In message <mpro.p8nswy015vzd501wv.lists@xxxxxxxxxxxxxxxxxx>
Steve Fryatt <lists@xxxxxxxxxxxxxxxxxx> wrote:
On 10 May, Chris Hughes wrote in message
<037920f656.chris@xxxxxxxxxxxxxxxxx>:
I am not happy with these ROOL changes to Printers which is wasting a lot
of my time. I am very tempted to roll back to the early OS and earlier
!printers and never upgrade the OS.
Hang on a minute. You have absolutely *no* idea, from what I can see, what
is causing these issues. Further, I'm afraid that I don't recognise "OS10a"
as a ROOL release number.
These changes came out some time ago on the ROOL "unstable" test stream, and
I've not seen any discussion on the forums along these lines. There has been
plenty of opportunity for anyone -- including you, if you got involved in
testing new builds of the OS like so many of us do in our own time -- to
spot and report these issues.
Further, R-Comp supposedly check these builds and approve them for release
to their users in the form of things like "OS10a".
I do note that there's no stable 5.24 release for the ARMX6, so you're
always in the same boat risk-wise as a user who has just stuck the latest
nightly-build onto their machine.
There seems to be no real benefit to the forced changes in !Printers
What do you mean "forced" changes? Why do you think that they are "forced"?
Who "forced" them on you, any more than they "forced" any of the other
changes or updates?
I presume that there's a reason for the changes. I can see only a couple
made to Printers in the past year or so, and none of these are obvious
candidates: fixes to the way that Choices are handled (as they seem to have
been somewhat broken), and some fixes to things which were trying to read
printing data from NULL pointers (not good, whatever FUD you've been fed
about how such things don't matter). The rationale in the commit logs looks
sane to me.
I have no connection with ROOL, other then being extremely grateful for the
massive amount of work that their small team and all of the volunteers do to
ensure that we continue to have an OS to use on our ARM systems. In a
platform the size of RISC OS, there will always be problems that slip
through the net: testing is largely done by users like you and I who run the
odd numbered "unstable" builds of RISC OS (such as 5.23).
Your use of the word "forced" bothers me a lot. I've been concerned for a
while (as have other developers that I've spoken to at shows and by private
email) at the tone being used by R-Comp over an apparent disagreement with
ROOL over the direction of RO5. This seems to have started with the ZeroPain
stuff (which, as I hope we made clear on this list before being shut down,
is a non-issue for most users now that a new stable build is available) and
has gone on to other things before culminating in the public spat over 5.24
for the AMRX6.
The fact that you're claiming to have had these changed "forced" on to you
suggests that this corrosive attitude is spreading to users. The changes are
no more "forced" on to you than any other changes made by any other
developer. Remember that without ROOL and their "forced" changes, you would
not have had any version of RISC OS running on an ARMX6 -- because RISC OS 5
would not exist in its current ARMv7-friendly form. "OS8", "OS10a" and
whatever are just R-Comp taking the work of ROOL and their volunteers,
testing it a bit (why didn't they spot the printing issue, eh?) and bundling
it out to you.
Instead of making idiotic statements like "I'll never upgrade the OS again",
why not report this issue on the ROOL forums, engage with any responses that
you get, and find a solution? I've not yet seen you do this.
