Earlier on this thread people were asking for a bounty for MUI-OWB, my own personal opinion is that this should have been moved to Kickstarter months ago. That site is very popular and you will get interest from non-Amigans - eager to donate should you word it non-technically and enthusiastically.
AmigaBounty website has proven to be ineffective, big projects should be moved to Kickstarter in my opinion.
Look at the LibreOffice effort - could that not have been at the very least, tried on Kickstarter first? Same with Timberwolf/FireFox - did the developers exhaust these avenues first?
I've previously suggested Kickstarter, but don't forget www.indiegogo.com as an alternative to Kickstarter, if Kickstarter isn't suitable for some reason.
TW is soooo damn close to working perfectly on my X1k...just need to iron out the occasional hang ups when doing alot of scrolling but other than that rock solid here
Yeah just think how good it would be if it was hardware accelerated, updated to the latest build with full html5 support and able to support a pluggin for flash?
In its current state on my Pegasos 2 it is usable but the software rendering really slows it down, I hope Ssolie is able to have the crew bring us an update quickly.
@all If you ask me, odyssey bounty still a bit wrong in few places. There is few moments why it wrong:
1. There noted that libs are out of scope, because they "pure recompile with minimum changes". That wrong. Curl, fontconfig and some others have real changes, which no one will do, and ports will be not as good as they can be with all those libs as well. So, bounty text should contain word about opening them too. Sure Fab there says he will open necessary ones, but as bounty didn't says so, then its all can be ends with "its trivial, do it yourself, be real programmer, i have no sources by hands". But no one will rewrite curl from scratch to use semaphores. I have some versions from Fab, but sadly its out of scope in bounty text (they can be updated and changed lately for anything else and porters will ends up to "its trivial, do it yourself"). That ugly fontconfig also have bunch of changes: to have nice progress bar, amiga pathes and stuff. Making all that from scratch its just pure annoying.
2. Its still again the same old story about not real opensource , but giving a half of candy : i.e. we will have now latest src, but then, _maybe_ if fab will want, and if he will like how ports done, he will open new versions, if he will not want, and think that ports done bad (by any of reasons) he will not open them. That also wrong. Bounty should have "open all new versions all the time does not matter what".
At least 7500$ its imho ok for such requirements with my remarks. But just as it currently only for current version only, without 3d party libs, without real svn, dunno. Imho too early and not real open sourcing.
I.e. there is 7500 for current version + html5 player (which also, will need rewrite for os4), without 3d party libs (on which someone should work, which is trivial for fab , but not trivial for others) and without normal SVN (damn). I.e. its the same old packed archive, the same "diffs", the same os4/aros coders works on their own, the same no aos4/aros ifdefs in the fab's root code so no need to recreate a wheel.. Which mean, with every new version, if fab will open it (if, because its optional, and only if he will like how ports done), everyone will need recreate wheels, as diffs will stop working by any of reasons. And, while we will have current sources, still, fab will works on his own, in his own branch, and maybe/if/when he will open new versions too.
Imho text of bounty was too early. 3d party libs important (those ones which have changes), have aos4/aros ifdefs in the root code are necessary too and one single damn SVN for everyone also necessary too, or it will be all the time the same as we have now.
Edited by kas1e on 2013/11/13 5:34:58 Edited by kas1e on 2013/11/13 5:35:46 Edited by kas1e on 2013/11/13 5:37:20 Edited by kas1e on 2013/11/13 5:41:25
we have some good and capable developer , plus with source open everyone can make a better and optimized version of it, or why not devs can make a AmigaOs Owb dev group. Hope Trevor/Aeon, Acube,the hardware developer and sellers in amiga land will make their small contribute because something like this is not only for user but will garant a better future for our platform.
I need only to know will be the Flash player opensourced too ?
______________________________________________________________ I just made my small contribute my wife love amigaland
@kas1e Oh dear, so the current bounty is going to have to be cancelled, and a new bounty created (and moneys moved over)? Just like happened to an earlier smaller OWB bounty on that site?
@ChrisH I don't know. As for me it miss some moments. I.e:
--- all 3d party opensource libs which have any single change should be opened too.
--- flash plugin latest version opened too.
--- any new version should be released all the time, and that should be not optional, but one of bounty requirements. For now it literally says:
"as long as the resulting ports stay true to the original version of the Odyssey Web Browser in terms of user experience and feature sets, Fabien Coeurjoly plans to share all source code that he will develop himself for future versions of Odyssey Web Browser for free"
What mean that "user expirience" ? I.e. if one will say os4 version suck, then no source code of new versions ?:)
--- "how to compile" file should be good to have too. As at least when i port 1.9 and 1.16, build systems changes, options changes, etc.
Its ok to not have svn, and to not have ifdefs in source code (its suck , but ok). But have to deal again with all those libs, plugins and cry for new versions to be released all the time again and again will be for sure not so cool. Imho some bits in bounty just missed.
But i think ppls will collect that summ anyway. Just porting will be harder, and latest "user experience" will be worse. Just if we already do bounty, if we already collect some summ, why not do everything one time at once, to sort all possible problems in future, and to avoid all that mess which we have all the time.
Crikey, the bounty is worse than I thought. $7500 for code snapshot, with no dependencies. And conditions on further merging? Fab is really taking the piss!
I can understand Fab wanting to retain some control of his project, but that's quite possible with a truly open sourced solution (blender is the ultimate example of that as they still have full control of the main source tree).
You know Yvon Roznin released a fully functional completely self contained browser which he wrote from scratch and had been a commercial product for free!
@kas1e Who is responsible for creating the bounty? I'd like to know why they created it without first getting agreement from all the parties involved (esp. Kas1e)? That could lead to problems down the road.
And this part of the bounty is especially bad: "as long as the resulting ports stay true to the original version of the Odyssey Web Browser in terms of user experience and feature sets, Fabien Coeurjoly plans to share all source code that he will develop himself for future versions of Odyssey Web Browser for free"
A bounty is basically a contract, and a contract must NOT have vague terms (that nobody understands or agree upon the meaning of). Vagueness spells disaster down the road, potentially with bad feeling on all sides.
If Fab has concerns about what the end result will be like, then he should be able to list what elements he considers requirements of the "user experience and feature sets". This should be easy, because he already has the existing MUI-OWB port to refer to. Presumably he has seen some things he doesn't like there, and doesn't want the proper port to be the same.
My worry is that he considers something like "non-blocking network connections" to be central the user experience, but that is dependant upon a "Sub library" (curl) which is stated to be "out of the scope of this bounty". That risks Kas1e/etc saying they can't meet Fab's unstated "user experience" requirements because he doesn't agree to giving the Curl source code, and Fab saying it's not his fault when it was stated in the bounty, hence both sides blaming the other.
WORSE, what happens if the AmigaOS4 port meets his definition of "user experience and feature sets" but the AROS port doesn't? Does that mean he will refuse to "share all source code" with anyone (neither OS4 nor AROS porters)?
These sort of problems should have been sorted out BEFORE the bounty was created, which means getting agreement from everyone involved.
"Please note that the bounty terms have also been reviewed by other developers who have worked on Odyssey ports. While we did receive suggestions for modifications, which we complied with, they did not concern any of the things you pointed out."
I do not know what it mean, but its me worry about os4 version before (Deniil help me with some mui arts, and Thore help with mui itself). Maybe they just contact with deadwood (aros port) and that all ?
Quote:
My worry is that he considers something like "non-blocking network connections" to be central the user experience, but that is dependant upon a "Sub library" (curl) which is stated to be "out of the scope of this bounty". That risks Kas1e/etc saying they can't meet Fab's unstated "user experience" requirements because he doesn't agree to giving the Curl source code, and Fab saying it's not his fault when it was stated in the bounty, hence both sides blaming the other.
Fab already give me sources of threaded curl, bugs if in the Roadshow (olaf noted about), but just Fab maybe already add something new to it, or fix something: all latest stuff will be good to have, on all latest version of libcurl, with all latest makefiles and whatever fab have for.
Quote:
WORSE, what happens if the AmigaOS4 port meets his definition of "user experience and feature sets" but the AROS port doesn't? Does that mean he will refuse to "share all source code" with anyone (neither OS4 nor AROS porters)?
:)) Right. That pretty strange phrase as well.
Quote:
These sort of problems should have been sorted out BEFORE the bounty was created, which means getting agreement from everyone involved.
I told the same, but "Please note that the bounty terms have also been reviewed by other developers who have worked on Odyssey ports. While we did receive suggestions for modifications, which we complied with, they did not concern any of the things you pointed out." :)
So far this Odyssey port bounty feels like a slow-motion train-wreck: We can see exactly what is going to go wrong, but can't do anything to stop it from happening .
I hope that who-ever created the bounty will listen to such concerns, and revise the bounty as necessary (after appropriate discussions).
Until that happens, I think AmigaOS4 users (and AROS users for that matter) should avoid putting money towards the bounty. Just my humble opinion of course.
@kas1e Please note that I said something like non-blocking network connections. i.e. Similar issues with other (non-curl) 3rd-party libs.
Until that happens, I think AmigaOS4 users (and AROS users for that matter) should avoid putting money towards the bounty. Just my humble opinion of course.
sadly i think you're right on this. this is a huge amount of money (well, for our community anyway), and if future versions are not guaranteed, it seems a bit much to me. additionally the idea that dependencies with code changes will not be opened either sounds like someone just trying to make things more difficult.
-- eliyahu
"Physical reality is consistent with universal laws. When the laws do not operate, there is no reality. All of this is unreal."