[Date Prev][Date Next][Thread Prev][Thread Next][Date Index][Thread Index]

Re: emacspeak and w3



Hi Raman,

well, I have to say, I'm pretty impressed with chromevox and even Lois
isn't that bad. In fact, I'm sending this email from within chrome
using gmail's standard web UI rather than from VM. It took a little
bit to get the layout etc, but wasn't that hard. The Lois TTS has a
few pops and clicks, lags a little sometimes, but is quite usable.

For others who may be interested, this was so easy to install on my
xubuntu 12.04 box. All I had to do was

1. Remove the Ubuntu chromium-browser package

2. Download the deb version of google.chrome from the google website.
I'm running the 64 bit version

3. Install the deb padkage

4. Go tot he chrome store and install the Loix TTS extension and then
the chromevox extension.

Thats it. I then just read most of the uses guide and I was up and
going. Very simple and very fast.

The browsing model for chromevox seems very good. It is easy to move
from a high level grop browsing to objet, paragraph, sentence, word or
right down to character. Jumping back and forward between links or
headers is easy. I quite like the idea of skimming around the page at
a group level and then drilling down into object or paragraph etc once
I find bits I'm interested in.

Really looking forward to seeing how well this works on other
websites. Would also like to see if it is possible to make it work
with other TTS engines or not. Lois is OK, but it would be nice to
have the familiar viavoice engine.

I'm really impressed with how well this works when you consider it is
all just running as browser extensions. I'm going to go and read the
article Raman posted about that describes this work at
http://google-axs-chrome.googlecode.com/svn/trunk/developer/chromevox-overview-2012/paper.html

regards,

Tim

On 13 May 2012 15:10, Tim Cross <theophilusx@gmail.com> wrote:
> OK, will try and get a version of the chrome browser from google which
> has the macl stuff built in - rather irritating if Ubuntu doesn't do
> this.
>
> Willl also work on fixing up the mm-* stuff so that we have a current
> setup that fits with current defaults packaged with emacs 24 and
> possibly some docs for those who may prefer to use their own config
> rather than customize. I noticed when I whent to setup mine that
> something I'm loading is already customizing things outside the
> standard customize framework, which may cause unreliable results if I
> don't track it down.
>
> Thanks for the clarification the other stuff is either old kruft or
> errors from cut and paste. Again, will sort things out and report
> back.
>
> Tim
>
>
> On 13 May 2012 00:58, T. V. Raman <tv.raman.tv@gmail.com> wrote:
>>
>> responses in-lined:
>>
>>>>>>> "Tim" == Tim Cross <theophilusx@gmail.com> writes:
>>    Tim> I installed lois and chromevox from the chrome
>>    Tim> store. Running version Chromium 18.0.1025.168 from
>>    Tim> ubuntu 12.04, but no luck getting it working yet. Not
>>    Tim> sue how to verify nacl support - looks like it was added
>>    Tim> in v14 and it is listed/mentioned in the copyright file
>> See Jason's note. Not all debian/ubuntu builds include nacl. Get
>> package google-chrome -- rather than chromium.
>>
>>    Tim> of this version, but no luck yet.
>>    Tim>
>>    Tim> On another note, seem to have run itno a bug with emacs
>>    Tim> 24 url, so now I cannot follow links in w3. Seems to be
>>
>> this is transient. A couple of weeks ago, they broke all google
>> search result links -- see the advice in emacspeak-w3.el
>> that "canonicalizes google result urls when passed to the url
>> library.  I  filed a bug and I suspect they broke something else
>> in fixing that. Make sure you're running emacspeak from svn and
>> that you have the latest snapshot as well.
>>
>>    Tim> losing the 'host' value somewhere, so throws a stingp
>>    Tim> nil error. Was working with a version of emacs from
>>    Tim> earlier in the week, so probably just a temporary hitch.
>>
>>
>>    Tim>
>>    Tim> Was working through your mm-customize.el file and
>>    Tim> noticed something which I found a bit confusing. In it,
>>    Tim> you have some tests under the mm--inline-media-tests
>> that is old cruft, and the or is a no-op -- consequence of cut
>>    Tim> and paste.
>>
>>    Tim> setting of the form
>>    Tim>
>>    Tim> (lambda (&rest ignore) (or mm-text-html-renderer
>>    Tim> mm-text-html-renderer))
>>    Tim>
>>    Tim> I don't understand the or statement. It has the same
>>    Tim> varible for both paths - seems like a weird thing to do,
>>    Tim> unless there is something I'm missing? I would have
>>    Tim> thought just returning the var would achieve the same
>>    Tim> result? Just curious really as I always like to try and
>>    Tim> understand bits of elisp I come across.
>>    Tim>
>>    Tim>
>>    Tim> Tim
>>    Tim>
>>    Tim>
>>    Tim> On 12 May 2012 13:05, T. V. Raman
>>    Tim> <tv.raman.tv@gmail.com> wrote:
>>    >> nds good.
>>    >>
>>    >> My own choice for general purpose browsing is Chrome with
>>    >> chromevox loaded -- give that a shot.
>>    >>
>>    >> You'll need to:
>>    >>
>>    >> 1. Install ChromeVox and the Lois voice from the Chrome
>>    >> Web Store; 2. Ensure that your Chrome has nacl support.
>>    >>
>>    >> Eventually I'd like to use Chrome's remote debugging API
>>    >>  to connect emacs to Chrome --- and do the following:
>>    >>
>>    >> Use ChromeVox to work with the general purpose Web.
>>    >>
>>    >> 2. Use the remote port connection between Emacs and Chrome
>>    >> to get Web content into Emacs so you can further
>>    >> manipulate it as desired. All of this is a fair bit of
>>    >> work.
>>    >>
>>    >>>>>>> "Tim" == Tim Cross <theophilusx@gmail.com> writes:
>>    >>    Tim> I could be way off the mark, but my thoughts were
>>    >> that    Tim> the w3 rendering code could be a worthwhile
>>    >> reference    Tim> wrt rendering, but most of it would need
>>    >> to be    Tim> updated/modified - same with the css
>>    >> module. I would    Tim> like to see something with very
>>    >> separate and independent    Tim> modules that each have a
>>    >> well defined interface that are    Tim> all hooked up
>>    >> together to provide a usable browser AND    Tim> provide a
>>    >> useful set of utilities that could be used for    Tim>
>>    >> other processin and provide an interesting tool set for
>>    >>  Tim> experimentation and personal customization.    Tim>
>>    >>    Tim> It is probably this last part, experimentation and
>>    >>    Tim> customization for specific tasks I find most
>>    >> appealing    Tim> about w3. For general browsing, I am
>>    >> finding it    Tim> increasingly necessary to use something
>>    >> like firefox or    Tim> chromium. However, as the epub
>>    >> stuff you did shows,    Tim> there is still a lot of
>>    >> potential for a reliable text    Tim> browser with a very
>>    >> close integration with emacs that    Tim> will allow us to
>>    >> manipulate components with custom    Tim> elisp.    Tim>
>>    >>  Tim> Tim    Tim>    Tim> On 11 May 2012 13:23,
>>    >> T. V. Raman    Tim> <tv.raman.tv@gmail.com> wrote:    >>
>>    >> shr doesn't use any of the W3 rendering code.    >>    >>
>>    >> Unfortunately the parse structure that the W3 parser    >>
>>    >> builds is not exactly the same shape as that returned by
>>    >>  >> libxml -- so you'd have to do a bit of work before the
>>    >> W3    >> renderer can eat the libxml parse tree. But it
>>    >> shouldn't    >> be a huge amount of work.    >>    >>>>>>>
>>    >> "Tim" == Tim Cross <theophilusx@gmail.com> writes:    >>
>>    >>  Tim> Now that is very interesting. Will certainly be
>>    >>  >> checking    Tim> out shr - this is exactly along the
>>    >> same    >> lines as I was    Tim> thinking wrt moving w3
>>    >> forward and    >> helping to ensure it    Tim> doesn't end
>>    >> up dying of bit    >> rot.    Tim>    Tim> Since emacs
>>    >> added the interface to    >> libxml, I've been    Tim>
>>    >> wondering about how much work    >> this would take, so
>>    >> I'm    Tim> quite excited to find that    >> thiswork has
>>    >> already started.    Tim>    Tim> Tim    Tim>    >> Tim> On
>>    >> 11 May 2012 03:47, T.V. Raman    >>
>>    >> <tv.raman.tv@gmail.com>    Tim> wrote:    >> From an    >>
>>    >> Emacspeak perspective, W3 will always trump W3M    >>  in
>>    >>    >> terms of features --- w3m in emacs is nice, but the
>>    >>    >>    >> integration leaves a lot to be desired -- in
>>    >> that emacs/w3    >> >> only gets its hands on the content
>>    >> after w3m itself has    >> >> done the bulk of the
>>    >> rendering.    >>    >> In the best    >> of all worlds,
>>    >> the parsing and dom    >> construction    >> would happen
>>    >> in the native layer --- and the    >>    >> rendering
>>    >> happen in the lisp layer.  Module shr is a    >>    >>
>>    >> beginning to that end --- in that it uses libxml to do the
>>    >>    >> >> parsing --    >>    >> On 5/9/12, Robert
>>    >> D. Crawford    >> <robdcraw@gmail.com> wrote:    >>> "Tim
>>    >> Cross"    >> <tcross@une.edu.au> writes:    >>>    >>>>
>>    >> The really nice    >> thing is that since machines have
>>    >> gotten    >>>> so much    >> faster in the past few years,
>>    >> the slower    >>>> rendering    >> time of w3 compared to
>>    >> w3m is pretty much    >>>>    >> unnoticeable except on
>>    >> really large html files.    >>>    >> >>> I've not used
>>    >> w3m in such a long time.  I've preferred    >> w3    >>>
>>    >> since I first used it.  Glad to hear I'm not    >> really
>>    >>    >>> missing out on speed.    >>>    >>>> thanks    >>
>>    >> again, you have made my day!    >>>    >>> You are quite
>>    >>  >> welcome.  Glad to have been of assistance.    >>>
>>    >>  >>>    >> rdc    >>> --    >>> Robert D. Crawford    >>>
>>    >>    >> robdcraw@gmail.com    >>>    >>>    >>
>>    >> -----------------------------------------------------------------------------
>>    >>    >> >>> To unsubscribe from the emacspeak list or change
>>    >> your    >> >>> address on the emacspeak list send mail to
>>    >>    >>>    >> "emacspeak-request@cs.vassar.edu" with a
>>    >> subject of    >>>    >> "unsubscribe" or "help".    >>>
>>    >>  >>>    >>    >>    >>
>>    >> -----------------------------------------------------------------------------
>>    >>    >> >> To unsubscribe from the emacspeak list or change
>>    >> your    >> >> address on the emacspeak list send mail to
>>    >>  >>    >> "emacspeak-request@cs.vassar.edu" with a subject
>>    >> of    >>    >> "unsubscribe" or "help".    >>    Tim>
>>    >>  Tim>    Tim>    >> Tim> -- Tim Cross    >>    >> --    >>
>>    >> Best Regards, --raman    >>    >> --    >> Best Regards,
>>    >> --raman    Tim>    Tim>    Tim>    Tim> -- Tim Cross
>>    >>
>>    >> --
>>    >> Best Regards, --raman
>>    Tim>
>>    Tim>
>>    Tim>
>>    Tim> -- Tim Cross
>>
>> --
>> Best Regards,
>> --raman
>
>
>
> --
> Tim Cross



-- 
Tim Cross

-----------------------------------------------------------------------------
To unsubscribe from the emacspeak list or change your address on the
emacspeak list send mail to "emacspeak-request@cs.vassar.edu" with a
subject of "unsubscribe" or "help".



If you have questions about this archive or had problems using it, please send mail to:

priestdo@cs.vassar.edu No Soliciting!

Emacspeak List Archive | 2010 | 2009 | 2008 | 2007 | 2006 | 2005 | 2004 | 2003 | 2002 | 2001 | 2000 | 1999 | 1998 | Pre 1998

Emacspeak Files | Emacspeak Blog | Search the archive