Re: [Fsfe-uk] Mac OS X refund

[Chris Croughton]

On Thu, Jan 31, 2008 at 11:24:09PM +0000, Kevin Donnelly wrote:

> On Saturday 26 January 2008 15:53, Chris Croughton wrote:
> > Well, I want most of the features of Noteworthy certainly.  Not
> > necessarily in exactly the same way, though, if that's what you mean by
> > "feature for feature".  What  Ido want particularly is something where a
> > lot of free software written for GUI use breaks down, and that is both
> > the keyboard and mouse interfaces must be capable of doing everything
> > with comparable ease (OK, some things might need selecting a menu item
> > with the mouse but be a single keystroke, others may need a couple of
> > keystrokes to do a single mouse click, but on average they should be
> > about the same).  If there are any functions, es[pecially common ones,
> > which can only be done with the mouse or only with the keyboard then as
> > far as I'm concerned it is bad.  This is an area where I hold up
> > Noteworthy as great UI as an example for any system.
> 
> Fair enough, but of course  you only get that sort of development when you 
> get 
> users giving feedback to the developers, so we're back to the chicken and egg 
> situation.

For something where there is already a working version (for some value
of 'working'), or where users are pulled in from the beginning, we can
give that feedback.  The problem with a lot of free software is that
it's rushed at by geeks who do something which seems adequate for
themselves at the time but which hasn't had that sort of thing designed
in from the start.  I know, I'm a geek just like that, I write software
for myself without designing it for other people (which means most of
the time that it's command-line only).  Yes, the same sort of thing
happens with a lot of shareware as well, Sturgeon's Law applies (90% of
everything is crap).

> > > Have you tried using Rosegarden, one of the unsung heroes of free
> > > software? The latest version of this has dealt with a lot of the notation
> > > issues earlier versions had.  I've just entered a 5-page piano piece, and
> > > that went swimmingly - I found the interface quite easy.  You also get
> > > output in Lilypond, widely recognised (even by some Noteworthy users on
> > > their forum) as providing very good music typography.  Try it and see how
> > > you get on - if you have difficulties getting it running, email me
> > > offlist.
> >
> > If it produced output in XML I'd be more interested.  As far as I could
> > tell the Lilypond format is very good at typography but less so for
> > performance or converting to MIDIfile or other formats.  I work with a
> > number of other people and we need to have things compatible, preferably
> > all using the same software.
> 
> Er ... the Rosegarden .rg file is a gzipped XML file, so you now have an 
> excuse to be more interested :-)  If that's what you're looking for (ie 
> you're going to handle typesetting yourself), you may not even need Lilypond.

Oh, OK.  When I get round to putting together a machine with both X and
audio I'll give it a try.

> > I have been pointed at MuseScore, and I like what I see.  It's nowhere
> > near complete yet but it's certainly going in the right direction.  And
> > being based on Qt4 it is available for Windows and Mac as well, so other
> > people can use it directly (see discussions on the success of Firefox,
> > it has become popular because it is available wherever people want it).
> > It saves in its own XML formast as well as MusicXML (interchangeable
> > pretty much using XSLT).  I need to find out about it and where they are
> > going with it, but they are certainly a candidate for my money.
> 
> Yes, it's quite good, though the last time I used it a couple of years ago I 
> found it very unintuitive.  I just installed it again, to test, and I still 
> think the same - even setting the time-signature is not that obvious.

Interesting, I found the time signature quite intuitive (I set up a 23/8
time signature, just because it could and I happen to have something I'm
writing in 23/8).  But I would certainly do quite a bit of keyboard
remapping (the nice thing is that it is possible to do a /lot/ of
keyboard remapping) for things I want.

> There 
> is also the Muse sequencer - they were originally one package.  Another 
> package worth looking at may be Canorus (canorus.org, when it's up), but it 
> is still at a very early stage - when I tried it last year it was very 
> crash-prone.

I'll have a look.

> > But shareware only gets money because it does what is needed when people
> > use it.  Shareware which isn't up to quality gets no money. 
> 
> Not the whole story - people often support shareware because it is good 
> enough, even if it's not perfect, in the hope that features they would like 
> will be added in future.  It's the same for all species of software, which is 
> why I think it's important to give free software a fair crack of the whip.

For ordinary people it has to have a certain amount of functionality to
even be considerd.  For instance, no one will buy a word processor which
can't save text yet.

As an example, KDiff3 is less than perfect, but it is fully usable.  I
would happily pay for it as shareware in a binary version if they asked
(because it saves me the hassle of building it myself).  And again it is
available for multiple platforms so I can recommend it to people who
can't change from Windows, and they can continue the process of making a
GNU+Windows platform.

> > And the same goes for advocates, if I recommend something which
> > doesn't do the job then my reputation suffers. 
> 
> Certainly, but doesn't it depend a bit on how you define the job in the first 
> place?  Everything has drawbacks and shortcomings - it's a matter of picking 
> the option whose shortcomings are least annoying.  Viewed in that light, it's 
> a question of putting options to people.

With a lot of things I don't define the job at all, it is defined by the
people who already have something usable.  In order to get them to
change I need to have something which at minimum does the same job as
their current software does, and then has advantages (cheaper, fewer
bugs, better features, more frequently maintained, etc.).  Very few
people will be willing to throw away something which works in favour of
something which is worse.

> > If I've got them to change a complete 
> > OS and they don't find it useful then they'll really be pissed not only
> > with me but with free software, and there's already too much "anything
> > free is worth (only) what you paid for it" attitude.
> 
> So that why there's all those BOGOFF offers in the shops, is it?

But BOGOFF isn't free, it's two for the price of one.  Have you ever
tried getting the 'free' one without the one you pay for?  For some
reason the shops won't allow it.  What they do is they get you to buy
two of something you probably wouldn't have bought at full price, and
still make a profit on it.

A better example may be people wanting free music, but from what I've
seen they treat that the same way as other free things, something which
can just be thrown away if it's not exactly what they want.

> > > If you want to use this sort of daft analogy, I could ask why people
> > > argue for closed software being allowed to retain a privileged position
> > > when they wouldn't stand for:
> >
> > I don't see anyone arguing for this "privileged position", we are
> > arguing for freedom.  Freedom to choose what we want is no more wanting
> > to force everyone to use proprietary software than freedom to choose to
> > have an abortion is wanting to abort every pregnancy.
> 
> Not quite.  The impression given in earlier posts was that free software 
> could 
> only go so far, and beyond that the only realistic option was closed 
> software.  This is a self-fulfilling prophecy, and free software would never 
> have got anywhere if people had accepted that it was only a niche product.  
> That's what I mean by giving closed software a privileged position. 

As far as I'm concerned at least, it isn't that free software /can/ only
go so far but that at the moment it /does/ only go so far.  In some
areas free software is better than proprietary (I'll recommend
OpenOffice and Firefox, for instance, over any of their non-free
equivalents), but in others the available free software offerings just
don't cut it yet.  And it's in those areas that I (and if I understand
them others) are arguing that for the moment there is nothing wrong with
using non-free software to get the job done.

> You're perfectly free to choose whatever software you want, but I think it's 
> a 
> bit hypocritical to "advocate" free software by saying that it has inherent 
> and unchanging limits on where it can be used,

I'm not saying that.  I haven't seen anyone say that.  Please point to
where anyone on this list has said that free software has "inherent and
unchanging limits on where it can be used".

> and then to suggest that any 
> who don't agree are being zealots (or in your words "fanatics").  It 
> currently does some things less well than closed software (and a lot of 
> things better) - but that's been the case for years - it hasn't stopped its 
> increasing use as it improves and gets better.  I read only today that the 
> French gendarmerie is going to switch 70,000 desktops to Linux 
> (http://www.theregister.co.uk/2008/01/30/french_open) - presumably another 
> set of unfortunates who don't realise that the Linux desktop isn't yet ready.

I haven't said that the Linux desktop isn't yet ready for anything.
It's obviously plenty ready for many people, including several
non-techie people I know.  Indeed, I would say that the desktop itself
is now of well merchantable quality.  There is, however, a lack in
applications in some areas.  Quel surprise, the same is still true of
all the other desktops, it takes an infinite time to write all of the
possible applications for every possible use.  I'm not at all surprised
that the gendarmerie has found that they have all the applications they
need, I suspect that most office environments could easily use free
software for everything they do.  But in many places they are hampered
because of the percieved costs of changing, or biases of non-technical
people (if the Big Boss says "we only use MS" there's not much the staff
can do about it).

> > > - using Citroen-branded petrol because that's all the manufacturer of
> > > their Citroen car allows them to use;
> >
> > Even Windows computers run on the same electricity as GNU+Linux ones.
> 
> Yes, which is very different from the tied analogies (eg traffic-lights) you 
> used earlier.  Imagine the fuss there'd be if people had to get a special 
> Microsoft power-line to run their Windows PCs.

Which they don't, so the analogy breaks down.

> > It's actually more like buying a new radio for the car, and people do
> > indeed do that and pay for it.
> 
> But that's exactly the point - they wouldn't take kindly to the carmaker 
> saying that they could only install that carmaker's own brand of radio, and 
> not one by Bose or whoever took their fancy.

Again, that's exactly my point.  No one says that you can only install
MS software on Windows systems.  You can happily install OpenOffice,
Firefox, KDiff3, Audacity, Vim, GIMP, etc., on Windows (that's just the
ones I can think of off the top of my head which I've used and
recommended to other people).

The last system I remember which said that you were only allowed to
install the manufacturer's own software was Apple, and they gave up that
idea years ago.  Windows has had third-party applications from its
beginning.

Actually, many car manufacturers do make it difficult to install other
radios and attachments, using proprietary connectors for CD
autochangers, radios which are integrated with the proprietary display,
etc.  I tried to get the information for the CD autochangers, assuming
that it was a standard, and found that it's blocked to third-party
developers by several major manufacturers.  It doesn't seem to stop
people from buying cars by Peugeot etc.  At least Windows has mostly
published APIs, you can even write Windows GUI software using gcc.

Chris C