Apps and things

1

This is my first post so I’m hoping not to get flamed too much, I’d also like to say I’m a non-programmer. :oops: I first tried BeOS before Be went bust, I think I was using Windows 98 at the time, I’d also tried Corel Linux. I was immediately impressed with BeOS it was fast, fast to boot, and easy to set up, in contrast to Corel Linux which didn’t install properly and left me frustrated and unable to get the thing working.

However, I didn’t stick with BeOS due to lack of useful apps, even the browser was rubbish, though you could get an out of date version of Opera. Sad but I went back to Windows. Since then I’ve installed and run Mandrake (Linux has certainly improved) and it comes with so many apps (too many). Open Office, the Gimp etc. I’ve also run BeOS again and although there is now a decent browser there are still too few apps.

I feel BeOS or Haiku has real potential – but the problem of apps is very real, I must admit I liked the idea of Blue eyed OS, but that seems to have died, all those Linux apps on something like BeOS (although I feared it would be difficult like Linux.) So here’s my question, are people working on porting Open Office, the Gimp etc to BeOS? Is it possible to make a compatibility layer in Haiku or something so that it will run Linux Apps? Could Wine be ported to BeOS?

2
humpmihk wrote:
So here’s my question, are people working on porting Open Office
Yes, http://www.beunited.org/index.php?page=developer
humpmihk wrote:
Is it possible to make a compatibility layer in Haiku or something so that it will run Linux Apps? Could Wine be ported to BeOS?
This stuff is already in multiple OS's, someone just has to put in the considerable amount of work to do it.
3
humpmihk wrote:
Is it possible to make a compatibility layer in Haiku or something so that it will run Linux Apps? Could Wine be ported to BeOS?

There is a port of X11 for BeOS, but from what I gather, it’s pretty clunky. It helps when porting x-based linux apps to BeOS so that a native UI doesn’t have to be built for the port. The Gimp requires GTK+ which has been ported in the past with the help of the X11 port…

see the following:

X11: http://www.bebits.com/app/337
(note that it was provided by The BeWine Project which used to be attempting to port Wine to BeOS)

GTK+: http://www.bebits.com/app/3585
Gimp: http://www.bebits.com/app/3586

And as already pointed out, BeUnited is working on an OpenOffice port.

Also for cross-platform support, BeUnited has already made much progress on a Java port :slight_smile: That will bring some additional apps to BeOS, and probably provide a better browsing experience.

4

A project hosted on BeUnited is not an indication of active development.

for example, JavaShare is hosted on BU, but is not actively maintained by Bryan Varner. IIRC he wanted to host it at BU in an attempt to spark the interest of other developers because he chooses to work on other projects, mainly Java. ( i don’t remember the link )

projects on BU should be considered as open-invitations for you to work on them :slight_smile:

5
mmadia wrote:
A project hosted on BeUnited is not an indication of active development.

for example, JavaShare is hosted on BU, but is not actively maintained by Bryan Varner. IIRC he wanted to host it at BU in an attempt to spark the interest of other developers because he chooses to work on other projects, mainly Java. ( i don’t remember the link )

projects on BU should be considered as open-invitations for you to work on them :slight_smile:

Yes, good point. I have in fact been wondering if there is any progress on the OpenOffice port… I have suspected for some time that there is very little… but then again, not many people were sure what the status of the Java port was until WalterCon - and I got the impression it is going very well :slight_smile:

6

Why does everybody want an OpenOffice port? I can’t understand this.
OpenOffice on Haiku or BeOS would most probably be even slower and more resource-hungry that it is on the systems for which it was originally designed.
How did be make such a good OS? They started from scratch.

I propose we rather make:

  1. OpenOffice format translators - the format is open and documented, so this shouldn’t be hard.
  2. A Gobe-style open source word processor / office suite, probably taking some ideas (architecture-wise more than UI-wise) from KOffice. It may be very basic at first, as long as the translators are there it doesn’t matter. It should just be made for Haiku / BeOS. we need something SLIM and FAST!
  3. If you have this, you can improve it - add spell checking, feature X,Y, and Z and so on. You can add a plugin API…

This can all be done in ~2 years by a bunch of coders. The OpenOffice.org port is (or isn’t? who knows) going on for more than two years now, with no results so far.

7
Quote:
This can all be done in ~2 years by a bunch of coders. The OpenOffice.org port is (or isn't? who knows) going on for more than two years now, with no results so far.

if we could turn the community into a utililitarian / dictatorship,. that may be possible :slight_smile:

Style of Organization:
A group of people coordinate and prioritize projects. Each project lists tasks for Research, Development, QA and who is working on each task. Everyone signs up for at least one thing. If you cannot code, you can do research or QA (Quality Assurance – methodically finding and reporting bugs).

This could even teach non-programmers how they can help the most; teaching them how to do proper QA and how to do proper and meaningful research that will assist developers.

While this will require people to work on stuff they may not want to do, it could be a very effecient way to organize and complete projects.

IMVHO, this type of organization could very well make or break the entire community. Some people would very easily be put off by the “forced labor” style of contributing. Though if it worked, which i admit is a long shot, it could dramatically increase our productivity – maybe even draw new people to the scene…“Hey, these people have a straight-forward plan for making progess. I can see what they want and I’d like to help. In fact, according to this list, i can start by doing FooBar”

I believe that most people in our community have a serious desire to help but they simply don’t know what to work on. …Look at other posts in this forum “I know how to do such and such, what can i do to help??” This style of organization can help to inspire people.

I fully understand that most dev’s are focused on R1. IMO, they should be. IMO, Haiku is priority #1. I also realize that Haiku has its own style of organization and right now re-adjusting to a new style of organization could hurt development ( at least in the short term. Maybe once R1 is released, something of this style should be implemented. It would re-focus everyone as Haiku’s goal of re-creating beos r5 in Haiku R1 would be completed.

On a closing note, this style of organization could also further distinguish Haiku apart from the crowd: An organized system of open-source workers. I know of no other open source platform that has such a system of coordination|organization. That could even draw additional interest, both in the form of News headlines and more newbies.

In short: Once R1 is released, Haiku Inc., needs a new method to organize its workers, as aiming for R5 no longer applies. My proposed organizational scheme is rational, well-informative to users, creates well defined development paths and lists the needed objectives for each.

(PM me if y’think i should drop this to the GE or Haiku/OBOS email lists)

8
mmadia wrote:
I believe that most people in our community have a serious desire to help but they simply don't know what to work on. ....Look at other posts in this forum "I know how to do such and such, what can i do to help??" This style of organization can help to inspire people.
*mmadia hits the nail on the head* I've been looking for something to do for a while but every time I think something up it's either already done or someone is already working on it.

Perhaps both of these systems could work in parrallel. People like me don’t have the low level skills to work on OS development, but can readily help with application development. Unfortunatly I’ve been away from the BeOS community for some years and I don’t know what’s done and/or what needs doing and/or what people are working on.

Perhaps Haiku should make a “wish list” (or even a team or sister project to do this) to manage these things. This way people who have my type of skill (which in my experience is the most common) will have a list of something that they can do contribute to and those that want to contribute to the OS can.

Quite frankly this has been a very fustrating issue for me. And I really think that the attitude of “look through the source and find something to do for yourself” turns a lot of people off. I has for me anyway, and when one is like this other surely are as well.

9
theturner wrote:
Why does everybody want an OpenOffice port? I can't understand this. OpenOffice on Haiku or BeOS would most probably be even slower and more resource-hungry that it is on the systems for which it was originally designed.

It was originally designed to be multiplatform… the fact that it is slow on every platform it runs on should be enough to discourage it’s use, although compared to the full featured microsoft products that it is meant to compete with instead of the smaller open source products that run very quickly but support no features, open office does pretty well.

10
Quote:
Perhaps Haiku should make a "wish list" (or even a team or sister project to do this)

The Wish List is the Suggestion Box … and the Team that looks after this is called Glass Elevator. That’s where they discuss all the new things for enhancing R1.

Quote:
I believe that most people in our community have a serious desire to help but they simply don't know what to work on.

All they have to do is ask … we have the forums, the mailing list and the IRC channel where Team Leads can be contacted.

Being one of the Team Leads, I’ve seen a lot of people say they can contribute, I give them all the info they need and assist where I can and then within 3 months, they are never heard from again.

Don’t worry I’m not bitter about it, just if you are going to contribute, no light weights please.

11
Quote:
On a closing note, this style of organization could also further distinguish Haiku apart from the crowd: An organized system of open-source workers. I know of no other open source platform that has such a system of coordination|organization. That could even draw additional interest, both in the form of News headlines and more newbies.

And it would be a very good thing to help keeping Haiku what we love in BeOS: It’s straightforward, not bloated, and simple.

12
Quote:
I give them all the info they need and assist where I can and then within 3 months, they are never heard from again.

Imagine that information being presented in an easy to display & easy to maintain website… the interested people can look over all the available “task listing” (researching protocols, creating GUI mockups, testing a certain aspect of a program, coding a specific functionality)… You, as team lead wouldn’t need to keep re-gathering, re-organizing, and re-sending the same information.

Quote:
And it would be a very good thing to help keeping Haiku what we love in BeOS: It's straightforward, not bloated, and simple.

Exactly. It brings order to chaos. After R1, Haiku will be on it’s own two feet. Let’s do our best to keep it on the straight and narrow.

13
Sikosis wrote:
The Wish List is the Suggestion Box ... and the Team that looks after this is called Glass Elevator. That's where they discuss all the new things for enhancing R1.
I was under the impression that the suggestion box is what people want from the Haiku project. What I was getting at is having the teams make a list of things that won't be included in R1 for whatever reason and (as has been mentioned above) putting them up on a web site somewhere.
Sikosis wrote:
Quote:
I believe that most people in our community have a serious desire to help but they simply don't know what to work on.

All they have to do is ask … we have the forums, the mailing list and the IRC channel where Team Leads can be contacted.

Being one of the Team Leads, I’ve seen a lot of people say they can contribute, I give them all the info they need and assist where I can and then within 3 months, they are never heard from again.

Don’t worry I’m not bitter about it, just if you are going to contribute, no light weights please.


Actually, I think having that web page would help to prevent getting light weights in the mix. Just post along with the said info that they’ll only be considered for contribution if in the email that states there intentions, they have a good chunk of code already compeleted and ready to submit to CVS.

And I did send an message to mphipps about a potential project (some legal stuff came up and I wanted him to decide what was best), also mentioning my fustration of having nothing to do, before the weekend and I know he read the message (the messaging system here tells me that) but as of yet have no response. (EDIT: He actually did respond to the first message but didn’t actually answer my question. I still await a response to the second where I put my question (and noted my fustrations (see below)) in no uncertain terms.)

I’m not bitter or bitching or anything, it just adds to my fustration. I mean this perceived attitude of “look through the code yourself and find something to do or come up with something to do on you own” works fine in the world of Linux because of the sheer volumn of people that are part of that community. But, with Haiku this pool of people is limited and as such requires a much greater level of organization.

14
theturner wrote:
Why does everybody want an OpenOffice port? I can't understand this. OpenOffice on Haiku or BeOS would most probably be even slower and more resource-hungry that it is on the systems for which it was originally designed. How did be make such a good OS? They started from scratch.

I propose we rather make:

  1. OpenOffice format translators - the format is open and documented, so this shouldn’t be hard.
  2. A Gobe-style open source word processor / office suite, probably taking some ideas (architecture-wise more than UI-wise) from KOffice. It may be very basic at first, as long as the translators are there it doesn’t matter. It should just be made for Haiku / BeOS. we need something SLIM and FAST!
  3. If you have this, you can improve it - add spell checking, feature X,Y, and Z and so on. You can add a plugin API…

This can all be done in ~2 years by a bunch of coders. The OpenOffice.org port is (or isn’t? who knows) going on for more than two years now, with no results so far.

There was a port of AbiWord 1.0 for BeOS. If anybody could revive it and port the BeOS-related code from 1.0 to 2.x, it would be perfect.

A spreadsheet app based on something existing, is a bit more difficuilt. Maybe use parts of Sum-It?

As graphics app there’s ArtPaint. IMO it’s quite good, but does also need an update.

15

instead of it being a website, …
what about an application that connects to the net,
downloads a database and intelligently displays it to the user.

the application could sort by:
-task type ( research, dev, QA),
-preferred worker class ( graphic artist, icon-maker, gui designer, network programmer, … )
-number of active participants
-priority
-Estimated hours involved.

to restate my goal: getting more people (devs and even more importantly, those who don’t know how to code or how to code well) involved with application development. As writing code is only a part to the entire process(research, theorizing ideas, coding, testing)

16

I had a similar idea, a while back. I exchanged a few emails with Micahel on the subject too, and Simon Gauvin and a few other people.

It’s changed a bit since it’s original inception, but you can see the full idea here: http://forums.begroovy.com/viewtopic.php?t=4348

ps: Now I don’t think entirely user-directed development is a good idea, but a lot of the organisational stuff still seems like a good plan to me.

Simon

17
tb100 wrote:
I had a similar idea, a while back. I exchanged a few emails with Micahel on the subject too, and Simon Gauvin and a few other people.

It’s changed a bit since it’s original inception, but you can see the full idea here: http://forums.begroovy.com/viewtopic.php?t=4348

ps: Now I don’t think entirely user-directed development is a good idea, but a lot of the organisational stuff still seems like a good plan to me.

Simon

I’d say yes to this type of structure. It’d save the the time (that I’m not willing to spend and I gather others feel this way as well) to find something to do.

18

Hmm… openoffice shouldn’t be compared with small featureless textwriters when it comes to speed. It should be compared with other officesuits with the same amount of features and then openoffice rules :stuck_out_tongue:

HOWEVER…

It is somewhat slow and not-so-intuitive and HUGE and SLOW an HUGE and SLOW… This is true for many modern applications. They tend to get bigger and bigger and slower and slower and more and more bulky… I think most applications are developed by stupid coders using giant RAD-tools, without truly optimizing the code later on :evil:

Compare the size of Wordperfect 6.0a to Wordperfect 8 or OfficeXP (2002) or openoffice and compare the functionality and the speed. It really makes one wonder what the heck the coders are thinking, releasing such lame apps now-a-days :?

I believe/know that size of openoffice and many other apps could be reduced with approx. 80% if the code was optimized. And no RAD-tools was involved that is. Or least no design-time code was left back. And the speed would most likely increase (if the code is optimized properly) and definitely result in a lot lower RAM-usage.

Let’s take openoffice and rewrite it according to BEOS/Haiku-style :stuck_out_tongue:

(BTW: RAD-tools are great for creating prototypes, but if speed and ram-usage and stability is important then the code must be optimzed and all design-time elements rewritten completely)

19

Mr. Jones,

I can understand you. Optimized code, both in size and speed, is a fine thing. But there are IMHO some problems with that:

When you manually optimize a big app, you need very much additional time and this slows down the development process.

And may be you can’t continue to use rapid prototyping tools because they are unable to reuse your hand-optimized code.

When you manually optimize code, you may be unable to use the source code for different platforms. So you give up an important synergy for an OS with a small user base, like BeOS/Haiku, here.

Just my $0.02

Thomas

20

Heh… schlesi, you forgot to mention that optimizing code is really BORING :stuck_out_tongue:

I have a lot to more to say about this, but I’m about to play CS, so you’ll have to wait :wink: