1. there is no support for plugins in os4 port at all 2. plugins itself should be recompiled for os4 (as its just native library) 3. i do not have in my hands sources of that plugin. If fab will share his sources of plugin later i then can worry about porting of plugin's support and plugin itself.
@Fab
Maybe you can also release sources of swfdec plugin as well when will release main src ? Sure its out of bounty, but worth of ask
The port is created by MUI through the application base name, not by odyssey explicitely. If you want it fixed, fix it in MUI.
Okay if it's MUI issue I will make a BZ against MUI
Quote:
That said, i don't see much point in still using RA OWB, but whatever. :)
It's not about using both OWBS but making sure duplicate port names aren't being created. If it's a MUI bug it'll come back to b ite some other app another time.
BTW when you start a second Odyssey do you get a whole new app or does it behave like AWeb and pass a message to the existing instance to open another window? Could be important for the MUI bug report.
It's not about using both OWBS but making sure duplicate port names aren't being created. If it's a MUI bug it'll come back to b ite some other app another time.
Any change to MUI would really be a workaround for a design shortcoming in Amiga ARexx. ARexx assumes that each application will use a UNIQUE port basename. Even if MUI checks for existing port numbers, it won't solve the problem. Kas1e's post that started this discussion points out that OpenURL didn't work because the wrong ARexx command was used for Odessey. Even if MUI adds a UNIQUE number to the port basename, OpenURL can end up sending the wrong command to the active browser. If Odyssey is opened first and has a port name of OWB.1, OpenURL has no way of knowing that it is sending the wrong command (OPENURL) if the OpenURL prefs were set up for RA-OWB.
The only way for OpenURL to send the correct command to RA-OWB and Odyssey is to set different ports for each browser in the OpenURL prefs. If the user is running RA-OWB and Odyssey at different times, the only way of assuring that OpenURL will send correct commands to the open browser is to have UNIQUE ARexx port basenames.
At this point it's too late to change the way ARexx assigns ports. However, Commodore should have set up an ARexx port registry to avoid these port name conflicts. Just out of curiosity, will SketchBlock scripts work correctly if someone releases a popular program with the ARexx port name "SKETCHBLOCK".
Edited by xenic on 2014/1/2 13:53:39
Amiga X1000 with 2GB memory & OS 4.1FE + Radeon HD 5450
At this point it's too late to change the way ARexx assigns ports. However, Commodore should have set up an ARexx port registry to avoid these port name conflicts.
A registry can't prevent programs from intentionally destroying data of other software. Odyssey doesn't only abuse OWB's ARexx port name but for example overwrites it's AppDir: path as well.
So i will close then your BZ for odyssey about port clasesh and in comment wrote why it can't be done as well as your solution.
That sounds like a reasonable course of action to me. As far as an AppDir: name conflicts, it's simple enough for users to rename one of their browsers. I renamed my Reaction OWB to RA-OWB when MUI-OWB was released.
Amiga X1000 with 2GB memory & OS 4.1FE + Radeon HD 5450
What about sys:Prefs/URL program? I`m using YAM and it seems that after clicking a URL in a mail while browsing, a new Odyssey session opens (it`s already running, and starts again!) with all the preexisting tabs, plus the newly clicked! Should I change any prameters in the prefs/url ? Can I choose if URL or Open URL will handle the browser (which one has precedence over the other?) ? Thanks.
As far as an AppDir: name conflicts, it's simple enough for users to rename one of their browsers. I renamed my Reaction OWB to RA-OWB when MUI-OWB was released.
I don't know anything about a Reaction OWB and if it works with that one, but you must not change the name of the executable of my AmigaOS 4.x port of OWB since for example the AmiUpdate support depends on the the correct name.
So i will close then your BZ for odyssey about port clasesh and in comment wrote why it can't be done as well as your solution.
That you don't, and don't even want to, port the MorphOS parts of Odyssey to AmigaOS but just try to get them compile somehow hoping the incompatible MorphOS code doesn't crash, or cause other problems, too much on AmigaOS is bad enough, but not even fixing such obvious and easy to change problems in your AmigaOS builds is ridiculous.
That you don't, and don't even want to, port the MorphOS parts of Odyssey to AmigaOS but just try to get them compile somehow hoping the incompatible MorphOS code doesn't crash
We already port it, and its already works and its better in compare with all the other we have (even if it bad port as you say). We already replace necessary parts to make it works on os4. I even reuse some of your old code (native mouse pointers and functions for unicode conversion), but in others its all same morphos code with simple changes (hooks, processes, mui, etc). I know you hate morphos and everything about it, but is there needs, after all those years of silence and no bringing to end users normal browsers, to bring that hate there again and saying that Fab's odyssey broken your no-more-indevelop-owb, because use the same name of binary and arexx port ?
Sure, some parts still may need some love, but morphos is not that different from os4 as you may think, its all the same, just with some little bits of differences. I assume that Timer problem about which you told me before is the last one to fix.
Sure you can do its all better, just you not do it anymore, so we do it as we can. And if you back in past not stop to works on owb, and not refuse to make gui on top of it with words "owb is not gui, sorry", then it all can be different today.
We do what we can because no one else do it. And because of that we have what we have.
Quote:
but not even fixing such obvious and easy to change problems in your AmigaOS builds is ridiculous.
Ridiculous ? Imho ridiculous is that you just can't see that because no one else work on any normal browser for os4 in accpeted timeline, we have to port one from morphos so users still will holds on os4. And there is no bugs in question, there is just name clashes, which we can't easy solve by changing names, because odyssey in develop, and scripts for it in develop. So if choice between non-in-developer app and app-in-developer, of course right logical choice is to choice one which in develop still, and not that one which imho will never be in develop anymore.
Why we should now change odyssey's binary name and arexx port to be no in clash with your owb, if odyssey in develop, and scripts for it in develop, while your one are not , and you only bring hate to morphos there with all those replies how bad it all. Even if, someday, you will worry to update webkit code in your version and it will works much more stable that my bad port, there still will be no gui, no download manager, and so it will be no normal browser just a good core. While users need right now browser which works fast and have gui, even, if it will crashes sometime.
From programming point of view you of course right, there all can be done better even for port, but for users matter to have browser right _NOW_, not after 10 years (maybe). As well as for them veeeery importnat to have browser with GUI, not just as core. GUI just must to have. Download manager just must to have. After few years you can't skip the reality: TW is not ready, your owb have no gui and outdated webkit => users like me do tasks which they do not want to do, and because none who can have time or interest in it anymore.
I will only happy if you will back on owb, and will bring in it fresh webkit as well as build at top of it gui. But till you not, we follow morphos as there is developer who do gui on top of webkit and users have modern browser. Not like we have before: good core of your owb with totally outdated webkit and no gui at all, and non finished and slow timberwolf port.
What you want in end ? I should say to fab "sorry i will change it as i want" , and so users later will cry why scripts do not works ? Or i should not change it, and so , Joerg will think that my port is very bad and "ridiculious" ? I only do logical choice : your owb not in develop, odyssey in develop, so i follow odyssey, as it bring fresh stuff for users.
Edited by kas1e on 2014/1/3 8:22:18 Edited by kas1e on 2014/1/3 8:24:42 Edited by kas1e on 2014/1/3 8:34:41 Edited by kas1e on 2014/1/3 8:37:22 Edited by kas1e on 2014/1/3 8:42:41
Just out of curiosity, will SketchBlock scripts work correctly if someone releases a popular program with the ARexx port name "SKETCHBLOCK".
It depends how the script is being run. If it is executed from within Sketchblock, then it does not need the "address sketchblock" part and will work. If it is a script run external, eg. from the Shell or another program, then it will potentially not work as the commands may go to the "wrong" Sketchblock.
It's up to the user to ensure they are addressing the correct ARexx port. It's a more serious issue that duplicate port names are created, than programs using the same basename (the latter, whilst annoying, can be worked around).
SketchBlock will not start if it's ARexx port name is used, it will simply send a command to the existing port to open any new file specified on the command line / workbench in the existing app.
Your right about setting the address when starting a script via the application but does it set the port, or the name of the address, if the latter then you are just as messed up if a duplicate port name exists. Depnding on the search / creation order.
@thread
BTW I duplicated the MUI port slot number bug with wookiechat and a python script using arexx module to occupy slot name. Bug report written.
I'm not back, I just logged in to remove a dead link from my profile and saw this usual crap, obviously nothing has changed in the last 3 years ..., and this will very likely be my last post here.
Quote:
A question about your OWB port, is there any future update planned ?
Nothing changed, the sand-labs/pleyo OWB project is still dead, since about 5 years already, and even all servers incl. the SVN with the sources are offline since several years. It doesn't make any sense to work on such an outdated browser. The only exception would be if someone ports gstreamer (used by OWB for HTML5 video, etc.) or some NPAPI plugins to AmigaOS, in this case I'd rebuild the last version with these features enabled, but since that didn't happen in the last 6 years it probably never will.
In case both OWB and Timeberwolf get too old and don't work with important sites I'm using regularly any more I'll probably port another browser like Chromium to AmigaOS, or help updating Timberwolf, but in any case I'll keep the results for myself and wont make the mistake of releasing something like that to this AmigaOS "community" again, OWB was definitely the last one.
Quote:
And if not, do you have the intention to release the latest source code ?
This question doesn't make any sense OWB was always open source, intentionally, and of course the AmigaOS parts were one the sand-labs SVN server as well (some bug fixes in platform independent parts were added by the author of the AROS OWB port, but IIRC he didn't submit the AROS specific parts). With the last release(es) the sand-labs OWB project was dead already but I've send the sources of OWB 3.32 by E-Mail instead to about 5 people myself, and very likely much more got it from them indirectly.
I could have kept the sources for myself by using shared objects like in the OWB Blastoise port (AmigaOS OWB versions 1.x) but additionally using shared objects for the OWB libraries (libJSC.so, libWebCore.so, etc.) as well, only that way it would have been possible at all to only release the sources of LGPL parts like libJSC.so, but the ports of OWB Dodou, Galekid, Galeking, etc. (AmigaOS OWB versions 2.x and 3.x) were statically linked and therefore not releasing the complete sources wouldn't even have been possible.
Nothing changed, the sand-labs/pleyo OWB project is still dead, since about 5 years already, and even all servers incl. the SVN with the sources are offline since several years. It doesn't make any sense to work on such an outdated browser.
I remember that when the original project was alive Sand-Labs did their merge with WebKit almost regularity, even if they didn't work on it anymore eventually you are free to continue to do such merge alone and so mantain the project alive, really i don't see where is the problem from your side ..
Quote:
In case both OWB and Timeberwolf get too old and don't work with important sites I'm using regularly any more I'll probably port another browser like Chromium to AmigaOS, or help updating Timberwolf
Maybe you could at least fix the latest important problems ?
- The displaying source are no more availible since you disable in 3.32 to avoid another problem, maybe you can fix such problem and so re-enabled also the displaying source ?
- Again, since latest 3.32 we have some blocks that prevent the closure of the program, but also of the tabs When this happen OWB will stay freezed for many seconds
Quote:
With the last release(es) the sand-labs OWB project was dead already but I've send the sources of OWB 3.32 by E-Mail instead to about 5 people myself, and very likely much more got it from them indirectly.
I can assure you that such code you sent privately was never shared with us nor published (atleast i never found it), perhaps if you decide to don't work on it anymore can you upload it somewhere, maybe on Aminet/OS4Depot ?
Beside OWB this code might be usefull also for other related project
joerg, what's up? why are you so hurt? you come out as as sore kid. pull yourself together. odyssey is a better effort than your owb port. live with it.
odyssey is a better effort than your owb port. live with it.
Maybe that line should read: odyssey is a further developed project than your owb port.
Let's give credit where credit is due. I was about to give up on WEB browsing with OS4 and switch to a MAC when OWB was first released. I was finally able to access my accounts and important WEB sites and am now still using OS4. Reaction OWB was an important step in modernizing NG Amiga WEB access.
Amiga X1000 with 2GB memory & OS 4.1FE + Radeon HD 5450
News feed
OS4 Feedback forum
OS4Depot Feedback forum
Amigabounty forum
OpenAmiga forum
AmiCygnix forum
ABC shell forum
AmiKit forum
Cinnamon Writer forum
CodeBench forum
Digital Universe forum
Dopus 5 forum
E-UAE forum
Gnash forum
Ibrowse forum
JAmiga forum
Odyssey forum
OWB forum
Qt forum
SmartFileSystem forum
Timberwolf forum
TouchDevice forum
TuneNet forum
Unsatisfactory Software forum
