On Sun, Jan 31, 2010 at 9:05 PM, rhn <[email protected]>wrote:

> On Sun, 31 Jan 2010 19:01:47 +0200
> Tom Hacohen <[email protected]> wrote:
>
> > On Sat, Jan 30, 2010 at 9:07 PM, rhn <[email protected]
> >wrote:
> >
> > > On Sat, 30 Jan 2010 18:33:22 +0200
> > > Tom Hacohen <[email protected]> wrote:
> > >
> > > > On Sat, Jan 30, 2010 at 4:40 PM, rhn <
> [email protected]
> > > >wrote:
> > > >
> > > > > On Sat, 30 Jan 2010 16:22:36 +0200
> > > > > Tom Hacohen <[email protected]> wrote:
> > > > >
> > > > > > On Sat, Jan 30, 2010 at 3:26 PM, rhn <
> > > [email protected]
> > > > > >wrote:
> > > > > >
> > > > > > > Hey.
> > > > > > > I finally decided to switch to the new SHR-T after using the
> old
> > > SHR-U
> > > > > > > because I was afraid of breakage and not having time to fix
> stuff
> > > on my
> > > > > own.
> > > > > > > As I expected, most of the bugs are ironed out now, but some
> > > remain.
> > > > > > >
> > > > > > > The most annoying thing is that sometimes, multipart SMS I send
> are
> > > cut
> > > > > > > off. This happens only when my SIM card is full (I remember
> sending
> > > SMS
> > > > > via
> > > > > > > ogsmd required copying it to SIM and then sending). I guess
> that
> > > could
> > > > > be
> > > > > > > solved by copying messages from the SIM and always keeping some
> > > space
> > > > > on the
> > > > > > > card. Or at least presenting the user with a message along the
> > > lines of
> > > > > "You
> > > > > > > have to clean up your SIM card".
> > > > > > >
> > > > > > > The second most annoying thing was a bug in the new screen
> dimming
> > > > > feature.
> > > > > > > In some cases, using the phone irregularly causes the screen to
> get
> > > > > stuck in
> > > > > > > the half-lit mode. After a while, it dims again, but doesn't
> light
> > > up
> > > > > fully
> > > > > > > until restart.
> > > > > > >
> > > > > > > One more thing is Unicode support in SMS. I daresay it's
> actually
> > > > > unwelcome
> > > > > > > in many cases. Most mobile phones and SMS sending services
> convert
> > > > > Unicode
> > > > > > > to plain ASCII, because UCS-2 encoded messages are heavy and
> > > costly.
> > > > > > > Someone will say: "don't use Unicode signs". Well, that would
> be
> > > fine
> > > > > if
> > > > > > > the only input was a dumb keyboard, but our keyboards support
> > > > > dictionaries
> > > > > > > that do contain Unicode letters.
> > > > > > > My proposition is to mgive users choice to convert Unicode
> letters
> > > to
> > > > > their
> > > > > > > ASCII approximations. I would have done that myself, but I
> couldn't
> > > > > find my
> > > > > > > way through the C code responsible for that. I would happily
> extend
> > > the
> > > > > > > settings app though.
> > > > > > >
> > > > > > > Having said that, I have to admit that current SHR-t is the
> most
> > > > > > > comfortable image I've ever used! Being able to store all
> messages
> > > I
> > > > > receive
> > > > > > > is something I've been waiting for so long. The alarms app - it
> > > > > literally
> > > > > > > feels like someone was reading my mind!
> > > > > > > Last but not least, I'm happy to finally have a notes
> application.
> > > I
> > > > > used
> > > > > > > to write my notes down in text files, and that required running
> > > > > terminal and
> > > > > > > then the all-keyboard vi... BTW, there's a need for a
> > > finger-friendly
> > > > > text
> > > > > > > editor, too.
> > > > > > >
> > > > > > > Cheers!
> > > > > > > rhn
> > > > > > > _______________________________________________
> > > > > > >
> > > > > > Concerning messages: We do send messages as 7bit gsm (i.e, not
> ucs2)
> > > when
> > > > > > possible, we do not try to convert to approximation because we
> want
> > > to
> > > > > send
> > > > > > EXACTLY what you asked us to send, you don't want to send unicode
> > > > > messages,
> > > > > > don't type unicode characters, it's as simple as that (in my
> pov).
> > > The
> > > > > > remaining characters counter shows ~70 when it's a unicode
> message
> > > and
> > > > > ~160
> > > > > > when it's 7bit gsm.
> > > > > > Don't you think I have a point?
> > > > > > --
> > > > > > Tom.
> > > > > >
> > > > > It's not as easy as that.
> > > > >
> > > > > I'm using the Illume keyboard with a normal dictionary, containing
> all
> > > the
> > > > > words exactly how they are written. I use the dictionary not only
> for
> > > SMS: I
> > > > > write notes, emails or communicate using Pidgin. I enter Unicode
> words
> > > there
> > > > > and I'm happy about it.
> > > > >
> > > > > I don't want to use two separate dictionaries for SMS and for
> > > everything
> > > > > else. It's impossible with the Illume keyboard anyway, because it
> > > shares the
> > > > > dynamic part regardless of the "main" dictionary.
> > > > >
> > > > > You're basically saying "don't use the dictionary at all or pay
> twice
> > > for
> > > > > messages". Yes, I know that it's not something you couldn't live
> > > without,
> > > > > but it helps *a lot* to be able to type quickly and not pay and arm
> and
> > > a
> > > > > leg for that.
> > > > >
> > > > > I agree that modifying all Unicode messages is unacceptable, but it
> > > > > shouldn't be bad if the user could choose.
> > > > >
> > > > > Cheers,
> > > > > rhn
> > > > > _______________________________________________
> > > > >
> > > > If you can come up with an algorithm that clearly states what you
> want to
> > > > change to what, we can probably make it a config option. Mind
> providing
> > > an
> > > > example sentence (in unicode) and what you want it to translate to?
> (in
> > > 7bit
> > > > gsm).
> > > > Thanks.
> > > > --
> > > > Tom.
> > > >
> > >
> > > About the algorithm - a simple lookup table with Unicode letters would
> be
> > > enough. Some might call them maps, I call them dictionaries.
> > > There was a reverse of that (guessing the Unicode possibilities from a
> > > basic alphabet input) being done in the Illume keyboard, so a table
> like
> > > that is already there:
> > > svn checkout
> > >
> http://svn.enlightenment.org/svn/e/trunk/e/src/modules/illume/e_kbd_dict.c
> > > It lacks some letters though.
> > >
> > > An example sentence:
> > > "Wstałem z łóżka" -> "Wstalem z lozka"
> > > The algorithm I propose is replacing the Latin-Supplement-1 and
> > > Latin-Extended-A letters with the base ones.
> > > Maybe replacing them with strings in case of "ß"->"ss" or "æ"->"ae",
> but we
> > > should first ask people who actually use these.
> > >
> > > I can implement it myself if you provide me with a brief introduction
> into
> > > the new libphoenui.
> > >
> > > Cheers,
> > > rhn
> > > _______________________________________________
> > >
> > I know how to do it, I guess I wasn't clear enough, I need the table, not
> > anything else :P
> > I don't know how to do it alone (as I don't speak those languages) so
> please
> > provide with a table of to and from values, like:
> > ß ss
> > æ ae
> > and more.
> > --
> > Tom.
> >
>
> Sure.
> It's divided into parts: what I found in Illume keyboard source (only
> mapped to lowercase), the Polish conversions I use daily and the rest. I
> don't know how many other people read this thread, so the last section will
> be my guesses about other languages - so that you can choose to include them
> or not.
> If you do, then people will probably start to notice the feature and
> request improvements in the table.
> Also, it probably contains a fair number of characters that will never be
> used.
>
> The list is in attachment.
> Cheers,
> rhn
>
Please open a ticket and attach this table to it, I already have too many
mails marked as important atm :P


-- 
Tom.
_______________________________________________
Shr-User mailing list
[email protected]
http://lists.shr-project.org/mailman/listinfo/shr-user

Reply via email to