Google Groups no longer supports new Usenet posts or subscriptions. Historical content remains viewable.
Dismiss
Re: HongKong 1.67 released
10 views
Skip to first unread message
Peter Young
Aug 4, 2012, 5:59:47 PM8/4/12
to
On 4 Aug 2012 Steve Drain <st...@kappa.me.uk> wrote:
> HongKong is a version of the Mahjong patience game Shanghai.
> It was written as a demonstration of Basalt and its Toolbox keywords.
> It uses a drag-and-drop interface and differs in some other ways from
> alternative versions, so it might be worth a look if you are a fan.
> This version cures some serious problems with newer versions of RISC OS
> on both ROL and ROOL forks. [1]
Confirmed that it's now working on this ARMini, RISC OS 5.19; thanks.
Peter.
Peter.
--
Peter \ / zfc Ta \ Prestbury, Cheltenham, Glos. GL52
and \/ __ __ \ England.
family / / \ | | |\ | / _ \ http://pnyoung.orpheusweb.co.uk
/ \__/ \_/ | \| \__/ \______________ pny...@ormail.co.uk
> HongKong is a version of the Mahjong patience game Shanghai.
> It was written as a demonstration of Basalt and its Toolbox keywords.
> It uses a drag-and-drop interface and differs in some other ways from
> alternative versions, so it might be worth a look if you are a fan.
> This version cures some serious problems with newer versions of RISC OS
> on both ROL and ROOL forks. [1]
Confirmed that it's now working on this ARMini, RISC OS 5.19; thanks.
Peter.
Peter.
--
Peter \ / zfc Ta \ Prestbury, Cheltenham, Glos. GL52
and \/ __ __ \ England.
family / / \ | | |\ | / _ \ http://pnyoung.orpheusweb.co.uk
/ \__/ \_/ | \| \__/ \______________ pny...@ormail.co.uk
Steve Fryatt
Aug 4, 2012, 5:58:56 PM8/4/12
to
On 4 Aug, Steve Drain wrote in message
<ap.4f245552ba.a...@argonet.co.uk>:
> The ROOL RO 5.17+ OS_Heap has been rewritten, and changes the previous
> API, which Basalt relied on. This has caused problems for some other
> developers, too.
Which API has been changed? The internal heap structure has been changed,
but that was never an API: IIRC, it was documented in the PRMs as "this
might be interesting, but don't rely on it".
--
Steve Fryatt - Leeds, England
http://www.stevefryatt.org.uk/
<ap.4f245552ba.a...@argonet.co.uk>:
> The ROOL RO 5.17+ OS_Heap has been rewritten, and changes the previous
> API, which Basalt relied on. This has caused problems for some other
> developers, too.
Which API has been changed? The internal heap structure has been changed,
but that was never an API: IIRC, it was documented in the PRMs as "this
might be interesting, but don't rely on it".
--
Steve Fryatt - Leeds, England
http://www.stevefryatt.org.uk/
Brian Howlett
Aug 5, 2012, 5:12:41 AM8/5/12
to
On 4 Aug, Steve Drain wrote:
> This version cures some serious problems with newer versions of RISC OS
> on both ROL and ROOL forks.
Working on RISC OS 5.18 now, but still shows the version number as
> This version cures some serious problems with newer versions of RISC OS
> on both ROL and ROOL forks.
1.66 (03 Aug 2012).
Cheers,
--
Brian Howlett
-----------------------------------------------
Karl Marx's grave. Just another Communist Plot?
Steve Drain
Aug 5, 2012, 7:39:42 AM8/5/12
to
Steve Fryatt wrote:
<OS_Heap>
> Which API has been changed? The internal heap structure has been changed,
> but that was never an API: IIRC, it was documented in the PRMs as "this
> might be interesting, but don't rely on it".
Looking back, you are, of course, quite right.
The bit about allocations always being in blocks of 8 bytes is over the
page and several paragraphs after the preamble. It was easy to read that
information quite independently of any warning, and I must have absorbed
it a very long time ago.
As far as programming goes, assuming the 8 byte block size saved an
instruction or two, but clearly that was not worth it.
Steve
<OS_Heap>
> Which API has been changed? The internal heap structure has been changed,
> but that was never an API: IIRC, it was documented in the PRMs as "this
> might be interesting, but don't rely on it".
The bit about allocations always being in blocks of 8 bytes is over the
page and several paragraphs after the preamble. It was easy to read that
information quite independently of any warning, and I must have absorbed
it a very long time ago.
As far as programming goes, assuming the 8 byte block size saved an
instruction or two, but clearly that was not worth it.
Steve
Steve Drain
Aug 5, 2012, 7:53:04 AM8/5/12
to
On 05/08/2012 10:12, Brian Howlett wrote:
> Working on RISC OS 5.18 now, but still shows the version number as
> 1.66 (03 Aug 2012).
Thanks. Corrected now.
> Working on RISC OS 5.18 now, but still shows the version number as
> 1.66 (03 Aug 2012).
Steve
0 new messages