[News.eclipse.foundation] Re: [NEWCOMER] - Evaluation Summary: What is E

[Jean Couillaud <couillaud@xxxxxxxxxx>]

Ilias,

Although some (I dont want to extend on the exact percentage, it's not 
the point) of your points are right, I respecfully think you are missing 
some major points that I will try to explain.

First, the eclipse.org foundation is not a charity business. Either you 
accept the way it seems to work or you pay the fee to have a vote in 
important decisions or you try to be heard by the right people in a way 
that will not seem too rude (It is of no use to shout at directors 
directly). You can not expect it (the foundation) to bow before you for 
your newsposts and promote you in one week _because you seem to know 
what you are talking about_.

Second, opening so much threads (interesting or not, again, that is not 
my point) and waiting so little time for a detailed answer and a 
reflexion leading to some interesting change, is quite surprising and 
makes me question the fact that you already did such work in a 
professionnal environment. If you really worked for 20 years in the IT 
world, you should now that everyone has a pride and should be granted a 
minimum amount of time to accept any failure.

Third, as convinced as you might be of the accuracy of your comments, it 
might not be so obvious to people that matter (Namely: architecture 
choices in eclipse). People spent a long time on it and you might have 
missed some important points that encouraged them to think the 
architecture as it is now. Maybe you could try to make more than mere 
suggestions and try to be part of the project.

Fourth, as Marcus already told it, the situation is not that bad. It is 
very easy to come from nowhere, analyze the situation, criticize it and 
go away because none of your remarks are taken into account quickly 
enough in your opinion. Usually, people behaving like that are 
disregarded, plain and simple. In this case, it would be a shame since 
some of your comments are worth reading.
I completly agree with the fact that there seems to be some problems 
with the eclipse foundation and the EMO (but I will come back on this 
later), but if you really want to help, I think you should stop 
criticizing people because the changes you _demand_ are not done quickly 
enough and try to do it _the open source way_ (as some other eclipse 
members do):
Modify the parts you think should be modified, communicate on your 
efforts during the process to avoid your contribution to be unusable at 
the end, and see if your work is appreciated (you could even sell it at 
some point).
To tell a bit more about the foundation (it is a bit off topic, sorry). 
I have the impression that the foundation has a problem: On one hand, it 
tries to attract open source people, on the other hand, it tries to 
attract commercial entities to back it up financially. It seems to 
desire to work as in an open source project, but it finds it really 
difficult to do it without putting aside its commercial backers.
On one hand, it provides a way for the community to influence 
development, make choices, on the other hand, it seems it tries to 
preserve the influence of its biggest members. I feel the eclipse 
foudation's biggest problem IMHO is about money, which is kind of paradoxal.

Fifth, back on topic, you might be a genius in terms of management, Java 
and Webbased development, and quality assurance, but you can not truly 
expect the foundation to listen to every single person in the community 
who thinks he detains the real knowledge, and apply the changes he 
_demands_ immediatly.

Instead, I trully advise you to stay in the community (it is a trully 
great community and a trully great topic), and help do the necessary 
changes. Sorry for being rude, but it seems to me it is a bit weak, if 
not childish, to post 4-5 comments on this newsgroup in 4-5 days and 
leave the community with so much fuss because no director answered you 
directly to tell you that your comments have been heard and that 
modifications will be done immediatly.

Respectfully,

Jean

ilias wrote:
> [followup to .foundation]
>
> -
>
> Please take 15 min. to read this carefully. This will save you most  
> possibly many time.
>
> -
>
> What is eclipse NOT.
>
> eclipse is _not_ a no-cost production environment
>   - you have to get commercial plugins to make it work
>     - for Webdevelopement
>     - for Model Driven development
>     - for efficient crossplatform GUI design
>     - [...]
>
> eclipse is not complete & mature WYSIWYG environment
>   - Visual GUI editors (VE project) are not mature
>     - Visual GUI editors do not work across platforms
>   - HTML / JSP and orther essential visual editors are not provided
>   - the UML2 project refers to the UML _model_
>     - a visual UML designer is not provided
>
> eclipse is not a real cross-platform solution
>   - Windows NT and others are _not_ officially validated
>     - think twice if you want to validate such huge code-base yourself
>   - Several eclipse projects do not support e.g. Apple Mac.
>
> eclipse is not a "write once run everywhere" solution
>   - the promise is not kept within eclipse
>   - you have possibly to create platform specific GUI code
>     - you have to provide different installations per platform
>
>
> *RCP*
>   - inherits all described cross platform problems
>
>
> *SWT*
>
> eclipse is not an overall elegant and efficient design
>   - SWT (platform specific GUI) does not provide an efficient 
> abstraction  layer
>     - you need to code platform dependent
>       - no framework is provided to encapsulate this task fully
>   - several projects suffer already from this (MAC incompatibility)
>   - SWT is organized as a component
>     - should be at minimum a subproject
>
>
> *COMPATIBILITY*
>
> eclipse JDT (Java Development Tools) are not fully compatible
>   - Eclipse own compiler is not fully compatible to official specs
>     - Feeback on compatibility issues is not treated in a special way
>
> eclipse is not Java 5.0 capable
>   - many constructs of latest java release (1.5/5.0) were unsupported
>     - implementation will delay till 2nd or 3rd quarter of 2005
>   - although requested by community, adequate early support was not  
> provided
>     - depending users and even eclipse.projects suffer now from this
>
>
>
> *GOVERNANCE*
>
> eclipse is _not_ honest to new users
>   - Limitations are not stated clearly on the website
>     - Many time can be lost, until limitations are uncovered
>
> eclipse foundation's board has _no_ user representatives.
>
> eclipse governance is not transparent
>   - Board meetings and contents are not announced
>   - Board meeting _abridged_ minutes are published one month later.
>   - Staffing decisions (e.g. Webmasters) are kept behind the scenes.
>     - no further information given
>   - EMO (Eclipse Management Organization) is _invisible_ in its operation.
>   - critical feedback is many times ignored.
>
> eclipse developement infrastructure is not really open
>   - no process defined for influence of voting on Issues
>   - many Issues are simply ignored.
>   - several procedures happen effectively 'undercover'
>
> eclipse does not follow their governance and developement processes
>   - not all described processes are followed in practice
>     - no issue tracking category available to channelize complaints.
>
>
>
> *WEBTEAM*
>
> - the organizational unit "webteam" does not exist
>   - no issue tracking category available to channelize complaints.
>     - request to create category were simply _ignored_ by webmaster &  
> direction
>
>
>
> *CONTROL*
>
> eclipse has no decentralized control.
> - The Platform Project has very much control.
>   - depending projects have no strictly defined process to influence  
> platform
> - JDT subproject keeps many highly important subsystems.
>   - e.g. Compliler (which is at minimum a subproject)
> - SWT subproject keeps many highly important subsystems.
>   - should be at minimum a subproject
>
>
>
> *EFFICIENCY*
>
> eclipse foundation is not fast and efficient
>   - User feedback is not processed immediately
>     - Even _critical_ issues (like privacy policy) are not processed  
> immediately
>   - most essential tasks (like newsgroup monitoring) are not ensured.
>     - simplest things, like an "office@xxxxxxxxxxx" email are missing.
>
> [...] (some more things)
>
>
> *FINALE*
>
> - I would not be able to work with a tool where the infrastructure is 
> in  effect not working.
> - I've suggested to assist the eclipse foundation practically as an  
> independent, with the goal of a >=5% efficiency increase within 3 
> months  from now.
> - No answer.
>
> -
>
> - I close the eclipse case now quickly, as I cannot deal anymore with 
> such  irrationality and inefficiency (apology! my aggression level has 
> raised  due to several unanswered issues, including privacy policy 
> complaints)
>
> - Moving on to NetBeans.org now, picking it, starting to work.
>
> The feedback/influence/contribution/information channels look much more  
> efficient @netbeans.org!
>
> As a conclusion, the product should have a much higher quality.
>
> -
>
> If there is enouth interest, I will possibly write a more detailed 
> report  about eclipse.
>
> Please contact me via email.
>
> Thank you!
>
> -
> -
> -
>
> Some History / Details:
>
> After an initial compact review of netbeans/eclipse, I  selected 
> eclipse  for a deeper evaluation (based on a 'better feeling' that I had).
>
> The things that give me an positive impression are finally the things 
> that  make eclipse fail (SWT, Openness, Transparency, Independency, ...).
>
> Many times expectations were raised based on marketing stuff, or simply  
> due to the need to find finally the so long awaited solution.
>
> Of course a product cannot always fulfill this expectations in practice.
>
> Same thing for Open Source Projects.
>
> -
>
> You will possibly read some replies on this message, which will state  
> rationales and justifications against my findings.
>
> I can provide for nearly every statement one or more links within  
> eclipse.org.
>
> But I am tired. I apologize for not making a detailed report.
>
> The newsarchive of eclipse.org has a bad quality, but here are the 
> links  to my messages:
>
> http://eclipse.org/search/search.cgi?q=ilias%40lazaridis.com&ul=%2Fnewslists%2F&ps=20&m=all 
>
>
> -
>
> I continue monitoring of eclipse.foundation for a while.
>
> cu!
>
> .