From my conversation with GoBE my impression was that V3 did have a Windows UI layer added onto the BEOS code. From a design perspective it was a clean layer. From a practical stand point, the code base is huge. The manpower to port it is man, many man months. Maybe man-years. Man meaning persons. Porting Productive source to Haiku is just not a realistic effort.
What is realistic is running V2 binary on Haiku. Get yourself a trial version of Productive V2 and give it a try. It’s actually hugely better than any of the options proposed in this thread.
So, the only option for the future of Haiku is to forever maintain compatibility with BeOS so that the old applications existing only as binaries can still be run?
Without an integrated “office application”, Haiku will always just remain an experimental curiosity for me.
In terms of porting efforts, which one appears the less onerous - bringing LibreOffice to Haiku or bringing Productive back to Haiku?
Seriously, get a copy of GoBE Productive and give it a try. It is the single-most sophisticated app written for BeOS or Haiku. It does now more than any other office productivity app ever will.
Productive 3.0.3 (for Windows) was released around 2001. That’s about 15 years ago.
Productive 2.0.2 (for Beos) was released a few years before that and, by its age, is approaching what is often considered to be the “vintage computing” category.
When I started this thread I was only thinking about GoBe 2.0.1 getting open sourced to the Haiku community, mostly because that version was a native BeOS app that already runs well on Haiku. If it was made available to the community, it would be possible to create a build recipe, package it in hpkg format, and development could be picked up again.
I had not retraced the discussion to its origin and thus had not realized that 1) it started 2-1/2 years ago, and 2) I contributed ideas and suggestions earlier.
Going through the discussion history:
There were four developers who had signed a non-disclosure agreement and may still have a copy of the Productive 3.0.3 code. The link provided to a repository has been broken for some time.
Copies of the Productive 2.0.1 code may still be exist somewhere in the files of the original developers. Given its age, there is a high probability that it has vanished unless the repository was maintained with some form of a source version control and “cloning” the 2.0.1 branch might be all that is required.
The third-party code is said to relate to the dictionary/spell-checker. There are open-source spell-checkers, for example GNU Aspell. Their licenses may conflict with the desire of the original Productive developers and owners of the intellectual property to only allow distributions of binaries.
Productive 3.0.3 for Windows requires around 16 MB of hard disk space for its installation. This implies that the underlying code is large but much smaller than Calligra/LibreOffice.
Yes, it will be a massive effort to get Productive in source code form to “work” with Haiku as it now exists. However, that is the only way which would allow it to be usable on Haiku-ARM and Haiku-X64.
To get to the finish line, one needs to start walking. Where would we be now if the walking had started 2-1/2 years ago.
Here’s my recommendation for starting work on GP V3 for Haiku. Start a fundraiser for 1 month’s salary for a professional C++ developer. 3.5 weeks to dig into the code, assess issues resulting from BeOS/Haiku change, old C++ compiler/new C++ compiler, old IDE build system/new Jam build system, 3rd party code removal, other issues. .5 week to write a detailed report itemizing issues, and answer questions. After the assessment is provided and discussed we would have a better idea of the true scope of the development effort. Aside from the 3rd party code, the true issue is the huge amount of code in the codebase. This is people-years worth of development. Multiple expert coders writing code for multiple years. Perhaps just the word processor could be ported first, then the spreadsheet, then the presentation app. But the code would need to be assessed first. But this can only be done by another expert C++ coder and would take literally multiple full-time weeks to complete correctly. This is my recommendation as a professional project manager, as well as someone who has ported multiple BeOS apps to Haiku.
" Perhaps just the word processor could be ported first, then the spreadsheet, then the presentation app"
GP is not like that. It is a Works-style package with a single executable and a single file format. So removing the other sections would mean that the word processor could no longer do tables, because it actually uses the spreadsheet code to do that, and so on. If all we want is a word processor, Beatware Write would be a better starting point.
It looks like there is a plan (using AndrewZ’s scope outline)::
-Know how much one month’s salary for a C++ developer of the necessary caliber and skill is.
-Renew contact with owner(s) of Gobe Productive and ensure access to a copy of the Version 3.0.x source code.
-Do a fundraiser for that amount plus some contingency funds.
-Find the developer, establish the contract and any necessary non-disclosure agreement.
-Get going with tasks of identifying third party code for replacement, defining method of reversal of APIs from Windows to BeOS/Haiku, identifying IDE/compiler/build system issues and path for resolution, and summarizing the findings.
-Define the full scope based on the assessment, its cost and schedule.
-Either go ahead and do a fundraiser to execute, or don’t go ahead and forever close the question.
It does appear that “Haiku Inc.” will need bless this as it will have to be involved regarding the fundraising and the contract with the developer.
Left aside so far is the whole question of user documentation (there was only an on-line manual) and the generation of test documents.
I can’t speak for the Haiku Inc. guys, but I doubt they’ll get involved with all that. People been clamouring about using kickstarter for years and this looks like a perfect project for it.
I doubt though that anyone will actually do it. It’s a huge task to create a viable kickstarter campaign and, should it be successful (which I doubt), actually see the project through. And that’s only for the recon mission into the code. The actual work on the codebase to release a new GP will be a hundred times harder.
A kickstarter to get LibreOffice to Haiku seems to me more realistic, future-proof and useful. At the same time I doubt the few existing Haiku users will be able to come up with the required funds.
Of course, I’d love to be proven wrong, so don’t let this little rain keep you from your parade…
Damoklas: Well, Beatware gave us Sum-It for free, and turned a blind eye to the widespread use of Get-It and Mail-It for years. So if that code is still lying around on a HD somewhere, it’s not earning anyone any money now.
“A kickstarter to get LibreOffice to Haiku seems to me more realistic, future-proof and useful.”
As long as we realize what we’re getting into. I actually looked it up about two years ago and LibreOffice is a larger code base than the whole of Haiku!
Actually it will work to first finish the word processor only… you can test it out… Spreadsheet and Tables, Image, Presentation are all plugins… so if they are not there (if you delete them)… the Menu that you click will simply not work…
So its totaly possible
I stand corrected, although having a wordprocessor with a bunch of menu entries that do nothing won’t be good for Haiku’s image, but I suppose those could be commented out.
After this is done, you have an idea what needs to be done, and the only guy with access to the code still can’t share it with others.
Really, I think it is not worth the effort to get GoBe up and running again. It doesn’t know how to read and write modern document formats, it is most likely a huge codebase, the latest version isn’t even designed to run on BeOS.
If you are serious about it, the funding should be enough for one people or a team to:
Get the sourcecode from whoever owns the rights, with proper license agreements,
Remove all 3rd-party code, etc
Get the thing released under an appropriate free software license, again with agreement from whoever owns the rights.
Then you have an open source piece of code that doesn’t even build, and you need someone to take it from there.
Last but not least, there have been several annoucements, fundraisings, etc and nothing happened. Here is a good summary of the story, with relevant links:
So we had:
At least 2 attempts from GoBe people to get the development started again (including funding new companies, trying to acquire the sourcecode, etc)
A fundraising at beunited reaching $10000, and a negative answer to their attempt, saying that the funds to just get the thing open sourced would be more like $200000.
Not to mention, you’d have to first discover who actually owns the right and the sourcecode backups these days.
These are very relevant and valid points representing hurdles to proceed.
Maybe it is possible for Haiku Inc. and the owners of the Productive to establish what could be described as a master non-disclosure agreement for essentially two purposes:
for Haiku Inc. to organize fundraising and hire developers with the purpose of providing binaries built from the resulting code with Haiku. Any developer hired would have to sign a non-disclosure agreement as well.
for Haiku Inc. to be custodian of a copy of the original source code within its archives (electronic as well as a physical like a CD-ROM).
From past history, there is a very high probability that the venture will not go further than this. What will be avoided though is again “loosing” the source code as the ultimate custodian would be Haiku Inc., not an individual contract developer…
AndrewZ estimated a month of experienced C++ developer to perform the assessment and recommendations. This should be subjected to a sanity check and adjusted as necessary. From the resulting scope of work, it is a matter of establishing the budget, fundraising, and finding a suitable developer. With a well defined objective and scope, the fundraising would be hopefully more positively received than in the past.
The full scope established from the assessment would allow to approach the major fundraising campaign on a firmer footing than with budgets and time frames pulled from the air. One will need to be creative as a crowd funding campaign is typically for a product that the funding participants will get first and/or with a discount compare to the sale price once established. In this case, the product will be “free” to who-ever obtains a copy of Haiku. So, where is the incentive for one to participate in the campaign?
As I understand the history of the founding of GoBe, Productive is the third iteration for these people to develop an integrated “office suite”. The underlying concepts, refined for the BeOS/Haiku API, are likely more mature than anything which can be ported from other APIs.
I believe it is important and relevant to bring Productive to a state from which it can evolve along with Haiku (ARM binary?, X64 binary?) rather than forever being stuck in BeOS compatibility mode. Actually, bringing Productive to Haiku could be a more effective rallying point for the community than a mascot!
If you read the history thread I linked to, you will see that a past fundraising attempt reached $10000. Back then it was run by BeUnited.org, which was the entity coordinating efforts toward evolution of the BeAPI between Haiku, Zeta, Cosmoe and BlueEyedOS. However, when they approached GoBe, the result was that cleaning up the source code and releasing it into the open source world would cost something around $200000. This was way out of reach back then, and still is today.
I don’t understand why people insist that Haiku, Inc should be in charge of everything. Haiku, inc is a support association for developping the Haiku operating system. It is not their goal to act as a funding platform for such things or buy deprecated sourcecode from defunct companies. And, they already barely have time to handle their current activities, anyway.
If someone is really serious about getting GoBe open sourced, the first step is locating whoever owns the sourcecode these days, and ask them for their current estimate of the cost for open sourcing it or acquiring the codebase from them (no NDA, just buy the whole damn thing and full rights for it). Then, run a Kickstarter or so to fund your effort. Once you have done that, you can donate the source to Haiku, inc if you really want them to own the code.
Alternatively, I’d be fine with GoBe being a commercial app for Haiku and I’d buy a copy, if it allows working with modern fileformats (docx and the like or odt). So, is someone willing to run a company to maintain and distribute it?
