Shawalphabet YahooGroup Archive Browser

hF /filip
nFs t nO sum wun iz stil Qt HX lisaniN.
wut dM V HiNk v H pOatrI v /rybxt /fryst
it hAz bIn kumpXd wiT H pOatrI v
/yOhyn /wUlfgaN /von /gOT.
vXI sYlem.
cCz, /pYl /vI.
__________________atAct___________________________
--- In shawalphabet@yahoogroups.com, "Philip Newton"
<philip.newton@...> wrote:
>
> On 4/30/06, paul vandenbrink <pvandenbrink11@...> wrote:
>> f wut sImz t bI a lYN tFm, nuTiN hAz bIn pOstid
>> if its Just mI bIiN a bit pXanqd, F apolaJFz.
>
> its not H fxst lul in konvDsESan, n probablI wOnt bI H lyst :)
>
> cCz, /filip

On 4/30/06, paul vandenbrink <pvandenbrink11@...> wrote:
> wut dM V HiNk v H pOatrI v /rybxt /fryst

t bI onast, F didant rId H pOamz... Fm not tM intristad in pOatrI.

/filip
--
Philip Newton <philip.newton@...>

hF /filip
hE HAts O-kE.
t bI onest, pOatrI iz nyt f evrI wun.

it tUk mI a lYN tFm t lxn t undDstAnd
n enJQ a smYl sub set v Yl H avElabal
pOams. F lFk t rId fAst,
n pOatrI OnlI wxks wen V slO dQn n
wxk yD tuN arQnd H wxdz.

it kAn be hAndI az a nemynik.
fP eksAmpal.

30 dEz haT /septembD
/E-pril, /JMn n /nOvembD
Yl H rest haV 31
sEv /febrMXI glum.

rigRdz, /pYl /vI.
_____________atAct______________________________
--- In shawalphabet@yahoogroups.com, "Philip Newton"
<philip.newton@...> wrote:
>
> t bI onast, F didant rId H pOamz... Fm not tM intristad in pOatrI.

helO YlivD
TANks f H kwOt abQt pOatrI bF /rybxt /brQniN "Robert Browning"
its kFnd v lFk a puzal.

> rQnd n rQnd,
> lFk swiraliN snO,
> in a dAzaling drift,
> aZ its gRdIanz, gO,
> flOtiN hiz wimen
> fEdad f EJaz,
> but skulpcxd in stOn
> f H pOats pEJaz.

it iz O-kE, but F lFk /rybxt /fryst betD.

rigRdz, /pYl /vI.
--- In shawalphabet@yahoogroups.com, "paul vandenbrink"
<pvandenbrink11@...> wrote:
> t bI onest, pOatrI iz nyt f evrI wun.
>
> it tUk mI a lYN tFm t lxn t undDstAnd
> n enJQ a smYl sub set v Yl H avElabal
> pOams. F lFk t rId fAst,
> n pOatrI OnlI wxks wen V slO dQn n
> wxk yD tuN arQnd H wxdz.
>
> it kAn be hAndI az a nemynik.
> fP eksAmpal.
>
> 30 dEz haT /septembD
> /E-pril, /JMn n /nOvembD
> Yl H rest haV 31
> sEv /febrMXI glum.

Ph. D. wrote:

>paul vandenbrink said:
>
>
>>Star Raven wrote:
>>
>>
>>>what about the homonyms? PERfect vs. perFECT ect?
>>>
>>>
>>It should be PERF-ect (Verb) vs per-FECT (Adjective)
>>which not only has a different syllable boundary but also a
>>distinct Shaw Spelling.
>>
>> pxf-ekt ----- pD-fekt
>>
>>There are going to be homonyms in Shavian, perhaps even more
>>than T.O In Shavian, won = one
>>
>>Overall, it still will be miles ahead of T.O.
>>
>>
>
>
>This thread started out by suggesting that large amounts
>of literature could be transcribed by using a computer
>program and a look-up table. I believe Star's question
>was how to handle these homonyms in a look-up table.
>("one" vs. "won" is not a problem; only those which
>can have a different stress.)
>
>As someone else pointed out, natural language processing
>in English is not trivial. I would suggest that the look-up
>table simply have a flag on those words with more than one
>Shavian transcription. The program which does the trans-
>scription could flag those words in the output. Then a
>human could select the correct spelling. Thus 95% (or more)
>of the text would be automatic, with a minimal amount of
>human intervention .
>
>--Ph. D.
>
>
My thoughts exactly. The flag can be as simple as having more than one
solution to a word in the lookup table. Portions of the lookup table
might look like this:

Douglas = duglas, Gduglas
douglas = duglas
dour = dQD
douse = dQs, dQz
dove = duv, dOv
dovecote = duvkOt
dovekie = duvkI


reactive = rIAktiv
reactively = rIAktivlI
reactiveness = rIAktivnas
reactivity = rIAktivitI
reactor = rIAktD
read = rId, red

When the text conversion program comes across the word "reactor", it
outputs "rIAktD" and goes on to the next word. But when it finds the
word "read", it sees two choices, which can be nearly impossible to
ascertain from context even for people at times! The simplest solution
might be to pop up a dialog asking the user to choose the output: "rId"
or "red".

Example:
"I read that book." Could be "F rId HAt bUk." or "F red HAt bUk."

--
Ethan

Sorry about the reply to an ancient message, but I somehow ran into some
old messages I had missed from last year and thought I was in the newer
ones!

Ethan wrote:

>My thoughts exactly. The flag can be as simple as having more than one
>solution to a word in the lookup table. Portions of the lookup table
>might look like this:
>
>Douglas = duglas, Gduglas
>douglas = duglas
>dour = dQD
>douse = dQs, dQz
>dove = duv, dOv
>dovecote = duvkOt
>dovekie = duvkI
>
>
>reactive = rIAktiv
>reactively = rIAktivlI
>reactiveness = rIAktivnas
>reactivity = rIAktivitI
>reactor = rIAktD
>read = rId, red
>
>When the text conversion program comes across the word "reactor", it
>outputs "rIAktD" and goes on to the next word. But when it finds the
>word "read", it sees two choices, which can be nearly impossible to
>ascertain from context even for people at times! The simplest solution
>might be to pop up a dialog asking the user to choose the output: "rId"
>or "red".
>
>Example:
>"I read that book." Could be "F rId HAt bUk." or "F red HAt bUk."
>
>
>

--
Ethan

Hi Ethan
Better late than never.
I think that there are relatively few English words
that have two equally valid pronunciations inside a particular
accent group.
If we specify to the transliteration program, which accent group to
convert the
written T.O. English into, there will only be a small number of
exceptions. If there is a exception, we can embed both Shavian
spellings
into the text with curly brackets around them.
For example, the word perfect, becomes {pxf-ekt pD-fekt}

I don't think this is much of hard-ship, and shouldn't prevent
automatic translation.

There is another issue. Some words (i.e. names, etc.) and
abbreviations are not common enough to be registered in
the Database. By the way there are a lot of duplicate abbreviations
that represent different things. These
other Words that are not successfully translated, should be marked
with an asterisk and if possible retain their Roman spelling.

How many additional letters, would have to be added to Shavian
to allow a Roman equivalent pronunciation.
Offhand, I would think you would only need a way to represent
C, Q and X in Shavian. All the other Roman Letters have Shavian
Equivalents.
That would better than having to fiddle with 2 Fonts.

Regards, Paul V.
__________________________attached_______________________________
--- In shawalphabet@yahoogroups.com, Ethan <ethanl@...> wrote:
> >>Star Raven wrote:
> >>
> >>>what about the homonyms? PERfect vs. perFECT ect?

> >>It should be PERF-ect (Verb) vs per-FECT (Adjective)
> >>which not only has a different syllable boundary but also a
> >>distinct Shaw Spelling.
> >>
> >> pxf-ekt ----- pD-fekt
> >>
> >>There are going to be homonyms in Shavian, perhaps even more
> >>than T.O In Shavian, won = one
> >>
> >>Overall, it still will be miles ahead of T.O.

> >This thread started out by suggesting that large amounts
> >of literature could be transcribed by using a computer
> >program and a look-up table. I believe Star's question
> >was how to handle these homonyms in a look-up table.
> >("one" vs. "won" is not a problem; only those which
> >can have a different stress.)
> >
> >As someone else pointed out, natural language processing
> >in English is not trivial. I would suggest that the look-up
> >table simply have a flag on those words with more than one
> >Shavian transcription. The program which does the trans-
> >scription could flag those words in the output. Then a
> >human could select the correct spelling. Thus 95% (or more)
> >of the text would be automatic, with a minimal amount of
> >human intervention .
> >
> >--Ph. D.
> >
> >
> My thoughts exactly. The flag can be as simple as having more than
one
> solution to a word in the lookup table. Portions of the lookup
table
> might look like this:
>
> Douglas = duglas, Gduglas
> douglas = duglas
> dour = dQD
> douse = dQs, dQz
> dove = duv, dOv
> dovecote = duvkOt
> dovekie = duvkI
>
>
> reactive = rIAktiv
> reactively = rIAktivlI
> reactiveness = rIAktivnas
> reactivity = rIAktivitI
> reactor = rIAktD
> read = rId, red
>
> When the text conversion program comes across the word "reactor",
it
> outputs "rIAktD" and goes on to the next word. But when it finds
the
> word "read", it sees two choices, which can be nearly impossible to
> ascertain from context even for people at times! The simplest
solution
> might be to pop up a dialog asking the user to choose the
output: "rId"
> or "red".
>
> Example:
> "I read that book." Could be "F rId HAt bUk." or "F red HAt bUk."

Thanks, Paul. Now, since I didn't catch a bunch of messages from that
time last year, did anybody actually do anything with the suggestions
that were made? Do we have such a thing as a Roman to Shavian
translation program? I've been pondering the possibility of making such
a thing for a while now, even though I'm not much of a programmer
myself. I kind of like the idea of having a program which can operate
with a designated substitution list, or will make a list by prompting
the user if none has been supplied ahead of time.

Embedding the alternates in the the output text, using curly brackets,
sounds like a good idea to me. It requires no user intervention during
the conversion process, thus makes it possible to run lots of text
through the convertor without interruption. Then the only thing
necessary is to open the text with a text editor and do a search for
curly brackets, and change each word to the one the context demands.

I'm not quite sure I follow you regarding the adding of C, Q, and X to
Shavian. Could you explain that a bit further?

--
Ethan



paul vandenbrink wrote:

>Hi Ethan
>Better late than never.
>I think that there are relatively few English words
>that have two equally valid pronunciations inside a particular
>accent group.
>If we specify to the transliteration program, which accent group to
>convert the
>written T.O. English into, there will only be a small number of
>exceptions. If there is a exception, we can embed both Shavian
>spellings
>into the text with curly brackets around them.
>For example, the word perfect, becomes {pxf-ekt pD-fekt}
>
>I don't think this is much of hard-ship, and shouldn't prevent
>automatic translation.
>
>There is another issue. Some words (i.e. names, etc.) and
>abbreviations are not common enough to be registered in
>the Database. By the way there are a lot of duplicate abbreviations
>that represent different things. These
>other Words that are not successfully translated, should be marked
>with an asterisk and if possible retain their Roman spelling.
>
>How many additional letters, would have to be added to Shavian
>to allow a Roman equivalent pronunciation.
>Offhand, I would think you would only need a way to represent
>C, Q and X in Shavian. All the other Roman Letters have Shavian
>Equivalents.
>That would better than having to fiddle with 2 Fonts.
>
>Regards, Paul V.
>__________________________attached_______________________________
>--- In shawalphabet@yahoogroups.com, Ethan <ethanl@...> wrote:
>
>
>>>>Star Raven wrote:
>>>>
>>>>
>>>>
>>>>>what about the homonyms? PERfect vs. perFECT ect?
>>>>>
>>>>>
>
>
>
>>>>It should be PERF-ect (Verb) vs per-FECT (Adjective)
>>>>which not only has a different syllable boundary but also a
>>>>distinct Shaw Spelling.
>>>>
>>>> pxf-ekt ----- pD-fekt
>>>>
>>>>There are going to be homonyms in Shavian, perhaps even more
>>>>than T.O In Shavian, won = one
>>>>
>>>>Overall, it still will be miles ahead of T.O.
>>>>
>>>>
>
>
>
>>>This thread started out by suggesting that large amounts
>>>of literature could be transcribed by using a computer
>>>program and a look-up table. I believe Star's question
>>>was how to handle these homonyms in a look-up table.
>>>("one" vs. "won" is not a problem; only those which
>>>can have a different stress.)
>>>
>>>As someone else pointed out, natural language processing
>>>in English is not trivial. I would suggest that the look-up
>>>table simply have a flag on those words with more than one
>>>Shavian transcription. The program which does the trans-
>>>scription could flag those words in the output. Then a
>>>human could select the correct spelling. Thus 95% (or more)
>>>of the text would be automatic, with a minimal amount of
>>>human intervention .
>>>
>>>--Ph. D.
>>>
>>>
>>>
>>>
>>My thoughts exactly. The flag can be as simple as having more than
>>
>>
>one
>
>
>>solution to a word in the lookup table. Portions of the lookup
>>
>>
>table
>
>
>>might look like this:
>>
>>Douglas = duglas, Gduglas
>>douglas = duglas
>>dour = dQD
>>douse = dQs, dQz
>>dove = duv, dOv
>>dovecote = duvkOt
>>dovekie = duvkI
>>
>>
>>reactive = rIAktiv
>>reactively = rIAktivlI
>>reactiveness = rIAktivnas
>>reactivity = rIAktivitI
>>reactor = rIAktD
>>read = rId, red
>>
>>When the text conversion program comes across the word "reactor",
>>
>>
>it
>
>
>>outputs "rIAktD" and goes on to the next word. But when it finds
>>
>>
>the
>
>
>>word "read", it sees two choices, which can be nearly impossible to
>>ascertain from context even for people at times! The simplest
>>
>>
>solution
>
>
>>might be to pop up a dialog asking the user to choose the
>>
>>
>output: "rId"
>
>
>>or "red".
>>
>>Example:
>>"I read that book." Could be "F rId HAt bUk." or "F red HAt bUk."
>>
>>

I have been considering this very topic for a while now. The only
thing holding me back is my overall lack of time. My thought was to
take an existing phonetic breakdown of English and use that to
transliterate between latinized English and Shavian. The dataset I was
considering is from here:
http://www.speech.cs.cmu.edu/cgi-bin/cmudict. Has anyone looked at
this approach before?

--- In shawalphabet@yahoogroups.com, Ethan <ethanl@...> wrote:
>
> Thanks, Paul. Now, since I didn't catch a bunch of messages from that
> time last year, did anybody actually do anything with the suggestions
> that were made? Do we have such a thing as a Roman to Shavian
> translation program? I've been pondering the possibility of making
such
> a thing for a while now, even though I'm not much of a programmer
> myself. I kind of like the idea of having a program which can operate
> with a designated substitution list, or will make a list by prompting
> the user if none has been supplied ahead of time.
>
> Embedding the alternates in the the output text, using curly brackets,
> sounds like a good idea to me. It requires no user intervention during
> the conversion process, thus makes it possible to run lots of text
> through the convertor without interruption. Then the only thing
> necessary is to open the text with a text editor and do a search for
> curly brackets, and change each word to the one the context demands.
>
> I'm not quite sure I follow you regarding the adding of C, Q, and X to
> Shavian. Could you explain that a bit further?
>
> --
> Ethan
>
>
>
> paul vandenbrink wrote:
>
> >Hi Ethan
> >Better late than never.
> >I think that there are relatively few English words
> >that have two equally valid pronunciations inside a particular
> >accent group.
> >If we specify to the transliteration program, which accent group to
> >convert the
> >written T.O. English into, there will only be a small number of
> >exceptions. If there is a exception, we can embed both Shavian
> >spellings
> >into the text with curly brackets around them.
> >For example, the word perfect, becomes {pxf-ekt pD-fekt}
> >
> >I don't think this is much of hard-ship, and shouldn't prevent
> >automatic translation.
> >
> >There is another issue. Some words (i.e. names, etc.) and
> >abbreviations are not common enough to be registered in
> >the Database. By the way there are a lot of duplicate abbreviations
> >that represent different things. These
> >other Words that are not successfully translated, should be marked
> >with an asterisk and if possible retain their Roman spelling.
> >
> >How many additional letters, would have to be added to Shavian
> >to allow a Roman equivalent pronunciation.
> >Offhand, I would think you would only need a way to represent
> >C, Q and X in Shavian. All the other Roman Letters have Shavian
> >Equivalents.
> >That would better than having to fiddle with 2 Fonts.
> >
> >Regards, Paul V.
> >__________________________attached_______________________________
> >--- In shawalphabet@yahoogroups.com, Ethan <ethanl@> wrote:
> >
> >
> >>>>Star Raven wrote:
> >>>>
> >>>>
> >>>>
> >>>>>what about the homonyms? PERfect vs. perFECT ect?
> >>>>>
> >>>>>
> >
> >
> >
> >>>>It should be PERF-ect (Verb) vs per-FECT (Adjective)
> >>>>which not only has a different syllable boundary but also a
> >>>>distinct Shaw Spelling.
> >>>>
> >>>> pxf-ekt ----- pD-fekt
> >>>>
> >>>>There are going to be homonyms in Shavian, perhaps even more
> >>>>than T.O In Shavian, won = one
> >>>>
> >>>>Overall, it still will be miles ahead of T.O.
> >>>>
> >>>>
> >
> >
> >
> >>>This thread started out by suggesting that large amounts
> >>>of literature could be transcribed by using a computer
> >>>program and a look-up table. I believe Star's question
> >>>was how to handle these homonyms in a look-up table.
> >>>("one" vs. "won" is not a problem; only those which
> >>>can have a different stress.)
> >>>
> >>>As someone else pointed out, natural language processing
> >>>in English is not trivial. I would suggest that the look-up
> >>>table simply have a flag on those words with more than one
> >>>Shavian transcription. The program which does the trans-
> >>>scription could flag those words in the output. Then a
> >>>human could select the correct spelling. Thus 95% (or more)
> >>>of the text would be automatic, with a minimal amount of
> >>>human intervention .
> >>>
> >>>--Ph. D.
> >>>
> >>>
> >>>
> >>>
> >>My thoughts exactly. The flag can be as simple as having more than
> >>
> >>
> >one
> >
> >
> >>solution to a word in the lookup table. Portions of the lookup
> >>
> >>
> >table
> >
> >
> >>might look like this:
> >>
> >>Douglas = duglas, Gduglas
> >>douglas = duglas
> >>dour = dQD
> >>douse = dQs, dQz
> >>dove = duv, dOv
> >>dovecote = duvkOt
> >>dovekie = duvkI
> >>
> >>
> >>reactive = rIAktiv
> >>reactively = rIAktivlI
> >>reactiveness = rIAktivnas
> >>reactivity = rIAktivitI
> >>reactor = rIAktD
> >>read = rId, red
> >>
> >>When the text conversion program comes across the word "reactor",
> >>
> >>
> >it
> >
> >
> >>outputs "rIAktD" and goes on to the next word. But when it finds
> >>
> >>
> >the
> >
> >
> >>word "read", it sees two choices, which can be nearly impossible to
> >>ascertain from context even for people at times! The simplest
> >>
> >>
> >solution
> >
> >
> >>might be to pop up a dialog asking the user to choose the
> >>
> >>
> >output: "rId"
> >
> >
> >>or "red".
> >>
> >>Example:
> >>"I read that book." Could be "F rId HAt bUk." or "F red HAt bUk."
> >>
> >>
>

Or instead of just picking out curly brackets by eyeball, why not let
the program pick them out last, sort of a hold until it's done with the
easy part.

--Star in stormy Mid-TN

--- Ethan <ethanl@...> wrote:

> Thanks, Paul. Now, since I didn't catch a bunch of messages from
> that
> time last year, did anybody actually do anything with the suggestions
>
> that were made? Do we have such a thing as a Roman to Shavian
> translation program? I've been pondering the possibility of making
> such
> a thing for a while now, even though I'm not much of a programmer
> myself. I kind of like the idea of having a program which can
> operate
> with a designated substitution list, or will make a list by prompting
>
> the user if none has been supplied ahead of time.
>
> Embedding the alternates in the the output text, using curly
> brackets,
> sounds like a good idea to me. It requires no user intervention
> during
> the conversion process, thus makes it possible to run lots of text
> through the convertor without interruption. Then the only thing
> necessary is to open the text with a text editor and do a search for
> curly brackets, and change each word to the one the context demands.
>
> I'm not quite sure I follow you regarding the adding of C, Q, and X
> to
> Shavian. Could you explain that a bit further?
>
> --
> Ethan
>
>
>
> paul vandenbrink wrote:
>
> >Hi Ethan
> >Better late than never.
> >I think that there are relatively few English words
> >that have two equally valid pronunciations inside a particular
> >accent group.
> >If we specify to the transliteration program, which accent group to
> >convert the
> >written T.O. English into, there will only be a small number of
> >exceptions. If there is a exception, we can embed both Shavian
> >spellings
> >into the text with curly brackets around them.
> >For example, the word perfect, becomes {pxf-ekt pD-fekt}
> >
> >I don't think this is much of hard-ship, and shouldn't prevent
> >automatic translation.
> >
> >There is another issue. Some words (i.e. names, etc.) and
> >abbreviations are not common enough to be registered in
> >the Database. By the way there are a lot of duplicate abbreviations
> >that represent different things. These
> >other Words that are not successfully translated, should be marked
> >with an asterisk and if possible retain their Roman spelling.
> >
> >How many additional letters, would have to be added to Shavian
> >to allow a Roman equivalent pronunciation.
> >Offhand, I would think you would only need a way to represent
> >C, Q and X in Shavian. All the other Roman Letters have Shavian
> >Equivalents.
> >That would better than having to fiddle with 2 Fonts.
> >
> >Regards, Paul V.
> >__________________________attached_______________________________
> >--- In shawalphabet@yahoogroups.com, Ethan <ethanl@...> wrote:
> >
> >
> >>>>Star Raven wrote:
> >>>>
> >>>>
> >>>>
> >>>>>what about the homonyms? PERfect vs. perFECT ect?
> >>>>>
> >>>>>
> >
> >
> >
> >>>>It should be PERF-ect (Verb) vs per-FECT (Adjective)
> >>>>which not only has a different syllable boundary but also a
> >>>>distinct Shaw Spelling.
> >>>>
> >>>> pxf-ekt ----- pD-fekt
> >>>>
> >>>>There are going to be homonyms in Shavian, perhaps even more
> >>>>than T.O In Shavian, won = one
> >>>>
> >>>>Overall, it still will be miles ahead of T.O.
> >>>>
> >>>>
> >
> >
> >
> >>>This thread started out by suggesting that large amounts
> >>>of literature could be transcribed by using a computer
> >>>program and a look-up table. I believe Star's question
> >>>was how to handle these homonyms in a look-up table.
> >>>("one" vs. "won" is not a problem; only those which
> >>>can have a different stress.)
> >>>
> >>>As someone else pointed out, natural language processing
> >>>in English is not trivial. I would suggest that the look-up
> >>>table simply have a flag on those words with more than one
> >>>Shavian transcription. The program which does the trans-
> >>>scription could flag those words in the output. Then a
> >>>human could select the correct spelling. Thus 95% (or more)
> >>>of the text would be automatic, with a minimal amount of
> >>>human intervention .
> >>>
> >>>--Ph. D.
> >>>
> >>>
> >>>
> >>>
> >>My thoughts exactly. The flag can be as simple as having more than
>
> >>
> >>
> >one
> >
> >
> >>solution to a word in the lookup table. Portions of the lookup
> >>
> >>
> >table
> >
> >
> >>might look like this:
> >>
> >>Douglas = duglas, Gduglas
> >>douglas = duglas
> >>dour = dQD
> >>douse = dQs, dQz
> >>dove = duv, dOv
> >>dovecote = duvkOt
> >>dovekie = duvkI
> >>
> >>
> >>reactive = rIAktiv
> >>reactively = rIAktivlI
> >>reactiveness = rIAktivnas
> >>reactivity = rIAktivitI
> >>reactor = rIAktD
> >>read = rId, red
> >>
> >>When the text conversion program comes across the word "reactor",
> >>
> >>
> >it
> >
> >
> >>outputs "rIAktD" and goes on to the next word. But when it finds
> >>
> >>
> >the
> >
> >
> >>word "read", it sees two choices, which can be nearly impossible to
>
> >>ascertain from context even for people at times! The simplest
> >>
> >>
> >solution
> >
> >
> >>might be to pop up a dialog asking the user to choose the
> >>
> >>
> >output: "rId"
> >
> >
> >>or "red".
> >>
> >>Example:
> >>"I read that book." Could be "F rId HAt bUk." or "F red HAt bUk."
> >>
> >>
>
>
>


=========
http://www.livejournal.com/users/wodentoad

An idle duck is the devil's playground.

__________________________________________________
Do You Yahoo!?
Tired of spam? Yahoo! Mail has the best spam protection around
http://mail.yahoo.com