AmigaOS 3.9 BoingBag #4

[Gulliver]

AmigaOS 3.9 BoingBag #4 version 0.2b
=======================

NO WARRANTIES: USE AT YOUR OWN RISK!

The unofficial BoingBag 4 project is a community effort that is pursuing the considerable
task of upgrading, enhancing and bugfixing AmigaOS 3.9.

This is a preliminary release. It is anticipated that further updates
will be developed and incorporated before the final release. Please
report any bugs immediately.

To get the latest updates and information, visit http://lilliput.amiga-projects.net/bbag4.htm

You will need to install Boing Bags #1, #2 and #3 before installing BoingBag 4.

BoingBag 1 and 2 are available from http://os.amigaworld.de/index.php?lang=en&page=12
BoingBag 3 is available from http://lilliput.amiga-projects.net/bbag3isdone.htm
BoingBag 4 is available from http://lilliput.amiga-projects.net/bbag4.htm

__________________________________________________________________________________________

These components are updated:

                         BB3           BB4
graphics.library        40.24      -> 40.25b8
icon.ibrary             46.4.129   -> 46.4.159
SetPatch                44.38      -> 44.39
version.library         45.4       -> 45.5
wizard.library          42.100     -> 45.1

These components are new:

Canon BJC6000 printer driver          44.7
Epson Stylus Colour printer driver    45.0
SuperPlus monitor driver              40.2

Also, various components are replaced with their 68020-optimized
equivalents (generated with the PatchFor020 utility).

Thanks to all the developers who made this upgrade possible by their
continued support for OS3.9, and to all the beta testers of this package.


Changelog
---------

Canon BJC6000
    Includes limited support for photo printhead, and text mode printing.

Epson Stylus Colour
    It should work with the Stylus S20, D120 or Cx0 series which also
    support 1440 dpi.

icon.library 46.4.154
    The ProcessIcon bug and an enforcer zeropage read hit are fixed now.
    The following functions in the icon.library were broken:
    OptimizeImageSpace(), DropChunkyIconImage(), DropNewIconToolTypes(),
    OnlyUpdatePosition(), PreserveOldIconImages(). In some cases the
    planar images of non-V44 icons couldn't be displayed. This version
    works without conflicts under AfA_OS if the version of the
    AFA_OS_LIBS:Icon_Lib.exe is from 2010 or newer. More fixes and tweaks.

graphics.library 40.25b8
    Lots of fixes, optimizations and enhancements. You are now able to
    remove forever these patches:
        * AmyWarp
        * BlazeWCP
        * CPUBlit
        * FBlit
        * FText
        * IconBeFast
        * SetPatch 44.38 (R_WriteChunkyPixel and R_WritePixelLine8
          patches).
    They are now built in.

SetPatch 44.39
    Disabled R_WriteChunkyPixel and R_WritePixelLine8.

SuperPlus monitor driver 40.2
    An ECS/AGA driver with 800x600 70ns mode, about 70% faster than
    Super72 in 256 colours. It can provide support for ECS at 800x600 in
    16 colors out of 4096, and is also interesting for the AGA chipset,
    because it uses the "HiRes" mode with 70ns instead "SuperHiRes" with
    35ns pixel clock. The display DMA is not as heavily loaded as it is
    with Super72 and display operations are faster.

wizard.library 45.1
    Provides a nicer MUI/Zune look and is optimized for 68020 and up.

[kickstart]

I like the new graphics.library, good work.

[NovaCoder]

Hiya,

Is all of these updates considered stable 'enough' for a release?

Good work BTW

[XDelusion]

Oh great thanks!!!

[ognix]

Hats off to Cosmos especially for his work on graphics.library (since its importance).
Of course many thanks even to all the others who contribute!  

[Amigaz]

ohhh...nice! and with install script

*thumbs up!*

[recidivist]

Downloaded,thanks for your work.

[x303]

Under winuae, loadmodule11 doesn't load the graphics.library if the 'fast' option is set. Without this (and check) option, graphics.library is loaded.

:afro:

[kolla]

Also, various components are replaced with their 68020-optimized
equivalents (generated with the PatchFor020 utility).

Really? I mean, all I see are files that are exactly as they were before, now only with filenames in uppercase. Seems quite pointless.

[kolla]

Another thing... even with the new graphics.library I save ~160k of chipram by using fblit on my A1200.

[Gulliver]

Another thing... even with the new graphics.library I save ~160k of chipram by using fblit on my A1200.

The goal for graphics.library 40.25 was to be better than 40.24. If you find room for further improvements, why dont you drop your comments on the corresponding thread, so that the guy which is actually enhancing this library, Cosmos, improves it further? He is still looking for suggestions. Maybe in that way you can tell him with greater detail what doesnt seem to work as good as you expect to.

And I am sorry that uppercase filenames bother you.
I will see that at some point we convert them back to lowercase.

Thank you for your feedback.

[Karlos]

Another thing... even with the new graphics.library I save ~160k of chipram by using fblit on my A1200.

Now that you mentioned it, I took a closer look at my (fairly) vanilla 3.9/CGX install, which you can see a shot of here.

I didn't think about it before, but looking at that, the largest free block of chip ram I have is 2,036,624 bytes. That means I'm only using about 60K of chip ram to start with.

RTG FTW

[kolla]

The goal for graphics.library 40.25 was to be better than 40.24. If you find room for further improvements, why dont you drop your comments on the corresponding thread, so that the guy which is actually enhancing this library, Cosmos, improves it further? He is still looking for suggestions. Maybe in that way you can tell him with greater detail what doesnt seem to work as good as you expect to.

Hm, OK. I haven't really been paying much attention to what graphics.library 40.25 is meant to do, I'm just commenting on the "you can now forever remove" list. Appearantly FBlit still makes a rather big difference on AGA at least.

And I am sorry that uppercase filenames bother you.
I will see that at some point we convert them back to lowercase.

Great. I'm still pussled about the optimization though, the new files are exactly same size, but different md5sums. I would have expected small differences in filesize as well.

[Karlos]

Great. I'm still pussled about the optimization though, the new files are exactly same size, but different md5sums. I would have expected small differences in filesize as well.

Doesn't take much really. After all, even if there is one byte of data in the file that is different, the size isn't going to change but the md5 hash will.

[kolla]

Doesn't take much really. After all, even if there is one byte of data in the file that is different, the size isn't going to change but the md5 hash will.

Yes, Piru just explained it for me.

I must say though, I dont really like it, would be so much better if the version string could also be patched to reflect the changes done. Now we end up with lots of binaries that at first glance are the same, but yet are not. And god forbid you try to use one of those files from a BB4 system on a 68000 system. And for what gain, really? Optimized MEmacs? WTF? :laughing: