects they =
see (with real examples) projects are delivered on time and within =
budget.
3. Before OpenUP you did not have a way to mix and match the applicable =
parts/modules of any available agile (or other) methodology to customize =
it for your project eco-system. After using EPFC you have a way to blend =
and tailor one or more methodologies.
=20
I am hoping if you can indicate which (one or more) of these scenarios =
reflect your situation this discussion can go deeper in defining and =
understanding that problem domain.=20
=20
Kamal
=20
--------------------------------------------
Kamal Ahluwalia
Lead Solution Specialist
Osellus Inc.
750-144 Front Street West
Toronto, Ontario M5J 3L7 CANADA
Tel: +1 (416) 603-6667 Ext 5063
Email: kamal@osellus.com
Blog: http://www.osellus.com/blogs/author/kamal
----------------------------------------------
=20
=20
"XavierMehaut" <xavier.mehaut@free.fr> wrote in message =
news:fdr3sn$rd4$1@build.eclipse.org...
Tony,
To complete my previous posts, I would like to make notice that the=20
"subjective" part in choosing a method is very important, especially=20
when we have to convince our management or the customers that we =
should=20
use this method instead of another one.
I like Scrum, XP, and so on , and use them when I can.
But the notions, the vocabulary, the philosophy , the learning curve =
for=20
teams which don't know them is quite important and sometine not=20
"sellable" to the hierarchy.
OpenUP has this advantage to make up an agile method with traditional=20
concepts and vocabulary... It is also important! It is the same way =
why=20
Java is more used than Smalltalk... People had impression that it was=20
simpler because the vocabulary was the one of C or C++ ... but for the =
advanced developpers, Smalltalk was still the best language :)
regards
xavier
> Xavier,
>=20
> Keep in mind, that processes are meant to be tailored before use. So =
in=20
> many instances you take an agile, Unified Process (doesn=EF=BF=BDt =
have to be=20
> RUP) or other methodology and tailor it for your needs. For example, =
you=20
> can tailor a process to better support customer intimacy. You =
don=EF=BF=BDt take=20
> years and develop a new methodology (reinventing the wheel) when =
your=20
> problem can be solved by tailoring an existing process.
>=20
> When IBM initiated the EPF project they never clearly stated the =
problem=20
> that needs solving. Is it to fill the gap between RUP and Agile=20
> methodology?
>=20
> Tony
>=20
>=20
>=20
>=20
>
------=_NextPart_000_0059_01C80432.666BB5F0
Content-Type: text/html;
charset="UTF-8"
Content-Transfer-Encoding: quoted-printable
=EF=BB=BF<!DOCTYPE HTML PUBLIC "-//W3C//DTD HTML 4.0 Transitional//EN">
<HTML><HEAD>
<META http-equiv=3DContent-Type content=3D"text/html; charset=3Dutf-8">
<META content=3D"MSHTML 6.00.6000.16527" name=3DGENERATOR>
<STYLE></STYLE>
</HEAD>
<BODY bgColor=3D#ffffff>
<DIV><FONT face=3DArial size=3D2>
<P class=3DMsoNormal style=3D"MARGIN: 0cm 0cm 0pt">Xavier, Tony and=20
others<?xml:namespace prefix =3D o ns =3D =
"urn:schemas-microsoft-com:office:office"=20
/><o:p></o:p></P>
<P class=3DMsoNormal style=3D"MARGIN: 0cm 0cm 0pt"><o:p> </o:p></P>
<P class=3DMsoNormal style=3D"MARGIN: 0cm 0cm 0pt">I see this as a great =
discussion=20
in trying to identify the root problem domain that can be solved with =
something=20
like OpenUP. Moreover by keeping this discussion on a technical =
investigative=20
level we can make good progress.<o:p></o:p></P>
<P class=3DMsoNormal style=3D"MARGIN: 0cm 0cm 0pt"><o:p> </o:p></P>
<P class=3DMsoNormal style=3D"MARGIN: 0cm 0cm 0pt">Xavier, I am having =
some=20
trouble getting clarity on the problems you mentioned that =
make OpenUP=20
attractive to you. Does the problem you are trying to solve relate to =
the=20
management not being open in trying existing methods such as XP, =
Scrum etc=20
because they don=E2=80=99t understand them? <o:p></o:p></P>
<P class=3DMsoNormal style=3D"MARGIN: 0cm 0cm 0pt"><o:p> </o:p></P>
<P class=3DMsoNormal style=3D"MARGIN: 0cm 0cm 0pt">Perhaps if you share =
your=20
experience in terms of a before-after situation it can throw some more =
light on=20
this discussion and bring out the real value of introducing OpenUP. =
After=20
going through this thread I have tried to summarize a few scenarios =
below.=20
The first two points relate to OpenUP (the methodology) the third =
to EPFC=20
(the tool):</P>
<P class=3DMsoNormal style=3D"MARGIN: 0cm 0cm 0pt"><o:p></o:p> </P>
<P class=3DMsoNormal style=3D"MARGIN: 0cm 0cm 0pt"><SPAN=20
style=3D"mso-fareast-font-family: Calibri; mso-bidi-font-family: =
Calibri"><SPAN=20
style=3D"mso-list: Ignore">1.<SPAN style=3D"FONT: 7pt 'Times New =
Roman'"><FONT=20
face=3DArial size=3D2> </FONT></SPAN></SPAN></SPAN>Before being =
aware of OpenUP=20
management did not try the existing agile-family methodologies =
because they=20
did not understand them due to unfamiliar terms. After you showed them =
OpenUP=20
they are willing to try it in projects because they are able to =
understand the=20
methodology. </P>
<P class=3DMsoNormal style=3D"MARGIN: 0cm 0cm 0pt"><SPAN=20
style=3D"mso-fareast-font-family: Calibri; mso-bidi-font-family: =
Calibri"><SPAN=20
style=3D"mso-list: Ignore">2. </SPAN></SPAN>Before using OpenUP =
management is=20
unconvinced that any of the existing agile-family methodologies will =
result in=20
projects that are delivered on time and within budget. After using =
OpenUP in=20
projects they see (with real examples) projects are delivered on time =
and within=20
budget.</P>
<P class=3DMsoNormal style=3D"MARGIN: 0cm 0cm 0pt"><SPAN=20
style=3D"mso-fareast-font-family: Calibri; mso-bidi-font-family: =
Calibri"><SPAN=20
style=3D"mso-list: Ignore">3. </SPAN></SPAN>Before OpenUP you did not =
have a way=20
to mix and match the applicable parts/modules of any available =
agile (or=20
other) methodology to customize it for your project eco-system. After =
using EPFC=20
you have a way to blend and tailor one or more =
methodologies.<o:p></o:p></P>
<P class=3DMsoNormal style=3D"MARGIN: 0cm 0cm 0pt"><o:p> </o:p></P>
<P class=3DMsoNormal style=3D"MARGIN: 0cm 0cm 0pt">I am hoping if you =
can indicate=20
which (one or more) of these scenarios reflect your situation this =
discussion=20
can go deeper in defining and understanding that problem domain. =
<o:p></o:p></P>
<P class=3DMsoNormal style=3D"MARGIN: 0cm 0cm 0pt"><o:p> </o:p></P>
<P class=3DMsoNormal style=3D"MARGIN: 0cm 0cm 0pt">Kamal<o:p></o:p></P>
<P class=3DMsoNormal style=3D"MARGIN: 0cm 0cm 0pt"><o:p> </o:p></P>
<P class=3DMsoNormal style=3D"MARGIN: 0cm 0cm 0pt"><SPAN lang=3DEN-US=20
style=3D"FONT-SIZE: 10pt; COLOR: navy; FONT-FAMILY: =
'Arial','sans-serif'; mso-ansi-language: =
EN-US">--------------------------------------------</SPAN> <SPAN=20
lang=3DEN-US style=3D"mso-ansi-language: EN-US"><o:p></o:p></SPAN></P>
<P class=3DMsoNormal style=3D"MARGIN: 0cm 0cm 0pt"><SPAN lang=3DEN-US=20
style=3D"FONT-SIZE: 10pt; COLOR: black; FONT-FAMILY: =
'Arial','sans-serif'; mso-ansi-language: EN-US">Kamal=20
Ahluwalia</SPAN><SPAN lang=3DEN-US=20
style=3D"mso-ansi-language: EN-US"><o:p></o:p></SPAN></P>
<P class=3DMsoNormal style=3D"MARGIN: 0cm 0cm 0pt"><SPAN lang=3DEN-US=20
style=3D"FONT-SIZE: 10pt; COLOR: black; FONT-FAMILY: =
'Arial','sans-serif'; mso-ansi-language: EN-US">Lead=20
Solution Specialist<o:p></o:p></SPAN></P>
<P class=3DMsoNormal style=3D"MARGIN: 0cm 0cm 0pt"><SPAN lang=3DEN-US=20
style=3D"FONT-SIZE: 10pt; COLOR: black; FONT-FAMILY: =
'Arial','sans-serif'; mso-ansi-language: EN-US">Osellus=20
Inc.<o:p></o:p></SPAN></P>
<P class=3DMsoNormal style=3D"MARGIN: 0cm 0cm 0pt"><SPAN lang=3DEN-US=20
style=3D"FONT-SIZE: 10pt; FONT-FAMILY: 'Arial','sans-serif'; =
mso-ansi-language: EN-US">750-144=20
Front Street West</SPAN><SPAN lang=3DEN-US=20
style=3D"FONT-SIZE: 12pt; FONT-FAMILY: 'Times New Roman','serif'; =
mso-ansi-language: EN-US"><o:p></o:p></SPAN></P>
<P class=3DMsoNormal style=3D"MARGIN: 0cm 0cm 0pt"><SPAN lang=3DEN-US=20
style=3D"FONT-SIZE: 10pt; FONT-FAMILY: 'Arial','sans-serif'; =
mso-ansi-language: EN-US">Toronto,=20
Ontario M5J 3L7 CANADA</SPAN><SPAN lang=3DEN-US=20
style=3D"mso-ansi-language: EN-US"><o:p></o:p></SPAN></P>
<P class=3DMsoNormal style=3D"MARGIN: 0cm 0cm 0pt"><SPAN lang=3DEN-US=20
style=3D"FONT-SIZE: 10pt; COLOR: black; FONT-FAMILY: =
'Arial','sans-serif'; mso-ansi-language: EN-US">Tel:=20
+1 (416) 603-6667 Ext 5063</SPAN><SPAN lang=3DEN-US=20
style=3D"mso-ansi-language: EN-US"><o:p></o:p></SPAN></P>
<P class=3DMsoNormal style=3D"MARGIN: 0cm 0cm 0pt"><SPAN lang=3DEN-US=20
style=3D"FONT-SIZE: 10pt; FONT-FAMILY: 'Arial','sans-serif'; =
mso-ansi-language: EN-US">Email:=20
<A =
title=3D"blocked::mailto:kamal@osellus.com mailto:kamal@osellus.com" =
href=3D"mailto:kamal@osellus.com"><FONT=20
color=3D#0000ff>kamal@osellus.com</FONT></A></SPAN><SPAN lang=3DEN-US=20
style=3D"mso-ansi-language: EN-US"><o:p></o:p></SPAN></P>
<P class=3DMsoNormal style=3D"MARGIN: 0cm 0cm 0pt"><SPAN lang=3DEN-US=20
style=3D"FONT-SIZE: 10pt; FONT-FAMILY: 'Arial','sans-serif'; =
mso-ansi-language: EN-US">Blog:=20
<A title=3Dblocked::http://www.osellus.com/blogs/author/kamal=20
href=3D"http://www.osellus.com/blogs/author/kamal">http://www.osellus.com=
/blogs/author/kamal</A><o:p></o:p></SPAN></P>
<P class=3DMsoNormal style=3D"MARGIN: 0cm 0cm 0pt"><SPAN lang=3DEN-US=20
style=3D"FONT-SIZE: 10pt; COLOR: navy; FONT-FAMILY: =
'Arial','sans-serif'; mso-ansi-language: =
EN-US">----------------------------------------------</SPAN ><SPAN=20
lang=3DEN-US=20
style=3D"FONT-SIZE: 12pt; FONT-FAMILY: 'Times New Roman','serif'; =
mso-ansi-language: EN-US"><o:p></o:p></SPAN></P>
<P class=3DMsoNormal style=3D"MARGIN: 0cm 0cm 0pt"><o:p><FONT =
face=3DCalibri=20
size=3D3> </FONT></o:p></P>
<P class=3DMsoNormal style=3D"MARGIN: 0cm 0cm 0pt"><SPAN=20
style=3D"mso-bidi-font-family: 'Times New Roman'; mso-ascii-font-family: =
Calibri; mso-ascii-theme-font: minor-latin; mso-hansi-font-family: =
Calibri; mso-hansi-theme-font: minor-latin; mso-bidi-theme-font: =
minor-bidi"><o:p><FONT=20
face=3DCalibri size=3D3> </FONT></o:p></SPAN></P></FONT></DIV >
<BLOCKQUOTE=20
style=3D"PADDING-RIGHT: 0px; PADDING-LEFT: 5px; MARGIN-LEFT: 5px; =
BORDER-LEFT: #000000 2px solid; MARGIN-RIGHT: 0px">
<DIV>"XavierMehaut" <<A=20
href=3D"mailto:xavier.mehaut@free.fr">xavier.mehaut@free.fr</A>> =
wrote in=20
message <A=20
=
href=3D"news:fdr3sn$rd4$1@build.eclipse.org">news:fdr3sn$rd4$1@build.ecli=
pse.org</A>...</DIV>Tony,<BR>To=20
complete my previous posts, I would like to make notice that the=20
<BR>"subjective" part in choosing a method is very important, =
especially=20
<BR>when we have to convince our management or the customers that we =
should=20
<BR>use this method instead of another one.<BR>I like Scrum, XP, and =
so on ,=20
and use them when I can.<BR>But the notions, the vocabulary, the =
philosophy , the learning curve for <BR>teams which don't know them is =
quite=20
important and sometine not <BR>"sellable" to the hierarchy.<BR>OpenUP =
has this=20
advantage to make up an agile method with traditional <BR>concepts and =
vocabulary... It is also important! It is the same way why <BR>Java is =
more=20
used than Smalltalk... People had impression that it was <BR>simpler =
because=20
the vocabulary was the one of C or C++ ... but for the <BR>advanced=20
developpers, Smalltalk was still the best language=20
:)<BR><BR>regards<BR>xavier<BR><BR>> Xavier,<BR>> <BR>> Keep =
in mind,=20
that processes are meant to be tailored before use. So in <BR>> =
many=20
instances you take an agile, Unified Process (doesn=EF=BF=BDt have to =
be <BR>> RUP)=20
or other methodology and tailor it for your needs. For example, you =
<BR>>=20
can tailor a process to better support customer intimacy. You =
don=EF=BF=BDt take=20
<BR>> years and develop a new methodology (reinventing the wheel) =
when your=20
<BR>> problem can be solved by tailoring an existing =
process.<BR>>=20
<BR>> When IBM initiated the EPF project they never clearly stated =
the=20
problem <BR>> that needs solving. Is it to fill the gap between RUP =
and=20
Agile <BR>> methodology?<BR>> <BR>> Tony<BR>> <BR>> =
<BR>>=20
<BR>> <BR>></BLOCKQUOTE></BODY></HTML>
------=_NextPart_000_0059_01C80432.666BB5F0--
|
|
|
Re: OpenUP solution looking for a problem [message #583900 is a reply to message #41557] |
Mon, 01 October 2007 19:05 |
Peter Haumer Messages: 228 Registered: July 2009 |
Senior Member |
|
|
Hello.
Per and I worked very closely with the Jazz team (in particular Erich Gamma
and Kai-Uwe Maetzel) on formalizing Eclipse Way for EPF Composer as well as
defining integration points with Jazz for content as well as tools (e.g. the
ones we previewed at the RSDC in Florida this year). During this
collaboration a lot of feedback went both ways. However, principally
Eclipse Way and OpenUP have two different goals. Eclipse Way aims at
'precisely' describing how the Eclipse teams work. OpenUP is a generic
process framework that provides an essential and consistent collection of
best practices for development teams to use as a starting point to develop
and tailor their own processes. You can use OpenUP as is, but probably most
teams want to tailor it to their needs. Eclipse Way aims at describing only
the practices that are used in the development of Eclipse instead. In that
sense it is already the tailored 'end product', the process in use by the
Eclipse teams. You can see the clear influence the Eclipse practices had on
OpenUP looking at work items, micro iterations, stable builds, builds,
iteration reviews and feedback etc. The other way round the Eclipse and
Jazz teams started looking at Risk management after reviewing OpenUP with us
as well as Jazz' review workflows are close to RUP's compliance content. In
respect to AJ's comment it is correct that the EMF or GMF teams would not
say that they use OpenUP, but certainly OpenUP represents a lot of their
practices. Hence, for people interested in learning about such practices
OpenUP is a good way to start looking plus EPF Composer allows tailoring and
enriching that content as well.
Peter.
"Tony" <tony.tatil@gmail.com> wrote in message
news:61813296b370f62fcb59821f579dd20e$1@www.eclipse.org...
> In the most recent Rational Edge artical, Per says "The work was
> contributed to Eclipse under the Eclipse Process Framework (EPF) 5
> project, and we continued co-developing the process with roughly twenty
> people from a dozen companies. Some of us had done agile implementations
> of RUP, others were on the Dynamic System Development Method (DSDM) 6
> board of advisors or had created Agile Model Driven Development (AMDD); 7
> others practiced Scrum, 8 and some were key people behind the Eclipse
> Way...".
> I wonder who are the key people behind the Eclipse Way that made a major
> contibution to OpenUp.
>
> This to me is a good example of pure marketing (not enablement). Read and
> decide for yourself.
>
> http://www.ibm.com/developerworks/rational/library/sep07/kro ll/index.html
>
>
> Tony
>
|
|
|
Re: OpenUP solution looking for a problem [message #583925 is a reply to message #41985] |
Mon, 01 October 2007 19:22 |
Xavier Méhaut Messages: 133 Registered: July 2009 |
Senior Member |
|
|
Hi Kamal,
I see that this thread goes further than I wanted to go :) and my
different answers were not a general answer built with a specific intent
but were mere brought by the flow of the discussion.
I repeat myself in other words, I'm not an evangelist of OpenUP and I
just try to give counter-arguments to Tony to show that OpenUp is not a
bad method because it is made by people from IBM/Rational, and because
some roots are from RUP.
I'm method agnostic, because I used to use many ones from SART, SADT,
SA, Schlaer and Mellor, Merise, Hood, Hoora, AonixWay, and so on... With
time, we relativise the use of methods, and we only know that we need
some to develop apllications. If we look closer to all thoses methods,
we can see that there are finally quite few concepts to handle. And the
interesting part of a methodologist work is to arrange these few
concepts to make a process the most efficient as possible in the context
in which we are.
So the arguments I gave for OpenUP are some arguments among others to
show that a method must be tailored as told by Tony to match the
customer and development team needs. I could also add as arguments for
OpenUp the fact that the method insists on the architecture, which is a
very important point in the case of a quite big project.
In conclusion, OpenUp is :
- very young
- not yet tooled
- not yet implemented in public projects
- not a rupture with other methods
but
- exists,
- has a team for working on it
- can be enhanced, improved, derived thnaks to EPF (like Scrum or XP or
DSDM up to now)
- is clear
- has no visible bottleneck for the moment (or I haven't yet seen them)
- is "sellable" to a hierarchy
It is enough to give a chance to this method, isn't it?
Best regards
Xavier
>
>
> Xavier, Tony and others
>
>
>
> I see this as a great discussion in trying to identify the root problem
> domain that can be solved with something like OpenUP. Moreover by
> keeping this discussion on a technical investigative level we can make
> good progress.
>
>
>
> Xavier, I am having some trouble getting clarity on the problems you
> mentioned that make OpenUP attractive to you. Does the problem you are
> trying to solve relate to the management not being open in trying
> existing methods such as XP, Scrum etc because they don’t understand them?
>
>
>
> Perhaps if you share your experience in terms of a before-after
> situation it can throw some more light on this discussion and bring out
> the real value of introducing OpenUP. After going through this thread I
> have tried to summarize a few scenarios below. The first two points
> relate to OpenUP (the methodology) the third to EPFC (the tool):
>
>
>
> 1. Before being aware of OpenUP management did not try the existing
> agile-family methodologies because they did not understand them due to
> unfamiliar terms. After you showed them OpenUP they are willing to try
> it in projects because they are able to understand the methodology.
>
> 2. Before using OpenUP management is unconvinced that any of the
> existing agile-family methodologies will result in projects that are
> delivered on time and within budget. After using OpenUP in projects they
> see (with real examples) projects are delivered on time and within budget.
>
> 3. Before OpenUP you did not have a way to mix and match the applicable
> parts/modules of any available agile (or other) methodology to customize
> it for your project eco-system. After using EPFC you have a way to blend
> and tailor one or more methodologies.
>
>
>
> I am hoping if you can indicate which (one or more) of these scenarios
> reflect your situation this discussion can go deeper in defining and
> understanding that problem domain.
>
>
>
> Kamal
>
>
>
> --------------------------------------------
>
> Kamal Ahluwalia
>
> Lead Solution Specialist
>
> Osellus Inc.
>
> 750-144 Front Street West
>
> Toronto, Ontario M5J 3L7 CANADA
>
> Tel: +1 (416) 603-6667 Ext 5063
>
> Email: kamal@osellus.com <mailto:kamal@osellus.com>
>
> Blog: http://www.osellus.com/blogs/author/kamal
>
> ----------------------------------------------
>
>
>
>
>
> "XavierMehaut" <xavier.mehaut@free.fr
> <mailto:xavier.mehaut@free.fr>> wrote in message
> news:fdr3sn$rd4$1@build.eclipse.org...
> Tony,
> To complete my previous posts, I would like to make notice that the
> "subjective" part in choosing a method is very important, especially
> when we have to convince our management or the customers that we should
> use this method instead of another one.
> I like Scrum, XP, and so on , and use them when I can.
> But the notions, the vocabulary, the philosophy , the learning curve
> for
> teams which don't know them is quite important and sometine not
> "sellable" to the hierarchy.
> OpenUP has this advantage to make up an agile method with traditional
> concepts and vocabulary... It is also important! It is the same way why
> Java is more used than Smalltalk... People had impression that it was
> simpler because the vocabulary was the one of C or C++ ... but for the
> advanced developpers, Smalltalk was still the best language :)
>
> regards
> xavier
>
> > Xavier,
> >
> > Keep in mind, that processes are meant to be tailored before use.
> So in
> > many instances you take an agile, Unified Process (doesn�t have
> to be
> > RUP) or other methodology and tailor it for your needs. For
> example, you
> > can tailor a process to better support customer intimacy. You
> don�t take
> > years and develop a new methodology (reinventing the wheel) when
> your
> > problem can be solved by tailoring an existing process.
> >
> > When IBM initiated the EPF project they never clearly stated the
> problem
> > that needs solving. Is it to fill the gap between RUP and Agile
> > methodology?
> >
> > Tony
> >
> >
> >
> >
> >
|
|
| | | | |
Re: OpenUP solution looking for a problem [message #584127 is a reply to message #42312] |
Tue, 02 October 2007 18:49 |
Per Kroll Messages: 60 Registered: July 2009 |
Member |
|
|
Tony,
The intent with the sentence was to acknowledge a diverse background of
people, and influences from many sources. It is important to acknolwedge
that some content from Eclipse Way is in OpenUP. Specifically, the iteration
lifecycle, which is key to how to effectively manage an iteration so you end
up delivering a useful executable of high quality is from Eclipse Way, see
Figure 2 in
http://www.ibm.com/developerworks/rational/library/sep07/kro ll/index.html
The statement "some were key people behind the Eclipse Way" is however
incorrect. We did work with some key people behind Eclipse Way, which
resulted in content making its way into OpenUP, but they did not themself
contribute directly to OpenUP. The rest of the sentence is correct. "Some of
us had done agile implementations of RUP, others were on the Dynamic System
Development Method (DSDM) 6 board of advisors or had created Agile Model
Driven Development (AMDD); 7 others practiced Scrum, 8 .... "
So, thanks for finding the mistake, I will have it corrected.
Cheers
/Per
"Tony" <tony.tatil@gmail.com> wrote in message
news:05877041f556bfc34d1c48c9cc13ba94$1@www.eclipse.org...
> Hi Peter,
>
> What I am getting at is an extremely serious matter in both academic and
> corporate worlds, which is knowingly giving the public false information.
> The statement
|
|
| | | | |
Re: OpenUP solution looking for a problem [message #584215 is a reply to message #42558] |
Thu, 04 October 2007 17:51 |
AJ Bencomo Messages: 42 Registered: July 2009 |
Member |
|
|
Tony,
I am not sure what you are actually trying to say with "It's extremely
alarming that IBM has been the only contributor to EPF Composer". It
seems to me that you are trying to be either extremely
yellow/controversial or perhaps negative. So, I personally have to
disagree with you on this matter.
EPF is not the only tool/project supported by only IBMers. For instance,
the EMF project is remarkably well sustained by only 5 IBM developers, but
it has been embarrassed by multiple individuals and organizations.
It is obvious that you have a misconception/misunderstanding of why
projects/tools/code are donated to the Eclipse Open Source community.
The success of an open source project (at least in the Eclipse community)
is not determined or measured by the number of developer committing code.
Instead, it is determined by the benefits that it brings to the open
source community. So, I think EPF has clearly accomplish that.
Therefore, when you call it a "crony open source", basically you are
saying that the eclipse SDK is also crony because who do you think is
doing most the code development? Microsoft?
Ciao,
-Alfredo
BTW, I don't work for IBM ;-)
|
|
|
Re: OpenUP solution looking for a problem [message #584250 is a reply to message #42619] |
Fri, 05 October 2007 03:23 |
Tony Messages: 52 Registered: July 2009 |
Member |
|
|
Alfredo,
You are right, I need to clarify some of my previous points. Actually I
am very glad that you responded intelligently to my note. Intelligent,
in-depth and open discussions are the best way to learn.
As you know IBM, like many other IT vendors, is a public for-profit
company. Therefore, its first obligation is maximization of shareholder
value. Since, IBM has very good governance, all its business units have to
make sure their projects are aligned with the overall business objectives.
So, its realistic that next quarter a decision is made to move the
resources assigned to EPF to another more critical or lucrative project.
Many of us who are, or have been, working for large companies know this is
very possible.
At such a point if there is no IBM involvement, who is going to answer
your technical questions, who is going to fix bugs and who is going
enhance the tool? There is nobody else right now beside IBM employees who
get paid to do this work. You have spent the time learning how to use a
complex tool (plug-ins, variability, configurations etc.) and entered all
your methods. Now you are stuck with a dead-end tool. This is a major
risk that users are taking today. So, it may be helpful if we acknowledge
this risk (users be aware), or IBM can commit to support EPF for the
next two years.
I dont know if you are aware of the debates on the topics of IBM Eclipse
support and single vendor Eclipse projects at the Eclipse board of
director level. The Eclipse community has been alarmed by IBM focus on
Jazz (a commercial product) and shifting resources from Eclipse, hence
slowing the Eclipse related development. Also in their very latest
meeting the topic of single vendor project was discussed:
http://www.eclipse.org/org/foundation/boardminutes/2007_06_2 0-21_Minutes.php
From the second last paragraph in this page:
Mike Milinkovich introduced the topic of vendor neutrality.
Essentially in some circumstances we are sourcing open source code
originating from a single vendor and that is something we should
specifically consider the implications of. The lack of a definition of
open source, what constitutes a neutral project, and the fact that
this state could change over time was discussed and the associated
problems of managing any such distinction.
.
As far as I am concerned any Eclipse project (not just EPF) that is
dependent on a single vendor has the same problem.
How did you come up with the number of IBMers supporting this effort? Are
there really only 5 developers writing all the Java code for EPF composer?
There are more issues with EPF Composer being labeled as open source which
Ill cover in future posts.
Sorry for the long post,
Tony
BTW its not important to me who you work for. Only what you have to say
is important.
|
|
| |
Re: OpenUP solution looking for a problem [message #584302 is a reply to message #42778] |
Fri, 05 October 2007 16:17 |
Tony Messages: 52 Registered: July 2009 |
Member |
|
|
Ana,
Firstly, it is not my intension to be negative. As I stated multiple
times before, deep discussion helps everyone. We should be open to it.
I disagree with you on innovativeness of EPF Composer. Its a traditional
thick client application in an on-demand web application age. Just look at
what Adobe is doing with its Photoshop Lite. A rich media application like
Photoshop running in a browser. Now thats what I call innovative. I
would get very excited if someone works on web-based version of EPF
Composer. The other problem is that EPF Composer and the meta-model are
unnecessarily complex and not flexible. Some of the OpenUP concepts such
as work-items are not supported by the tool and the meta-model.
Also, it sounds like more and more wiki technology is being used for
OpenUP on going development.
In my opinion, OpenUP, or any methodology for that matter, is not a
suitable Eclipse community project. Keep in mind even Eclipse way is not
an Eclipse project. The best model for such a content-intensive
collaborative development is Wiki. Wikipidia is the best example of a very
successful and universally beneficial content-intensive project. Maybe
the RUP team selected Eclipse because its IBM friendly environment.
Hopefully IBM and other members of this project have open mind and
consider the above suggestions.
Ana, I have read carefully every one of the teams meeting minutes and
have my own idea how OpenUP 1.0 evolved from the initial IBM contribution.
You may remember since you were part of these discussions that IBM and
other members felt that it initial version was not innovative enough
(nothing new to offer) and not marketable so it was changed to look more
interesting (more window dressing). If you want I can provide you with
quotations from these meetings.
BTW, I came up with Crony Open Source from a very popular term Crony
Capitalism which was coined to describe South East Asian regimes. They
are skillfully made to look capitalist from outside but they are far from
it. Their economy is controlled by a single dictator, benefiting his
family member and close associates.
Tony
|
|
|
Re: OpenUP solution looking for a problem [message #584312 is a reply to message #42809] |
Sun, 07 October 2007 18:00 |
Ana Valente Pereira Messages: 18 Registered: July 2009 |
Junior Member |
|
|
Well ... it really sounded really negative to me ... maybe because I am
not a native English speaker and I may not had fully understood the
meaning of words like "crony"
.... also I was not referring to the innovativeness of EPF Composer ...
or OpenUP ... what I consider an innovation is the "EPF project" ...
a path to an open way of defining processes ... I find very interesting
and powerful the idea of having open source applied to processes and not
only to software products... even if this project has not created
something very "innovative" so far I really believe in its potential
Ana
Tony wrote:
> Ana,
>
> Firstly, it is not my intension to be negative. As I stated multiple
> times before, deep discussion helps everyone. We should be open to it.
> I disagree with you on innovativeness of EPF Composer. It�s a
> traditional thick client application in an on-demand web application
> age. Just look at what Adobe is doing with its Photoshop Lite. A rich
> media application like Photoshop running in a browser. Now that�s what
> I call innovative. I would get very excited if someone works on
> web-based version of EPF Composer. The other problem is that EPF
> Composer and the meta-model are unnecessarily complex and not flexible.
> Some of the OpenUP concepts such as work-items are not supported by the
> tool and the meta-model.
> Also, it sounds like more and more wiki technology is being used for
> OpenUP on going development.
>
> In my opinion, OpenUP, or any methodology for that matter, is not a
> suitable Eclipse community project. Keep in mind even Eclipse way is not
> an Eclipse project. The best model for such a content-intensive
> collaborative development is Wiki. Wikipidia is the best example of a
> very successful and universally beneficial content-intensive project.
> Maybe the RUP team selected Eclipse because it�s IBM friendly environment.
> Hopefully IBM and other members of this project have open mind and
> consider the above suggestions.
>
> Ana, I have read carefully every one of the team�s meeting minutes and
> have my own idea how OpenUP 1.0 evolved from the initial IBM
> contribution. You may remember since you were part of these discussions
> that IBM and other members felt that it initial version was not
> innovative enough (nothing new to offer) and not marketable so it was
> changed to look more interesting (more window dressing). If you want I
> can provide you with quotations from these meetings.
> BTW, I came up with �Crony Open Source� from a very popular term �Crony
> Capitalism� which was coined to describe South East Asian regimes. They
> are skillfully made to look capitalist from outside but they are far
> from it. Their economy is controlled by a single dictator, benefiting
> his family member and close associates.
>
> Tony
>
>
>
|
|
|
Re: EclipseWay in EPFC [message #584558 is a reply to message #42493] |
Tue, 09 October 2007 17:54 |
Eclipse User |
|
|
|
Originally posted by: david.osellus.com
Hello Xavier,
We were also interested in using The Eclipse Way process. After coming
across your post, I posted a question on the jazz.net newsgroup regarding
the licensing required to use the process and received a reply stating
that The Eclipse Way is currently only available from jazz.net and is
covered by the terms of the license required for jazz.net so it seems that
they may not be making The Eclipse Way publicly available.
Regards,
Dave
|
|
| | |
Goto Forum:
Current Time: Sun Jun 16 13:02:10 GMT 2024
Powered by FUDForum. Page generated in 0.13411 seconds
|