Is it safe to assume that a true native OOo will never happen? There appears to be not discussions or noise even any more.....Too bad I guess NeoOffice/J is the best we get....Dont get me wrong it is great but a real native app would have rocked.....
You know, now that Neo/J 1.1 is getting fairly stable, adding native menus, buttons, etc. to NeoOffice/J is a very feasible option.
Sun employs dozens of engineers ($$$) to work on OOo and pays Collab.net (more $$$) to run the OOo site and even with those levels of funding, they still don't have the time to implement native widgets on Windows.
In contrast, Ed, Dan, and I are keeping Neo or Neo/J reasonably in sync with Sun's OOo releases. Pretty amazing for three people working in their spare time with zero funding!
So, my point is that you should be patient. We will get there, just not fast.
You know, now that Neo/J 1.1 is getting fairly stable, adding native menus, buttons, etc. to NeoOffice/J is a very feasible option.
Sun employs dozens of engineers ($$$) to work on OOo and pays Collab.net (more $$$) to run the OOo site and even with those levels of funding, they still don't have the time to implement native widgets on Windows.
In contrast, Ed, Dan, and I are keeping Neo or Neo/J reasonably in sync with Sun's OOo releases. Pretty amazing for three people working in their spare time with zero funding!
So, my point is that you should be patient. We will get there, just not fast.
Patrick
You are quite right, amazing work. And we all love you guys for it . Keep it up. I think it is safe to say that pushing OOo to mac users could really take off with a native look.
I try to follow development as close as I can without programming experience. I read that there's going to be a new API for widgets to facilitate native interfaces (or something). Is this getting anywhere and if so, will it make your life easier? I rarely see anything about it on the mailing list.
- Oscar
P.S. One possible thing to fix: navigator and stylist windows refuse to stay on top of my document window. They are easily retrieved, but it takes a second or two each time.
Joined: May 25, 2003 Posts: 4752 Location: Santa Barbara, CA
Posted: Wed Sep 08, 2004 11:14 pm Post subject: Not dead...just redirected attention
Well, between spending more time with my photographic and backpacking passions and keeping trinity hacker free and stuff I've had to shift my own priorities. A fully Cocoa port is just perhaps too large a task, especially for three people. What I've come to believe is the following:
1) It doesn't matter if it uses Cocoa, Carbon, Java, Qt, or what-have-you. Folks want blue buttons, that's it. They don't give a rat's ass about the underlying technology. It could be built on top of an Atari SE emulator as far as they care, but if it has the blue buttons it must be an OS X application.
2) Mating the world-vision of OOo and Cocoa is hard. Not impossible, but definitely difficult.
3) Even if they were mated, people in the OOo world of mostly Windows-ish interface design can't communicate with Mac design expectations. The interface will need to be rewritten.
4) If the interface needs to be rewritten anyway, what good is the effort spent banging our heads against mating a non-Mac interface to Cocoa Mac buttons? Why just not rewrite and rework the interface from scratch with a Mac perspective from the get go instead of doing the years worth of work just to reach goal (1) and have people still complain?
5) To write a good Mac interface you need to go back to the drawing board, to the point of hiring graphic artists to draw you better icons.
6) For programmers, a good Mac interface isn't any good unless it is implemented using traditional Mac tools such as Interface Builder or even (gasp) resources.
7) It'll take a he** of a lot of effort to achieve either 5 or 6 without a major rethink of the source architecture of OOo to allow a split between UI frontend and abstract backend (it isn't in place for 1.1.x or even 2.0).
8 ) The Java implementation on Mac OS X is fuc*!@4ing sweet.
9) The Java AWT and Swing implementations on Mac OS X are f***ing sweet.
10) Neo/J is f***ing sweet. And stable. (!)
11) So...step back and ponder...
12) To get to step 7 requires getting past step one and then convincing people across other platforms which don't need it yet of the need of fundamental change (are there really any consistent user interfaces whatever on Linux?). Very difficult in an environment whose primary contributor is a team that needs to file earnings reports (or prospectuses at the very least) for only Windows, Solaris, and Linux-x86 and only those platforms.
13) Neo/J is more stable then the Cocoa or QuickDraw stuff has ever been thanks to amazing engineering by Patrick. And it's the most up-to-date with OOo release versions of any of the prototyping projects.
14) Mac OS X Java kicks ass...(see ...so...if the goal of the immediate future is to just get blue buttons and not undertake the massive work of reengineering everything (impossible for just 3 people, and maybe even impossible for purely volunteer open source even)...
15) Why not just give NeoOffice/J blue buttons through AWT and Swing?
Users on other platforms already think OpenOffice.org already uses Java, which it doesn't. So instead of just listening to them complain, the more innovative solution is to just realize that Java is the best and most functional answer for our short term goals (on Mac OS X at least) and to go ahead and use it instead of undertaking the Herculean task of doing an exact port in a severely understaffed situation, instead leveraging off of all of the development effort already expended by Apple in making such a great JVM and great native Swing implementation.
This is my logic.
Thus I've shifted all non-server development efforts to Neo/J over the last few months.
Dunno if this answers your question but...dead?...no.
In spirit, no.
In actuality to accomplish the long term goals it will take a long time to reach.
The pragmatic engineer in me says that it's best to focus on delivering what the users want...request number 1...blue buttons. And right now Neo/J is the best platform on which to build the blue buttons. Functionally there is absolutely no difference between doing it in Java or doing it in Cocoa for an end user. So I've been focusing on doing it in Java for the time being.
If pure Cocoa Neo has to stagnate to do it, so be it...it's what the users demand
Posted: Thu Sep 09, 2004 2:58 am Post subject: Re: Not dead...just redirected attention
Well, I agree with #1 and #14, I was surprised by #3, and appalled by #7.
If you want to do a cross platform app, having an abstraction in the UI layer seems like the basic thing to do, even for a newbie such as myself.
So basically, who not use good tools for leverage? As long as it doesn't kill performance and memory consumption and if it looks nice, few users will care. Cleanliness is next to godliness and users only look at the skin anyway .
Apple toutes Java as a fully feasible programming platform anyway. I don't know how much bigger or slower a Java implementation would be over a Cocoa one, but working is working. Currently, both OOo and NeoOffice/J are a bit slow, but I trust some pruning and optimizing would help.
I was also surprised that other OOo implementations don't use Java. Is it just the Sun association? I recall it said in the docs you need a Java VM to run it. All the major platforms for OOo have a JVM, afaik.
Again thank for the effort, and good luck!
OPENSTEP wrote:
Well, between spending more time with my photographic and backpacking passions and keeping trinity hacker free and stuff I've had to shift my own priorities. A fully Cocoa port is just perhaps too large a task, especially for three people. What I've come to believe is the following:
1) It doesn't matter if it uses Cocoa, Carbon, Java, Qt, or what-have-you. Folks want blue buttons, that's it. They don't give a rat's ass about the underlying technology. It could be built on top of an Atari SE emulator as far as they care, but if it has the blue buttons it must be an OS X application.
3) Even if they were mated, people in the OOo world of mostly Windows-ish interface design can't communicate with Mac design expectations. The interface will need to be rewritten.
7) It'll take a he** of a lot of effort to achieve either 5 or 6 without a major rethink of the source architecture of OOo to allow a split between UI frontend and abstract backend (it isn't in place for 1.1.x or even 2.0).
14) Mac OS X Java kicks ass...(see ...so...if the goal of the immediate future is to just get blue buttons and not undertake the massive work of reengineering everything (impossible for just 3 people, and maybe even impossible for purely volunteer open source even)...
15) Why not just give NeoOffice/J blue buttons through AWT and Swing?
Users on other platforms already think OpenOffice.org already uses Java, which it doesn't. So instead of just listening to them complain, the more innovative solution is to just realize that Java is the best and most functional answer for our short term goals (on Mac OS X at least) and to go ahead and use it instead of undertaking the Herculean task of doing an exact port in a severely understaffed situation, instead leveraging off of all of the development effort already expended by Apple in making such a great JVM and great native Swing implementation.
From a user's standpoint, I agree with you. I don't care if the underlying technology is Java or Cocoa (thought from a programmer's frame of mind, it makes a difference, but don't ask me why, It's just my being used to think different .
From all practical purposes, Neo/J is a native application. Right now, it is somewhat slower than the X11 version (you can feel it on a G4/400 MHz), but I guess that it will get better. It is way faster than the 0.8.4 which was barely useable on my machine. Neo/J, even in its alpha state is very stable for what I do, so for me it makes a lot of sense to concentrate on this version. I think it is a very good idea. Is it official?
A couple questions :
Can we now consider the X11 port to be obsolete (I don't know if Darwin users may need it)?
Can we expect at this stage Neo/J to closely follow the latest subversion number in 1.1.x? On which subversion number is 1.1 Alpha based, actually? I haven't seen this info anywhere.
Joined: May 25, 2003 Posts: 4752 Location: Santa Barbara, CA
Posted: Thu Sep 09, 2004 8:41 am Post subject: Re: Not dead...just redirected attention
ovvldc wrote:
If you want to do a cross platform app, having an abstraction in the UI layer seems like the basic thing to do, even for a newbie such as myself.
There's a layer of abstraction in there which is why it's as portable as it is to QuickDraw or Java2D...it's on the same level as how QuickDraw and Cocoa Graphics are really abstractions on top of CoreGraphics/Quartz. OOo 2.0 is extending that abstraction up to the level of buttons and menus using techniques similar to the ones we outlined in the Neo prototypes (the Native Widget Framework). It doesn't use native widgets, but allows them to be drawn differently on a platform dependent basis. In an effect it creates an environment kind of like Swing's pluggable look and feel modules. I think it's good that we had Neo to use as a testing ground to prove that it was feasible and could still make a functional product.
That makes it easy to change the look of the entire application, but it doesn't change the feel. On a Mac the look is really only half the battle. It's the feel that's important. Right now OOo has no way in 2.0 to make a pluggable "feel" (e.g. replace the entire interface with something that clones the old Word 5.1 UI).
Ideally I want to get it to that point, to where a Mac developer can simply whip out Interface Builder, set up some new bindings, and snap crackle pop out comes a different interface. A Mac interface. Or perhaps even a completely new interface. OOo doesn't become the final product...it becomes a framework, an enabling technology for other interfaces and even other applications (heck, I'd like to embed a functional spreadsheet in my own apps!). Reaching that ideal is what I want to do, but that's the one that's very difficult to do in the current OOo code structure and it's not really possible to do it alone on a part-time basis since it's too big
ovvldc wrote:
I was also surprised that other OOo implementations don't use Java. Is it just the Sun association? I recall it said in the docs you need a Java VM to run it. All the major platforms for OOo have a JVM, afaik.
I think with 1.1 they started to use Java for some things like the raw XML file export filters and some of the wizards. Essentially they're mostly scripts written in Java though...the 8 million lines are still all solid C++. You can run OOo just fine without any Java whatsoever (as evidenced by the broken Java support in Mac OS X OOo 1.1.x )
Historically Sun got OOo through an acquisition of a company named Star Division, not through internal development. But folks tend to forget that and think it was developed purely by Sun. Most folks also assume everything developed by Sun is written in Java (Solaris is definitely not Java code).
Posted: Thu Sep 09, 2004 11:17 am Post subject: hhmmm
Well let me start off with saying that I hope I never offended anyone.
Next I do want to thank and let you guys know the reason I ask was not because I was unhappy but rather so happy with you have delivered that I want more...(Doesnt everyone)
I have to say that I can totally understand you thoughts about moving your efforts to Neo/J and improving its look etc etc.(blue buttons) I have tried the new 1.1alpha and have to say that it is great. I had switched to the X11 version just because of the dramatic improvments that had been made with it latest release, but after trying out the 1.1alpha I am wanting to use it! The awesome font displaying and overall smooth appears vs X11 is great........
So Patrick and Ed I thank you both and everyone else who contributes to try and help your efforts. Again dont take anything negative.
Joined: May 25, 2003 Posts: 4752 Location: Santa Barbara, CA
Posted: Thu Sep 09, 2004 8:44 pm Post subject:
No, I don't think anyone's offended. It's definitely a valid question and I think it's an important one to ask
More often then not I have my own sense of an engineering "strategy" for attacking the project, but I don't always convey it properly or seek outside opinion. And I can just seem to disappear for days and weeks, probably because I'm doing something personal
I'd love to spend more time on Neo and really starting to get my head around the engineering necessary to do it right, but I just have this stupid thing called a job that's taking up too much of my time
If I may pipe in, one thing I'd say is that getting the menu up at the top is also important. Mac users aren't used to looking in the window for the file, edit, etc part.
do they really have to be blue? can't they be red or something? word is blue. =)
and damnit, you are supposed to send me stuff so i can help with the trinity admin. =P
actually i would reccommend against deleting the neooffice forums, rather just to put them off into an archive group (maybe make it so you can't post in them). too much useful info in there.
a revamp of the front has been up for a while, if you write up stuff, i'll put together a template and we can see about making a new face.
Joined: May 31, 2003 Posts: 219 Location: French Alps
Posted: Fri Sep 10, 2004 11:38 am Post subject: Disapointed
I like Neo/J, but I'd have much liked a Cocoa port.
Implementation DOES matter. It's not only look and feel but also performance and integration.
Flamming-Yeti (carbon) and Whatever-Gremlin were not usable applications but were (looked) already halfway.
But if it is to much work to be done, what can be said. I rather do not understand how you (3) guys did the work you did, beside a regular job to fulfill. Far beyond my potential.
Joined: May 25, 2003 Posts: 4752 Location: Santa Barbara, CA
Posted: Sat Sep 11, 2004 7:34 am Post subject:
jakeOSX wrote:
If I may pipe in, one thing I'd say is that getting the menu up at the top is also important. Mac users aren't used to looking in the window for the file, edit, etc part.
Yeah, after talking Patrick and I determined that was the most important step. I'm actually working on that right now (literally) for Neo/J. I hadn't expected the AWT menus to smoke more crack then the Cocoa menus, but they do
After that I think the next things will be starting with buttons and the like. Window backgrounds may wait until last, though, since the grey kind of looks a little bit like Metal
Joined: May 25, 2003 Posts: 4752 Location: Santa Barbara, CA
Posted: Sat Sep 11, 2004 7:37 am Post subject:
jakeOSX wrote:
actually i would reccommend against deleting the neooffice forums, rather just to put them off into an archive group (maybe make it so you can't post in them). too much useful info in there.
What I did do is move them below Neo/J in the list
jakeOSX wrote:
a revamp of the front has been up for a while, if you write up stuff, i'll put together a template and we can see about making a new face.
Yeah, updating the look of the main servers would be good to do. I haven't had time
You cannot post new topics in this forum You cannot reply to topics in this forum You cannot edit your posts in this forum You cannot delete your posts in this forum You cannot vote in polls in this forum You cannot attach files in this forum You cannot download files in this forum
I guess NeoOffice/J is the best we get....Dont get me wrong it is great but a real native app would have rocked..... 
. Keep it up. I think it is safe to say that pushing OOo to mac users could really take off with a native look.
...so...if the goal of the immediate future is to just get blue buttons and not undertake the massive work of reengineering everything (impossible for just 3 people, and maybe even impossible for purely volunteer open source even)...
