Hello,
please don't take these words personally, i know that some of them may
sound a bit bitter.

On Sat, May 2, 2009 at 7:43 PM, Adriano Marques <[email protected]> wrote:

> By saying that, Luis meant that it will be possible to distribute UMPA
> as a separate package for others to use just as you currently do. The
> *only* difference, is that people using umpa will have to do import
> umit.umpa instead of import umpa. I know someone could say it is
> easier to do import umpa, and I agree with that. But... come on! We
> have auto completion, and the goal here is to make our organization
> grow better and stronger with a sane pattern for all current projects
> and all projects to come.

Saying that user has to add 'umit.' and he can use auto-completion is
not an argument. it's the "solution". more, this solution make some
dependecies to the editor etc. in general, it sound like Microsoft
idea. But lets follow..

>> I don't know how much you understand structure of Apache code, but
>> not. It's not the same. Sorry.
> Yeah, that's just an example. We don't actually plan to follow every
> step from Apache. Our goal here is to have every project under a main
> module, which is umit. That will make it easier to integrate, maintain
> patterns, and distribute. It may sound ok to keep as we currenlty are,
> but think of how we'll maintain the pattern when we have more 10 or 20
> other projects and everyone decide to take a different path. It won't
> work. If everyone just keep using that methodology of putting their
> package inside umit, we'll have a better integration and it will be
> clear to everyone which pattern to follow.

I'm sorry Adriano, but please do not use 'better' and other adjectives
which are non-measurable in technical world. I wanna see some facts.
When I compress your words above i got only one thing 'it's for
integration'. But i don't see how it would help. Worse, I see how it
would break integration process. So, please write in some points,
benefits of it. I really wanna read and understand it. Maybe I'm
wrong.

>> Firstly, who decide it? And why someone wanna decide about some
>> projects which are made by other contributors? I don't like this kind
>> of words like "UMPA will be something", because I would like to have a
>> discussion first.
> Mentors propose the ideas, then, *if* it is something that people
> doesn't agree we bring that to the list to discuss. But, the final
> answer is upon the mentors, because if we leave to be decided by
> everyone it will take forever to decide on any subject. We decided to
> take this path. You didn't like it, that's ok. Now, you must come with
> solid reason on not to go on this way, and perhaps technical
> impossibilities that proves us wrong. Otherwise, there is no reason to
> change, as the goal of Summer of Code is to improve the project and
> that's what we believe well get if we take this path.

OK. Is it about Summer Of Code or about Umit Project? Sorry, but if
you see this line between mentors and students, so I have a question.
There is 3 mentors, and 5 students. But there is more contributors.
More, Summer Of Code is 3 month period. What about the rest of 9
months? Then the rest of us have the same importance? I don't any
relation between restructing of Umit Project and Google Summer Of
Code. If you really see (I'm hoping you don't), then explain me what
about other 9 months where there is 3-4 active developers? If you
don't the relation, please stop use argument about mentors/students.

> I understand it it a boring work, and that it may sound awkwards now,
> but it is something that will help both your project and the
> organization, don't worry.

Sincerely, I would like to see how it would help us.

>> Secondly, this idea about UMPA e.g. is completely bad. I wanna hear
>> some real arguments first. Then I will say why it's bad.
> Let's try to do the oposite, tell us why it is bad, and then we try to
> adapt the organization goal to the project needs if it is necessary
> and bad indeed.

Francesco pointed a lot of arguments agains this idea. More e.g. UMPA
use exactly 0 line of Umit code. It doesn't need to be integrated in
any way. It makes only limitation for UMPA. It has different licence.
Umit uses UMPA. But UMPA is a library and we are hoping that other
projects might use UMPA in the near future. But library it's a
library. It's a bit distinct between library and anything else. If I
were in other project and I would like to use some library to modify
network packates, and I will see UMPA as a part of something bigger,
propably I would drop it, because I will get more than I expected. I
don't need then GUI and some application. I wanna library. Why any
projects always seperate library from the other parts of projects?
Because it's useful. Because someone else can use it and don't feel
confuse. Please, don't make another Microsoft Windows from this. Let's
keep UNIX style. since 60's it looks like it's a good way to separate
things. Why Qt is not included in KDE in any way? Why Gtk is not
included in Gnome package? Would you really want to use gnome.gtk ?
You know that it would makes some other barriers.
Please don't break some rules of computer science, just because.
Otherwise, please provide strong arguments first.

On Sat, May 2, 2009 at 7:57 PM, Luis A. Bastiao Silva
<[email protected]> wrote:

> Actually we release UMPA as library right? So we will continue release it as
> a library. It is just to clarify, what changes is, geek talking:
> mkdir umit
> mv umpa umit/
> and changes the imports

Noone said that it's hard to change it. Noone said that it's worth to change it.

>> I don't know how much you understand structure of Apache code, but
>> not. It's not the same. Sorry.
> At least in Java, it works org.apache.p1, etc...

Sorry, you don't really know a structure of Apache. Please stop in
this point, or read more about Apache.
Also, we don't use Java right? Are you going to use getters and
setters as well? Becuase Java's developers have to use it?

>> Firstly, who decide it? And why someone wanna decide about some
>> projects which are made by other contributors? I don't like this kind
>> of words like "UMPA will be something", because I would like to have a
>> discussion first.
> It is not just about UMPA. It is about Umit Project.

Of course. But I disagree with puting UMPA into Umit structure.
Because there is no relation. Don't mix tomato with a jam.

I don't say about others but if you wanna know my words...here is it:
I think that we should provide some kind of packages for GUI.
Something like Umit Project where most of projects should be
integrated (Mapper, NetworkInventory, BT etc), but some of them should
live alone. Like BT. I think a lot of people would be interest in
UmitBT, but they will drop it because not everybody wanna to see 10000
of others features. So, I would see it as one big umbrella and some
drops.

------------------------------------------------------------------------------
Register Now & Save for Velocity, the Web Performance & Operations 
Conference from O'Reilly Media. Velocity features a full day of 
expert-led, hands-on workshops and two days of sessions from industry 
leaders in dedicated Performance & Operations tracks. Use code vel09scf 
and Save an extra 15% before 5/3. http://p.sf.net/sfu/velocityconf
_______________________________________________
Umit-devel mailing list
[email protected]
https://lists.sourceforge.net/lists/listinfo/umit-devel

Reply via email to