GitLab

During BootRun, see if !Scrap is in the root of the boot drive.
During SetChoices, see if !Choices is in the root of the boot drive.

Version 1.31. Tagged as 'DesktopBoot-1_31'

The only part of !Boot which is TCPIPLibs related is BootVars. However, BootVars is run before we get chance to load a newer Internet module, so must run time select which to support.
Fortunately the member wanted from the returned union from socketioctl is in the same place in both BSD4.3 and BSD4.4 so all that's needed is to select the appropriate command.
Also, BootVar's local makefile rewritten to use shared makefiles.

Remove Philips prefix from Philips monitors in Philips directory (!ScrnSetup uses the monitor_title field). Fixes truncation on 10 letter FS.
Ref: https://www.riscosopen.org/forum/forums/11/topics/15342

Version 1.30. Tagged as 'DesktopBoot-1_30'

Detail:
  The NoVRAM obey files for 370/400/430 hooks actually attempted to select a higher resolution than their VRAM counterparts, as though transposed. Copy them from 360 hook.
Admin:
  Spotted during Window Manager regression testing.

Version 1.29. Tagged as 'DesktopBoot-1_29'

Detail:
  PreDesktop obey files updated to load missing sprites where required.
Admin:
  Requires LoadWimp-0_12.

Version 1.28. Tagged as 'DesktopBoot-1_28'

But only needed pre 3.50, so can live in 310 hook.

Version 1.27. Tagged as 'Boot-1_27'

Detail:
  This set of changes adds a "disc error 20" fix to ADFS for RISC OS 3.50 thru 4.02, using similar code to that used for the fix that was applied to ADFS itself (in ADFS 3.54). This is to resolve issues with modern IDE devices which don't adhere to the DRQ timeout that old versions of the ATA spec mandated.
  File changes (all within Source/ROMPatch):
  - patches/s/ADFS354 - Generic assembler source for the disc error 20 fix
  - ihf2c/c/ihf2c - Source for a C utility that was used to help produce the ROM patch data
  - patches/adfs354,feb - Utility script for building the different versions of the ADFS patch and converting to ROMPatch form
  - patches/350/h/ADFSpatch, patches/360/h/ADFSpatch, patches/370/h/ADFSpatch, patches/371/h/ADFSpatch, patches/402/h/ADFSpatch - Added/updated ADFS patches
  - patches/350/h/patch, patches/371/h/patch, patches/402/h/patch - Include new ADFSpatch files where relevant
  - UK/h/messages - Update ROMPatch version numbers
  - c/rompatch, s/module - Add a word of writable workspace to the module which the ADFS patch can use
  - patches/402/h/CDSFTpatch - Remove duplicate patch (duplicate of RESLVpatch)
  - patches/402/h/MSGSpatch - Document the patches
  - ReadMe, Install/ROMPatch/!RunImage,ff8 - Remove redundant/outdated files (the main Boot makefile will build and install ROMPatch for us)
Admin:
  Tested on a RISC OS 3.70 StrongARM RiscPC with a CF-IDE adapter, disc error 20 is no more
  3.50/3.60/3.70/3.71/4.02 also tested under RPCEmu, manually provoking the different code paths to check hook points are correct


Version 1.26. Tagged as 'Boot-1_26'

The extra AddToRMA in BootRun is no longer needed, the true cause was a bug fixed in Wimp-5_55 reading IRQsema's address, which caused the RMA to be clamped.
Tested on a real A5000.

Version 1.25. Tagged as 'Boot-1_25'

A handful of MDFs had opening 'startmode' but closing 'EndMode', which MakeModes complains about. Bulk edit these to 'endmode'.

Version 1.24. Tagged as 'Boot-1_24'

Correction to default run path !Boot was adding its private utilities (!Boot.Utils) to the global run path, which isn't the intent of that directory. Changed singular use of HOn to use a fully qualified path to !Boot.Utils.

User commands should go in !Boot.Library, as before.

Version 1.23. Tagged as 'Boot-1_23'

Bulk MDF shuffle 45 of the provided MDFs used names that are ambiguous when copied onto a 10 letter long drive. Since the MDFs are already grouped in directories corresponding to the manufacturer name there's no need to prefix the filename with that same text. Removing the prefix gets them all under 10 characters.

Removed duplicate ViewSonic VA2012W and VA2012Wseries MDF.

Version 1.22. Tagged as 'Boot-1_22'

BootVars was not accounting for a locked drive (with FSLock) when working out whether to set Boot$ReadOnly.
This meant that a locked drive would unhelpfully create a small RAM disc with a default setup in it rather than running the normal boot sequence (albeit locked against changes to everything except $.Public and !Scrap).
Tested on RISC OS 4.02.

Version 1.21. Tagged as 'Boot-1_21'

Detail:
  Source/ClrMonitor/c/main - On CMOS reset, zap both lines in the monitor configuration file instead of just the LoadModeFile line
  Will prevent the boot sequence terminating early after a CMOS reset when running Wimp 5.40, under which WimpMode will correctly error when given a bad mode
Admin:
  Tested on Raspberry Pi


Version 1.20. Tagged as 'Boot-1_20'

After setting BootResources$Things, Choices$Things, Boot$Things a new service call is sent to allow, for example, ROM based modules to react to any implications of that change.
R0 is defined as flags, currently all 0, might be useful one day.
Tested briefly.

Version 1.19. Tagged as 'Boot-1_19'

Before I forget what it all means.
Not tagged.

The ROM patcher relied on lots of grubby internal knowledge of kernel addresses, some of which moved for Ursula
* L1PT now discovered with OS_ReadSysInfo 6
* CAM now discovered with OS_ReadSysInfo 6
* The StrongARM cleaner flipflop now discovered with OS_ReadSysInfo 6
* Option to ignore the ROM checksum where it is known to be duff (eg. personalised copies of RISC OS 4)

Also
* Remove RO430Hook from the output pending more testing
* Readd SoundDMA for RISC OS 3.60


Version 1.18. Tagged as 'Boot-1_18'

Retagged as Boot-1_17.

When softloading the Wimp pre RISC OS 3.50 we must kill the BorderUtils module (which gets loaded as a side effect of the Toolbox being loaded when !System is run). While BorderUtils is careful not to fiddle with the workspace of Wimps it doesn't recognise this check is only made when it loads (which passes as the ROM Wimp is acceptable) but subsequent softloading is missed. Generally bad things happen when it pokes the Wimps workspace on every mode change.

PreDesktop scripts were trying to fix up malformed Boot$OSVersions, but since we know we get that right those steps are redundant.

In BootRun the use of AddToRMA hadn't spotted a subtle syntax change.
The AddToRMA utility ignored a trailing 'K' and assumed the number was always in kilobytes. AddToRMA in BootCommands requires a K (or M) suffix otherwise it assumes bytes. So for a number of years only 768 bytes have been added to the RMA, which isn't enough to softload a Wimp on RISC OS 3.10.
As RISC OS 3.10 has no free pool, the AppSize command drains the RMA dry. We push 256K back into it for any subsequent requirements (notably on RISC OS 3.10, sprites are held in the RMA, and the filer seeing some apps could trigger an annoying RMA full error message).

Disabled theme loading on RISC OS 3.10, there's really not enough free RAM to usefully load that many sprites.

Typo in RO310Hook/Boot/Desktop corrected.

Version 1.17. Tagged as 'Boot-1_17'

Previously 3.98 (the nested Wimp supplied with the browser/Java CD), now 5.15.
While not wanting to belittle 117 versions, many of which were STB resource refinements, the main highlights are probably
 * 4.10 UTF8 support
 * 4.82 features contemporary with RISC OS 4.02
 * 4.84 support hires Sprites11
 * 4.85 support searching Wimp$IconTheme
 * 5.13 zero page safe and VFP contexts
 * 5.15 preset Wimp$IconTheme <-- that's the change !Boot was waiting for

RO310/350/360/370/400Hook: PreDesktop files updated to load the newer Wimp
RO430Hook: Unfortunately the ROM copy of the Wimp in 4.39 is already > the themed nested Wimp, but since the RO430Hook is only a skeleton (to allow softloading of RISC OS 5) there's little to gain by loading a Wimp
RO360Hook: Trim out licence from installed !Run file
RO400Hook: Reenable the default monitor definition in the NoVRAM case

Version 1.16. Tagged as 'Boot-1_16'

Everyone deserves a marble backdrop.

Version 1.15. Tagged as 'Boot-1_15'

Makefile produces RO520Hook too, corrected rule for RO430Hook.
Removed unnecessary 'IfThere' from RO500/RO510, both of which are HAL based so have never provided BandLimit.

Version 1.14. Tagged as 'Boot-1_14'

Largely a clone of the RO400Hook.
Also
* Update RO400Hook/ResetCMOS to have the right OS version at the end
* In each 'Desktop' file only Filer_Boot the utilities of the same major version number
* BootRun accepts RISC OS Adjust's extra default library prefix, but outputs a corrected Run$Path (where the CSD is checked before the Lib)
* Makefile and docs updated

Tested on RISC OS 4.39, booting to the desktop with no errors.

Version 1.13. Tagged as 'Boot-1_13'

Clarify comment in 350 patches
Fix bug in SetChoices where it tried to create 2 levels of directory in one operation, leading to choices not getting copied over correctly

Version 1.12. Tagged as 'Boot-1_12'

The BootFX:PreDesk file is no longer required now it's clear that
the splash screen stuff can't really be themed due to where various
things happen in the boot sequence.

Version 1.11. Tagged as 'Boot-1_11'

Detail:
  Configure/Monitors/Other/Generic - Added an MDF entry for the Pandora's native resolution
Admin:
  Tested on Pandora


Version 1.10. Tagged as 'Boot-1_10'

To save a little disc space the default is the RISC OS 5 cream colour folder, hence only the deltas (Ursula and Morris4) are needed.

Version 1.09. Tagged as 'Boot-1_09'

To save a little disc space the default is the RISC OS 5 cream colour folder, hence only the deltas (Ursula and Morris4) are needed.

Version 1.08. Tagged as 'Boot-1_08'

Comes built in from 3.70 onwards, so resides in Hook directory.

Version 1.07. Tagged as 'Boot-1_07'

* Removed triplicate of AKF11/12; AKF30/40; AKF11-40 since (aside from the monitor title) they're all the same
* Renamed Fujitsu L22W2 to use a hyphen like the others
* VM1403 title expanded to Vision Master 1403
* 31 occurences of 'Diaganol' spelling mistake from EDID Decoder tool corrected

Version 1.06. Tagged as 'Boot-1_06'

Version 1.05. Tagged as 'Boot-1_05'

Same as in Boot-1_03 but the switches must come after the command tail according to the help.
Additionally, the option is omitted when an older BootCommands module is present.

Version 1.04. Tagged as 'Boot-1_04'

Version 1.03. Tagged as 'Boot-1_03'

* BootRun no longer bothers checking for CallASWI, since the SharedCLibrary (which is loaded prior to BootRun) does this anyway.
* Try loading the StrongARM app patcher. Note the check is conditioned on 500 since it is assumed any application which has been 32 bitted is now also StrongARM safe, the use of 'X RMLoad' uses cunning on the placement of the PatchApp module in System:Modules.370, thus skipping the load on versions prior to StrongARM too.
* Remove redundant sprite sets from Extras/500 which were no longer being used anyway.
* Import of UserIF specific backdrop textures (Morris4 ones clawed off an old CD, others courtesy of Richard Hallas).
* Makefile rejigged to allow a TARGET=TEXTURES to select which UserIFs to add.
* Updated 'RTexture' to scan the configured theme for its random backdrop in addition to the generics.

Version 1.02. Tagged as 'Boot-1_02'

Retagged as Boot-1_01.

Archive of items on the ROOL USB stick, this contains a number of handy screenmodes that even the most picky monitor shouldn't be upset by, so should be enough to start with.

Version 1.01. Tagged as 'Boot-1_01'

When 'protect ROM sprites' is turned off, and a theme is being loaded (after the desktop has started in order to pick the right eigen factor sprites file), the various built in ROM applications have already installed their iconbar sprites.
Therefore, when new ones are loaded the bounding boxes are wrong (eg. drive icons the wrong width, switcher icon too tall) and need recalculating.
Derived from 'FontChange' new configure helper 'IconChange' does this.

Also removed unnecessary 500/Tools22.

Version 1.00. Tagged as 'Boot-1_00'

By running !ThemeDefs in the desktop file, this was too late to influence the start banner. Moved earlier again.
To solve the problem of *ICONSPRITES and *TOOLSPRITES needing to be in the final desktop mode, the desktop file looks for some system variables and does the deferred load then instead.
Tested with and without 'Protect ROM sprites' turned on.

Version 0.99. Not tagged

Needs to be a little later on, after the desktop has been entered, in order that the pended *WIMPMODE command set in Choices:Boot.PreDesk.Configure has been acted upon (so any *ICONSPRITES commands know which is the best resolution to load).
Inlined the (optional) user setting of the theme, rather than unconditionally doing so in Choices:Boot.PreDesktop then having to go back and fix it up again later.
Tested with some different eigen factors with different sprite sets for each, to check the right one gets loaded.

Version 0.99. Tagged as 'Boot-0_99'