Any news from Neuss
  • Acolyte of the Butterfly
    Acolyte of the Butterfly
    Sprocki
    Posts: 113 from 2005/2/23
    From: Berlin - Germany
    BTW: to those who remember the PegasosPPC shirt, I was happy to collect almost all signatures from the devs team at the show as I have never seen so many of them in one place before. Just ignore the bad quality of my camera, please.

    EDIT: the picture upload here is broken. I attached the picture with correct extension, type and file size but got the reply:

    "Error: 1: /var/www/morph.zone/modules/newbb_plus/cache/attachments/ does not exist or is not writable.


    If the page does not automatically reload, please click here"

    So I uploaded it to: http://www.directupload.net/file/d/4141/lzif8frt_jpg.htm
  • »15.10.15 - 21:55
    Profile
  • Paladin of the Pegasos
    Paladin of the Pegasos
    redrumloa
    Posts: 1391 from 2003/4/13
    Quote:

    Sprocki wrote:
    BTW: to those who remember the PegasosPPC shirt, I was happy to collect almost all signatures from the devs team at the show as I have never seen so many of them in one place before. Just ignore the bad quality of my camera, please.

    EDIT: the picture upload here is broken. I attached the picture with correct extension, type and file size but got the reply:

    "Error: 1: /var/www/morph.zone/modules/newbb_plus/cache/attachments/ does not exist or is not writable.


    If the page does not automatically reload, please click here"

    So I uploaded it to: http://www.directupload.net/file/d/4141/lzif8frt_jpg.htm


    That is awesome! I had one when I got my Peg1, but I wore it until it wore out and got holes. I wore it locally, where nobody had a clue what it meant :-/
  • »16.10.15 - 00:03
    Profile
  • Paladin of the Pegasos
    Paladin of the Pegasos
    Yasu
    Posts: 1724 from 2012/3/22
    From: Stockholm, Sweden
    We should have our own official MorphOS merchandice online store. I could do that and donate the surplus back to the community if I get enough funds :-)

    If it's at all of interest of course.
    AMIGA FORUM - Hela Sveriges Amigatidning!
    AMIGA FORUM - Sweden's Amiga Magazine!

    My MorphOS blog
  • »16.10.15 - 10:54
    Profile Visit Website
  • Yokemate of Keyboards
    Yokemate of Keyboards
    Andreas_Wolf
    Posts: 10923 from 2003/5/22
    From: Germany
    > We should have our own official MorphOS merchandice online store.

    At least an inofficial one there is already:

    https://morph.zone/modules/newbb_plus/viewtopic.php?topic_id=9447&forum=3&start=4 :-)
  • »16.10.15 - 11:29
    Profile
  • Paladin of the Pegasos
    Paladin of the Pegasos
    Yasu
    Posts: 1724 from 2012/3/22
    From: Stockholm, Sweden
    Yes, I know about those but I think it's not the same as a officially sanctioned store. But it's better than nothing of course :-)
    AMIGA FORUM - Hela Sveriges Amigatidning!
    AMIGA FORUM - Sweden's Amiga Magazine!

    My MorphOS blog
  • »16.10.15 - 12:24
    Profile Visit Website
  • Order of the Butterfly
    Order of the Butterfly
    ppcamiga1
    Posts: 202 from 2015/8/23
    So, very little progress on MOS NG since 2011?
    I have simple solution how to make MOS for x86 leading amiga os on x86 in less than falf a year.
    Amiga OS for x86 should be just open source port of boopsi, MUI, datatypes to unix.
    Rest of api should be droped.
    boopsi, datatypes from MOS ale already open source.
    Make MUI avaible under LGPL and everbody will be using it.
  • »17.10.15 - 08:41
    Profile
  • ASiegel
    Posts: 1214 from 2003/2/15
    From: Central Europe
    Quote:

    Yasu wrote:
    We should have our own official MorphOS merchandice online store. I could do that and donate the surplus back to the community if I get enough funds :-)

    If it's at all of interest of course.

    I have looked into this a bit but I am not sure what the best provider is just yet. Ideally, the merchandising company would offer local production in Europe as well as Northern America to keep shipping costs down.

    As far as offline stores are concerned, there has already been some progess: AmigaDave will have some exclusive MorphOS polo shirts available for sale at AmiWest this weekend :-)
  • »17.10.15 - 09:19
    Profile
  • Order of the Butterfly
    Order of the Butterfly
    ppcamiga1
    Posts: 202 from 2015/8/23
    Quote:

    Andreas_Wolf wrote:
    Why should Reggae be deprecated?



    Because is only for MOS i thats why not widely accepted and used.

    Quote:


    http://www.oocities.org/siliconvalley/park/3883/zune/ ;-)



    Zune after many years is still not compatybile with mui.



    [ Edited by ppcamiga1 17.10.2015 - 09:17 ]
  • »17.10.15 - 10:16
    Profile
  • Yokemate of Keyboards
    Yokemate of Keyboards
    Andreas_Wolf
    Posts: 10923 from 2003/5/22
    From: Germany
    >> Why should Reggae be deprecated?

    > Because is only for MOS i thats why not widely accepted and used.

    Then why not suggest to open source it (as you do with MUI) so that it can also be used elsewhere? Reggae is clearly superior to the old datatypes concept.
  • »17.10.15 - 10:54
    Profile
  • Paladin of the Pegasos
    Paladin of the Pegasos
    Yasu
    Posts: 1724 from 2012/3/22
    From: Stockholm, Sweden
    Why does people keep assuming open source equals sucess? It hasn't helped AROS or the other small hobby OS's. Hell, it hasn't even helped Linux in the desktop market. All it has done there is to make every whim turn into yet another incompatible distro. Even AROS got forked with ARIX.

    People still has to do the work, and someone has to force people to pull in the same direction. Open source doesn't solve the first part, and makes the second part pretty damn hard.

    @Amigadave

    You got pictures of that sweater? :-)

    [ Edited by Yasu 17.10.2015 - 11:55 ]
    AMIGA FORUM - Hela Sveriges Amigatidning!
    AMIGA FORUM - Sweden's Amiga Magazine!

    My MorphOS blog
  • »17.10.15 - 11:54
    Profile Visit Website
  • MorphOS Developer
    geit
    Posts: 936 from 2004/9/23
    Quote:

    ppcamiga1 wrote:
    Make MUI avaible under LGPL and everbody will be using it.



    After what "they" have done with their MUI3.999999 stuff, it is already clear enough to be sure that keeping the stuff private prevents a gigantic mess.

    OpenSource only causes incompatible and uncontrollable forks. And we already have that with MUI.

    Speaking about Datatypes: They are mostly broken by design and Reggae is so much easier and flexible in any way.
  • »17.10.15 - 12:16
    Profile
  • Yokemate of Keyboards
    Yokemate of Keyboards
    Andreas_Wolf
    Posts: 10923 from 2003/5/22
    From: Germany
    > what "they" have done with their MUI3.999999 stuff

    AFAIK it is based on a 2006 MUI4.0, not on MUI3.9.

    > keeping the stuff private prevents a gigantic mess.

    Well, they wouldn't have to reimplement things, which in turn would prevent errors emerging from erroneous reimplementation.
  • »17.10.15 - 12:28
    Profile
  • MorphOS Developer
    geit
    Posts: 936 from 2004/9/23
    Quote:

    Andreas_Wolf wrote:

    > keeping the stuff private prevents a gigantic mess.

    Well, they wouldn't have to reimplement things, which in turn would prevent errors emerging from erroneous reimplementation.


    I am talking about wildly added features on their side, which make no sense if you want to stay compatible in the first place.
  • »17.10.15 - 13:33
    Profile
  • Moderator
    guruman
    Posts: 455 from 2003/7/22
    Ok, all fine and nice, but some more concrete news? I mean, newer gfx boards supported (or something like this) surely does not interest just me? :P Other updates/advancements?

    @ASiegel
    I know, but this time I really could not make it to Neuss... ;)

    Kind regards,
    Andrea
  • »17.10.15 - 14:16
    Profile
  • Order of the Butterfly
    Order of the Butterfly
    Tcheko
    Posts: 436 from 2003/2/25
    From: France
    Quote:

    guruman wrote:
    Ok, all fine and nice, but some more concrete news? I mean, newer gfx boards supported (or something like this) surely does not interest just me? :P Other updates/advancements?

    @ASiegel
    I know, but this time I really could not make it to Neuss... ;)

    Kind regards,
    Andrea


    The team never talks about planned features. We don't even know ourself what will be in the ISO in the end... Some stuff can be simply removed just before the freeze date because it is not enough polished.

    To make an idea for yourself, read previous release notes and expect roughly the same: new supported hardware (motherboard, graphics cards, or other kind of devices), improved performances here and there, new tools/apps/utilities, tons of bugfixes...
    Quelque soit le chemin que tu prendras dans la vie, sache que tu auras des ampoules aux pieds.
    -------
    I need to practice my Kung Fu.
  • »17.10.15 - 20:11
    Profile Visit Website
  • Butterfly
    Butterfly
    Georg
    Posts: 62 from 2004/4/7
    Quote:

    geit wrote:
    OpenSource only causes incompatible and uncontrollable forks. And we already have that with MUI.



    Of all the OS components nothing is nowhere near as horrible as MUI for creating compatible reimplementations because the autodocs and dev docs are nowhere near good enough. Some things are not documented at all (TAG_IGNORE, special treatment/filtering/... of attributes/methods in certain classes, ...). Some things are documented wrongly (attributes which know OM_GET).

    The "real/true" MUI "owners" often give the impression that given a specific API and (auto)docs there is only exactly one possible way to create a reimplementation that "matches" the docs/API. So just follow the docs/API headers and you end up with compatible code. Not so with MUI. From one and the same MUI doc/API you can create countless variations which all may result in slightly or hugely different behaviour, even if all of them follow 100 % the info from the doc and the API.

    There are also some buggy things in MUI itself. Like MUIA_ShowMe. To be compatible you are not even allowed to make a non-buggy version of this. You would need to reproduce the exact same buggy behaviour to be really compatible. Great ...

    And btw, didn't MUI4 introduce it's own share of incompatibilities with MUI38? (list class?)
  • »18.10.15 - 08:21
    Profile
  • MorphOS Developer
    itix
    Posts: 1516 from 2003/2/24
    From: Finland
    @Georg

    The biggest problem is that many programs abused undocumented features, methods, API calls and internal structures. Not only Stuntzi had an access to sources and IBrowse, ListTree etc had their own peculiar way to peek and poke undocumented stuff.

    MUI 4 removed lot of internal stuff like external support libraries. List class was streamlined but IIRC it didn't have major incompatibility issues. I know there is a render quirk in AmIRC lists I should look at some day. I know that the in Zune you are going to have major issue because some old software depends on internal structures you can't know about.

    Another problem is of course what you mentioned. MUI design is not perfect and documentation is lacking on many post-V19 stuff. However, stuff like TAG_IGNORE was well known in 90s but all this information has disappeared over time.

    [ Edited by itix 18.10.2015 - 10:20 ]
    1 + 1 = 3 with very large values of 1
  • »18.10.15 - 08:51
    Profile
  • Order of the Butterfly
    Order of the Butterfly
    ppcamiga1
    Posts: 202 from 2015/8/23
    Quote:

    Yasu wrote:
    Why does people keep assuming open source equals sucess? It hasn't helped AROS or the other small hobby OS's.



    AROS developers do not care about tests with 68k software.
    Open source do not help AROS because AROS was untested crap.

    Oryginal MUI is different case.
  • »18.10.15 - 17:03
    Profile
  • Order of the Butterfly
    Order of the Butterfly
    ppcamiga1
    Posts: 202 from 2015/8/23
    Quote:

    geit wrote:
    After what "they" have done with their MUI3.999999 stuff, it is already clear enough to be sure that keeping the stuff private prevents a gigantic mess.



    MUI3.999999 is not compatybile with mui 4.0 because Amiga Os 4 developers have to implement things already done in mui 4.0.

    If they have mui 4.0 on LGPL, mui be the same on all amiga os platforms.
  • »18.10.15 - 17:06
    Profile
  • Moderator
    Kronos
    Posts: 1978 from 2003/2/24
    Giving their track-record with opensourced system components they would use an LGPL MUI to:

    - create an OS4 only incompatible version claiming that they don't have to give back sources misquoting an obsure licence clause

    - release updates for all systems with borked MorphOS binaries and then making sure that opensource MUI-apps require that exact version



    eh..... no thanks !
    --------------------- May the 4th be with you ------------------
    Mother Russia dance of the Zar, don't you know how lucky you are
  • »18.10.15 - 17:26
    Profile
  • MorphOS Developer
    jacadcaps
    Posts: 2005 from 2003/3/5
    From: Canada
    Quote:

    Georg wrote:
    Some things are not documented at all (TAG_IGNORE, special treatment/filtering/... of attributes/methods in certain classes, ...). Some things are documented wrongly (attributes which know OM_GET).



    That's called the "know how" ;) From a MUI application programmer's POV you really don't need to know how those internals work, do you?

    Quote:

    Georg wrote:
    There are also some buggy things in MUI itself. Like MUIA_ShowMe. To be compatible you are not even allowed to make a non-buggy version of this. You would need to reproduce the exact same buggy behaviour to be really compatible. Great ...



    MUIA_ShowMe was actually fixed in MUI4. Still, a lot of internal hickups do come from the fact that certain applications; namely Vaporware (does anyone remember Voyager?) and, even worse, IBrowse do rely on many MUI internals. That's not just internal methods/attrs, but the layout of object data structures. Trying to get those legacy (cr)apps to continue working with a modern MUI was a challenge and the result is that some stuff may be hard for you to understand just by testing/reading docs. Especially since MUI does have a lot of code that actually works around bugs in applications (IBrowse ;) ...
  • »18.10.15 - 21:43
    Profile Visit Website
  • MorphOS Developer
    jacadcaps
    Posts: 2005 from 2003/3/5
    From: Canada
    Quote:

    ppcamiga1 wrote:
    If they have mui 4.0 on LGPL, mui be the same on all amiga os platforms.



    Except MorphOS wouldn't boot with such a MUI build :)
  • »18.10.15 - 21:46
    Profile Visit Website
  • Yokemate of Keyboards
    Yokemate of Keyboards
    Andreas_Wolf
    Posts: 10923 from 2003/5/22
    From: Germany
    > If they have mui 4.0 on LGPL, mui be the same on all amiga os platforms.

    No, as MorphOS MUI went from 4.0 over 4.1 and 4.2 to current version 5 (since MorphOS 3.8).
  • »18.10.15 - 23:26
    Profile
  • Butterfly
    Butterfly
    Georg
    Posts: 62 from 2004/4/7
    Quote:

    jacadcaps wrote:
    That's called the "know how" ;) From a MUI application programmer's POV you really don't need to know how those internals work, do you?



    The MUI app coder can and will fiddle and shuffle around his code until it does what he wants. May work and be perfectly fine/bug free on his system but he may still have created illegal code if he does not know some of this internal stuff. Will only ever show up with different version of MUI and/or external MUI classes.

    For example he may do:

    Code:

    struct TagItem tags[] = {MUIA_Something, bla, TAG_DONE};

    SetAttrsA(extclassobj, tags);
    SetAttrsA(extclassobj2, tags);


    Some time later a new version 2 of extclass is released which changed mSet method to do
    Code:

    if (val != oldval)
    {
    ...
    }
    else
    {
    tag->ti_Tag = TAG_IGNORE;
    }


    (which it did not do in version 1) and with this the app all of the sudden becomes buggy.

    MUI contains far too few info on what things apps/classes may do and may not do. If doing certain things or settings certain attributes in methods is allowed or not. Resulted in some weird stuff out there. For example texteditor mcc to be able to draw inside handleevent code does some trickery by calling MUI_Redraw() and flag'ing itself that what it really want to do in the following MUIM_Draw is to do the handleevent stuff. What then happens is that you have code during/inside MUIM_Draw which normally isn't there. Like OM_SETs. Those may trigger notifications. Which may trigger rerendering of other objects. But what if the texteditor obj is inside a virtual group and MUI set up a clip region to restrict rendering to visible parts of virtual group. And that other notified object is outside that clip region. Depending on MUI version or MUI clone maybe behaviour differs. One version may do some optimization and drop clipping of the texteditor obj to virtgroup, if texteditor obj is fully visible within virtgroup. Another maybe future version may decide that it would make sense to always clip every drawing to object's boundaries (even if they are no virtgroups involved), because most GUI toolkits on other systems do so, too?
  • »19.10.15 - 07:56
    Profile