►
Description
Presented by
Richa Kumar, Sr. Director, Software Engineering @ GitHub
Kailash Nadh, CTO @ Zerodha
Nikhil Malhotra, Chief Innovation Officer @ Tech Mahindra
India is a land of many languages. While English is a common language for software development and open source, making open source more accessible to the non-English audience in India is important to advance the ecosystem. In this session, Richa Kumar from GitHub chats with the creators of two such projects—BHAML, an open source markup language for India, and Olam, an open source regional language dictionary—on the role of regional languages in open source, the challenges they faced in the development, and how other developers and contributors can make open source software more accessible to the masses.
As always, feel free to leave us a comment below and don't forget to subscribe: http://bit.ly/subgithub
Thanks!
Connect with us.
Facebook: http://fb.com/github
Twitter: http://twitter.com/github
LinkedIn: http://linkedin.com/company/github
About GitHub
GitHub is the best place to share code with friends, co-workers, classmates, and complete strangers. Millions of people use GitHub to build amazing things together. For more info, go to http://github.com
A
A
So
platforms
like
github,
provide
tools
and
capabilities
for
developers
anywhere
in
the
world
to
solve
problems,
learn
from
each
other
share
their
work
and
collaborate
towards
common
goals.
Today,
technology
is
at
the
heart
of
everything,
and
access
to
technology
is
critical
in
helping
us
solve
problems
and
to
innovate
in
any
area.
A
A
Luckily,
we
have
with
us
today
two
very
accomplished
people
who
have
been
thinking
about
this
problem
and
are
leading
innovation
in
this
space.
So
join
me
in
welcoming
nikhil
malhotra,
head
of
innovation
at
tech,
mahindra
and
creator
of
the
bharat
markup
language,
which
is
an
open
source,
markup
language
for
india
and
kailash
nad,
a
cto
of
zeroda,
hai,
kalash
and
creator
of
olam
and
lr,
which
are
open
source,
regional
language,
dictionaries,
nikhil
and
kailash.
We
are
eager
to
hear
about
you
know
what
has
what
have
been
some
of
the
motivating
factors
for
you.
A
B
I
sure
richard
first
of
all,
thank
you
so
much
to
github.
I
know
this
is
an
open
forum
in
an
open
community
and
then
taking
it
forward
to
this
community
is
a
is
a
great
privilege
as
well.
I
think
the
problems
always
been
there
richard
if
you,
if
you
notice
right,
the
90
percent
of
us
in
india
speak
vernacular
and
we
are,
you
know
about.
10
percent
of
us
are
only
well
versed
in
english
language.
B
My
inspiration
comes
from.
You
know,
one
of
the
mentorship
that
I
was
doing
with
the
utterly
incubation
centers
with
you
know
a
few
of
the
kids,
and
this
was
a
program
called
a
student
innovator
program
or
student
to
society,
and
we
had
students
from
both
urban
and
rural
sectors
and
one
of
the
you
know.
The
teams
that
won
was
a
rural
sector
team
from
chhattisgarh.
B
These
four
students
used
to
walk
about
45
kilometers
every
day
to
go
to
school,
and
you
know
you
call
it
scientific
serendipity,
you
call
it
thought
they
dropped
in
red,
amaranthus,
green
amaranthus
and
some
kind
of
bhajra
on
a
piece
of
land,
the
wheat
that
grew
out.
They
actually
took
juice
out
of
it
and
that
particular
jew
started
solving
a
problem
of
malnutrition
in
that
area.
This
project
has
been
now
under
the
eyes
of
unicef
and
people
are
looking
at
it.
B
So
I
was
amazed
by
the
ingenuity
and
of
course
there
was
a
challenge
to
solve
and
these
guys
were
solving
it,
but
then
they
came
to
us
for
mentorship
and
we
gave
them
mentorship
regarding
how
can
you
put
it
into
a
packaged
product?
How
can
you
put
it
into
seeds,
format
etc,
but
one
of
the
students
asked
me,
sir.
B
I
want
to
learn
technology,
but
I
don't
know
english,
so
that
was
possibly
the
starting
point
that
got
me
thinking
about
an
year
back
saying:
should
india
not
have
the
ability
or
should
indian
you
know,
indian
kids
not
have
the
ability
to
quote
something
in
their
own
language,
and
that
was
the
origination
point
of
bhamal
or
bharat
markup
language.
Now
what
does
brahmal
do
it's
actually
open
source?
It's
available
on
bamal.org
people
can
go
and
have
a
look
at
it.
B
But
for
the
moment
that
we
started
off,
we
said
that
let
kids
is
at
least
code
html
by
simple,
passing
and
transliteration
in
their
own
language,
so
we've
given
12
languages
for
kids
to
code
on
it's
hindi,
marathi,
punjabi
gujarati,
and
then
we
said
that
you
know.
Once
we
go
through
this
phase,
let
kids
try
it
out.
B
A
That's
awesome,
nikhil,
it's
great
to
see
the
the
from
where
that
you
know
motivation
started
and
where
you've
been
able
to
take
it
wonderful
to
hear
about
that,
can
kailash
what
were
the
motivating
factors
behind
the
regional
language
dictionaries
that
you
have
started.
I
know
this
has
been
an
effort.
That's
been
going
on
for
some
time,
but
would
love
to
hear
what
kind
of
prompted
you
to
do
that
and
what
you
feel
has
been
the
impact
or
can
continue
to
be
the
impact
of
some
of
this
work.
C
So
all
in
the
first
dictionary
that
came
out
came
out
10
years
ago,
almost
11
years
ago,
2010
and
it
was
born
out
of
my
personal
frustrations,
of
not
being
able
to
find
a
decent
digital
malayalam
dictionary
is
my
mother
tongue.
That's
what
I
speak
and
that
pushed
me
down
the
rabbit
hole
of
figuring
out
how
the
landscape
of
indian
online
dictionaries-
and
I
realized
it-
was
an
abysmal
landscape.
It
wasn't
anything
and
the
few
online
dictionaries
that
were
available.
They
were
riddled
with
ads.
You
know
the
quality
wasn't
great.
C
Commerce,
so
I
figured
that
there
should
be
an
open
data,
high
quality
dictionary
that
is
freely
accessible
to
everyone,
not
just
people
looking
at
definitions
in
a
dictionary
but
research
students,
anyone
it
should
be
open
data,
and
I
put
together
a
dictionary
with
a
few
thousand
words
that
I
entered
myself
and
I
found
a
few
data
sets
small
little
datasets
floating
around
on
the
internet.
Unfortunately,
unattributed
I
put
everything
scraped
everything,
everything
together
and
built
column.
It
was
in
english,
malayalam
dictionary
and
it
went
live
in
2010
and
today
it's
it's
still.
C
C
So
that's
the
story
of
poland
is
a
recent
when
I
completely
randomly
by
happenstance,
came
across
v
krishna,
who
spent
40
years
building
a
dictionary
which
nobody
really
knew
about,
and
he
was
gracious
enough
to
open
source
it.
So
we
took
it
the
biggest,
authoritative,
canada,
english
dictionary
and
it's
it's
called
allah.
Now
it's
available
again
openly
directionally.
So
that's
two
big
open
data,
regional
dictionaries.
A
Nice,
nice
great
to
hear
about
that
kailash
and
just
kind
of
opening
this
up
to
both
of
you.
You
know
what
have
been
some
of
the
challenges
that
you
know
you
all
have
faced
as
you've
sort
of
embarked
on
these
journeys,
and
you
know
how
have
you
you
know
what
tools
or
approaches
have
you
used
to
help
you
make
progress?
You
know.
B
So
I
I'll
take
that
question
richard.
I
think
the
one
of
the
biggest
challenges
is,
of
course,
lack
of
awareness,
and
I
think
I
would
be
very,
very
open,
saying
that
the
moment
you
get
into
a
coding
world,
there
is
a
kind
of
a
predistant
or
a
kind
of
a
predisposed
notion
that
these
are
the
languages,
and
this
is
something
that
we've
got
to
do,
and
that
is
something
that
I
think
programs
like
github
we
going
on
the
open
source
channel
would
really
help
us.
B
Take
it
forward
number
one
number:
two:
a
lot
of
people
are
not
aware
about
the
strengths
of
what
local
regional
languages
have
you
know
and
and
and
have
been
produced
as
well.
Like
kalash
really
mentioned
that
malayalam
there's
a
there's,
a
very,
very
valid
dictionary
that
is
now
made.
You
know
my
research
started
about
three
years
back
on
sanskrit
as
well,
and-
and
I
found
out
that
there's
so
much
to
talk
about
natural
language
processing
right
over
there
in
that
language,
even
though
it
may
not
be
spoken
very,
you
know
very
openly.
B
It's
only
spoken
in
small
parts
and
small
sectors,
but
sanskrit
can
give
us
that
ability
to
understand
the
root
of
language
as
well
as
how
the
natural
language
processing
systems
could
take
place.
So
you
know
if,
if
there
is
awareness,
then
there
is,
there
is
a
sort
of
you
know
the
people
do
understand.
What's
the
strength,
what
are
we
doing
for
it?
I
think
number
one
we.
I
continue.
B
To
talk
about
it
in
all
forms
letting
people
know
you
know,
I
continue
to
say.
I
continue
to
talk
about.
What's
going
on
with
the
language,
how
is
sanskrit
basically
going
on?
What
should
be
the
next
step?
Number
two
is
I've
got
some
I've
got
a
very,
very
good
team
that
is
also
now
taking
this
language
out
to
to
young,
kids
and
and
even
rule
sectors.
B
Very
interestingly,
I
in
my
lab
in
maker's
lab
there
is
a
gentleman
who
was
earlier,
an
uber
driver
who
actually
learned
over
by
the
dent
of
his
hard
work
or
whatever
python
c
c,
plus,
plus,
etc,
and
he
joined
the
lab
now.
Kovit
pushed
him
to
the
to
the
villages
he's
now
in
one
of
the
villages
in
maharashtra,
so
he's
taken
upon
himself.
His
name
is
satee
she's
taken
upon
himself.
He
brings
in
students
from
about
30
40,
odd
school
in
the
rural
sector,
and
he
starts
teaching
them
harmal.
He
starts
teaching
them.
B
You
know
the
coding
language
itself,
so
that
has
been
more
of
a
generic
effort
that
we've
been
doing
to
raise
that
awareness.
There
are,
of
course,
other
channels
that
we
are
utilizing
as
marketing
etc,
but
awareness
is
one
of
the
biggest
things
that
we
feel
is
the
challenge
at
this
point
in
time
got
it.
C
B
C
Twice
about
english
string
search,
you
know
you
do
control
f,
you
put
your
string
in
and
you
know
search
happens.
There
are
many
different
algorithms
to
do
english
string,
search
and
that's
that's
the
case
for
latin
languages,
but
search
for
indig
languages
is
an
understudied
research
area
and
without
academic
research
you
can't
really
build
tools
that
people
can
use
so
even
to
to
get.
C
C
Large.
There
are
large
companies
who
offer
free
computing
tools,
but
it's
definitely
awareness
and
I
think
the
site,
one
of
the
side
effects
of
language
input
becoming
prevalent
across
mobile
devices
everywhere
is.
I
think,
people
assume
that
the
indic
language
problem
is
solved
because
you
can
read
your
language
on
the
phone.
You
can
write
your
language
on
the
keyboard,
but
in
fact
the
fact
that
language
keyboards
are
now
standard
on
all
indian
operating
systems
was
in
fact
a
community
push
that
happened
five
six
years
ago.
C
A
So
how
does
that,
like
you
know,
is
that
you
know,
especially
as
you
worked
on
two
different
languages:
dictionaries
are
there
patterns
that
that
can
scale
out
to
other
languages?
Is
it
every
time
you're
adding
support
for
a
new
language?
Are
you
kind
of?
Are
we
starting
from
scratch?
You
know
how?
Maybe
you
can
give
us
a
little
insight
into
you
know
how
how
how
much
effort
is
involved
and
what
are
some
of
the
ways
that
people
could
contribute.
You
know
along
this
effort.
B
No,
absolutely
richard,
I
think
digital
content
and
content,
except
remains
a
challenge,
as
kailash
rightly
mentioned,
and
I
think
my
area
has
been
to
find
the
middle
layer
for
sanskrit
as
well
and
I'm
finding
some
patterns.
You
know
wherever
we
as
of
languages
across
the
across
the
subcontinent.
There
are
different
patterns,
but
then
again,
these
patterns
need
to
be
adjusted
to
that
specific
language.
B
Now,
just
to
give
you
an
example,
a
lot
of
people
and-
and
that's
where,
if
I
may
talk
to
the
community
in
terms
of
what
has
been
the
genius
that
panini
was-
and
I
think
panani's
genius
was
to
understand-
where
exactly
is
information
encoded
in
our
language,
so
how
much
of
that
information
is
encoded
and
what
is
there
exactly?
Which
needs
to
be
grabbed
up?
B
The
problem
is
that
if
I
replace
the
order
of
the
words
in
english,
for
example,
village
go
man
the
meaning
completely
changes
in
sanskrit
because
of
inflection
the
meaning
remains
the
same.
So
what
and
panani's
genius
was
to
actually
say
where
exactly
is
this
information
included?
It's
not
in
the
purusha?
It's
not
in
the
grammar,
it's
actually
in
the
t
of
gachati,
where
exactly
it
makes
that
verb
available.
B
This
specific
construct-
and
this
specific
rule
applies
to
almost
all
sectors
of
indict
languages.
They
may
be
different
here
and
there,
but
if
you
look
at
the
pattern
in
which
we
speak
and
the
way
the
languages
have
been
handed
over,
this
particular
pattern
applies
to
all.
There
are
certain
other
patterns
which
don't,
but
the
research
has
been
on
in
different
sectors.
B
I've
seen
colleges,
we've
been
we've
seen,
academia
work
on
it,
we've
seen
researchers
work
on
it,
but
there
are
some
very,
very
good
patterns
to
help
us
in
natural
language
processing
which
we
are
finding
within
indic
languages,
and
that's
been
a
revelation.
So
I
think,
if
more
and
more
people
join
this
community
of
open
source,
they
would
find
that
these
patterns
would
also
help
them
solve
some
of
the
challenges
for
natural
language
processing
in
english
as
well.
So
that's
what
we're
finding
out.
A
B
C
So
yeah,
indeed
languages
share
quite
a
bit
in
common
and
research
research
developed
for
one
language.
I
think
there's
a
significant
overlap
with
what
you
could
use
for
other
languages.
To
cite
a
concrete
example
for
the
malayalam
dictionary
to
work
I
had
to,
I
ended
up
putting
metaphor,
which
is
an
english
phonetic.
Hashing
algorithm,
really
simple!
You
you
take
a
word
elephant
apple,
it'll
encode
it
to
you,
know
two
or
three
letters
like
apl
or
ebnt.
So
the
idea
is
that
similar
sounding
words
will
all
hatch
to
the
same
set
of
strengths.
C
It's
called
metaphor:
it's
available
in
the
public
domain,
so
that
was
important
for
malayalam
and
that's
how
phonetic
search
from
that
malayalam
dictionary
works,
but
because
the
phonetic
nature
of
kannada
is
extremely
similar,
and
for
that
matter
you
know
in
the
gujarati
etc.
This
was
easily
voted
overnight
to
kannada
and
we
certainly
have
canada
phonetic
search
on
color.
C
So
a
lot
of
things
that
we
develop
for
one
indict
language
can
be
outright
applied
with
very
minor
tweaks
and
exceptions
for
many
other
languages.
The
and,
like
you
said,
I
wanted
to
add
one
thing
to
what
you
said
earlier.
There
is
widespread
availability
of
tools
like,
for
example,
there
is
google
translate
for
almost
all
major
indic
languages
today
and
there's
google
input
tools,
but
these
are
all,
unfortunately,
proprietary,
how
these
tools
work.
We
don't
know
what
are
the
algorithms
that
power
these
tools?
C
A
Yeah
make
sense
make
sense,
and
so
how
would
you,
if
I
ask
both
of
you
that
you
know
if
our
open
source
community
would
want
to
contribute
on
this
front,
both
specifically
on
your
projects
as
well?
As
you
know,
how
should
they
be
keeping
all
of
this
in
mind
in
their
own
efforts?
You
know
any
any
advice
for
the
folks
listening.
A
C
I
think
I'll
go
first,
so
all
the
all
the
components
that
have
gone
into
building
these
dictionaries
in
english,
the
data
sets
themselves
part
of
speech,
tagged
and
the
software
that
runs
the
website,
the
phonetic,
algorithms
and
power
source.
These
are
all
open,
source
and
they're
all
available
on
github.
So
for
someone
wanting
to
start
to
do
simple
things,
especially
surrounding
open
data
and
dictionaries,
they
could
think
of
it.
You
know
the
data
they
could
think
of
with
the
code.
C
They
could
even
use
some
of
these
tools
to
build
their
own
little
local
dictionaries.
For
example,
you
could
very
easily
with
minimal
effort,
build
a
hindi
technical
word
dictionary.
It
could
be
5000
or
10
000
words
so
or
even
a
couple
thousand
words,
but
you
know
technical.
You
know
scientific
mathematical
world,
so
even
small
efforts
like
that
building
niche
open
data
dictionaries
for
different
languages
would
be
a
great
start.
So
there
are
many
things
one
could
do
without
actually
getting
into
proper
linguistic
research.
B
A
B
B
A
B
Now,
you're
absolutely
right
richard,
I
think
it's
a
mixture
of
both
it's
both
translation,
it's
both
transliteration,
but
it's
also
about
looking
at
natural
language
processing.
But
I
think
my
vision,
the
vision
that
I
carry
within
india
is
that
at
the
end
of
the
day
on
silicon,
it's
just
so
what
we've
got
to
do
is
basically
do
the
layer
and
change
that
entire
layer,
so,
whether
it's
the
layer
of
compilers,
it's
the
layer
of
assembler,
it's
even
the
layer
of
the
os
that
should
essentially
be
changed
to
understand
the
language
in
its
most
prime
form.
B
If
that
happens,
you
can
put
that
as
an
os
addendum
on
anything.
It
could
be
a
version
of
linux,
it
could
be
freebsd
or
whatever,
but
that
will
really
change
the
impact
of
a
language
and
the
usage
of
that
language
as
well.
So
while
we
all
say
that
c
is
the
mother
or
the
foundation
of
all
in
or
you
know,
all
programming
languages,
could
we
develop
another
language
which
is
like
c,
but
it's
basically
based
out
of
all
of
these
regional
language
structures
as
well?
Could
it
have
a
completely
different
structure?
B
Could
it
have
a
completely
different
database
structure
as
well?
So
I
think
those
components
is
something
which
is
basically
driving
this
vision.
This
work
is
very,
very
primitive.
What
you
know,
and
I
would
like
to
call
my
work,
I'm
not
privy
to
actually
say
what
kailash
is
doing,
but
it's
very,
very
primitive.
It's
actually
scratching
the
surface,
but
the
open
source
community
now
has
the
ability
to
actually
take
this
forward
and
bring
it
out
in
a
much
more
bigger
and
brighter
way.
C
C
There
are
bi
standards
managing,
but
all
should
have
indig
languages.
So
the
open
source
communities
are
what
really
gave
us
the
indic
digital.
You
know
computing
language
revolution,
we
don't
see
it
because
we
only
interface
with
the
consumer
devices,
and
you
know
consumer
software.
So
I
I
would
say
that
the
question:
how
can
open
source
help
is
moved
in
the
face
of
open
source?
Is
what
really
gave
us?
You
know
this.
C
This
indic
languages
in
the
digital
form
to
begin
with
and
going
forward.
Also,
as
I
said,
you
know,
could
be
simple.
Algorithms
could
be
open.
Data
sets
all
these
building
blocks.
They
have
to
be
in
the
public
domain
for
true
innovation
and
accessibility
to
happen.
You
can't
really
build
if
you,
if
you
end
up
building
a
great
language
translation
model
and
you
keep
it
proprietary
you're,
holding
language
hostage
from
a
large
part
of
the
population
and
and
everything
is
going
online,
especially
critical
government
public
services.
C
B
No,
I
think
I,
I
think,
kalasha's
pretty
much
covered
everything
I
think
the
only
and
I
think
the
reason
is
for
everyone
to
see.
I
think
the
reason
we
are
having
and
discussing
this
today
on
a
github
satellite
is
the
reason
enough
for
us
to
say
that
this
is
the
power
of
open
source
community
and
that
power
of
open
source
community
can
actually
be
factored
in
yeah.
So.
A
So
it's
been
wonderful
talking
to
both
of
you.
It's
really
fascinating.
To
hear
about
the
work
that
you
all
have
been
doing.
I
am
very
very
excited
about
the
possibilities
that
are
ahead
of
us
and
I
hope
that
both
at
github
and
all
you
know
our
amazing
open
source
community
that
we're
gonna
take
away
some
important.