Hi Phuc,
That's all for now. I will already be very happy if you could accommodate
all my previous requests. To summarize:
- Documentation about the Khmer braille code. Preferably some official
document, but something unofficial like you provided for Vietnamese is OK
too. A standalone document is better than a link to a web page.
- Documentation about the Burmese braille code. (Same comments as above.)
- Tests for Vietnamese. Preferably all the rules should be covered with
tests. Ideally it should be possible to make changes to the table or to the
Liblouis code, while still being sure that the behavior of the Vietnamese
table is correct. As a very minimum there should be a test file that covers
a part of the rules.
- Tests for Khmer. (Same comments as above.)
- Tests for Burmese. (Same comments as above.)
Op zo 24 feb. 2019 om 12:19 schreef Dang Hoai Phuc <danghoaiphuc@xxxxxxxxx>:
Ok thanks Bert for your suggestions and help to check the tables. Please
let me know any further requests/questions you have.
Happy Sunday and great week ahead.
Phuc
On 2/24/2019 5:28 PM, Bert Frees wrote:
Hi Phuc, thanks for your answers.
My answers inline:
Op vr 22 feb. 2019 om 11:16 schreef Dang Hoai Phuc <danghoaiphuc@xxxxxxxxx
:
Hi Bert,
Sorry for confusing. I didn't include the original En-Ueb-math.ctb file
because I wanted to make a saperate copy in order to easily examine the
code for handling Vietnamese math by adding an extra file with the "vi"
prefix as vi-ueb-math.ctb. In that file, I dodn't add my credit for it,
it's credited by the American Printing House.
And in my vi-puncsdef.uti, I also wrote down below codes are borrowed
from UEB grade 1 table. Please help me to update the credits for all those
files.
I'm not so worried about the credits. There are not that much parts that
are copied letter by letter. But you copied several sections and then
modified the dot patterns, also in the Burmese and Khmer tables by the way,
and whenever I see something like this I get suspicious. It is just not a
good way of working because you copy everything including all the flaws and
unnecessary stuff. Much better is to start from a YAML specification.
By the way, I've been looking some more at the tables and am having
questions with some things. For example, in my-g2.ctb you are using the
match opcode extensively but it seems to me you could have simply used the
always opcode instead. When there would be tests, I would be able to
quickly verify whether replacing match with always would work.
To prepare for the next release, I think we can remove the vi-ueb-math.ctb
and just simply include the "en-ueb-math.ctb"
OK.
The people helping me test the translation result are just my friends and
they help me voluntarily as individuals. I just want to add the name of
their organization where they are working at. Please let me know if you
need further info about these people.
All I wanted to know is what the relation is between these people that you
mentioned in the tables and the organisations that you mentioned in your
email :). You never said who is working where, or if there was a relation
at all. If they are just individuals and it is not relevant where they
work, we don't need to add more information to the tables, but since you
mentioned organisations in your email I gathered it was somehow relevant.
Regards,
Phuc
On 2/22/2019 4:27 PM, Bert Frees wrote:
Thanks,
2. That doesn't really answer my question. Why wasn't it possible to
include the UEB table? If it is just for experimentation, and there are no
actual requirements, then why is it important to have exactly the table you
have right now?
3. Yes, sure, I can help you with the credits, but if you decide to keep
it like this I really need some tests for these parts. I want to understand
why it isn't possible to include the UEB table, or a part of it, and
possibly override some of it with Vietnamese characters.
4. But what is the relation with the mentioned institutions?
Op vr 22 feb. 2019 om 08:33 schreef Dang Hoai Phuc <
danghoaiphuc@xxxxxxxxx>:
Hi Bert,
Thanks for your message. Below are my answers for your questions/remarks:
1. The The reference link to the Vietnamese Braille code/rules can be
checked at:
http://saomaicenter.org/en/blog/rules-and-guidelines-vietnamese-braille
I can make it all in one file and send to you so you can help put on any
possible space.
2. For the UEB-math, I tried to explore different options if it's the
best way to utilize existing UEB=math table or we should develop our own
one for our Braille translation software. By the end, we create a new one
for the Ascii math and don't use the UEB-math table with our software
anymore. I just include it in the table for users to experiment if
possible. As I don't have much time to test with it deeply.
3. Yes, I did copy some parts from other UEB and make some adjustment to
correct with Vietnamese characters. Sorry for forgetting to put the info of
such parts. Can you please help me list the credits for that?
4. Fyi, the persons from Myanmar and Cambodia, they helped me in
checking the Braille outputs to see if the translation results are correct
in different kinds of cases.
Many thanks and please let me know if you have further
requests/questions.
Phuc
On 2/18/2019 9:09 PM, Bert Frees wrote:
Hi Phuc,
The link you provided before does not work anymore:
http://saomaicenter.org/en/blog/accessible-formats/rules-and-guidelines-
vietnamese-braille. Do you have an alternative link? Or better, do you
have a document that we could put on our own space?
I also have a few more remarks after looking more closely at your PR.
- Is it really necessary to copy the ueb-math table? Wouldn't it be
better to include the existing one? If it is really important that the
behavior of the table is stable, a better solution is to create regression
tests.
- It looks like you also copied parts of other UEB tables without
mentioning anything about it in your table, and you even gave yourself
copyright to it. You shouldn't do that. Again, I would suggest to either
include an original UEB table, or create regression tests.
- You mention in your tables that you got help from Benedict and Pyae
Phyopaing for Myanmar, and Khemara Ly for Khmer. In your previous email you
also mention Myanmar Assistive Tech Research & Dev Center and the Myanmar
Association of the Blind, and the Krausao-mei school in Cambodia. How are
these people related?
As a general remark, I strongly encourage you to write tests for your
tables because it has nothing but benefits.
<http://saomaicenter.org/en/blog/accessible-formats/rules-and-guidelines-vietnamese-braille>
Op ma 7 jan. 2019 om 22:36 schreef Bert Frees <bertfrees@xxxxxxxxx>:
Ok thanks.
Op ma 7 jan. 2019 om 18:11 schreef Dang Hoai Phuc <
danghoaiphuc@xxxxxxxxx>:
Hi Bert,
Thanks for your help pulling the tables to LibLouis.
Fyi, the Vi-south.ctb table is actually the full contracted
Vietnamese. As you might know, we had two political systems before 1975
and
we have two different contracted Braille versions. The uncontracted
version
is used same for the whole country.
The vi-vn-g1 and vi-vn-g2 are officially used by the government.
However, the South one is still used by many visually impaired people
based
in the South so I decided that it should be given as I believe they have
right to choose to read with any Braille codes that they want. For the
official document of Vietnamese Braille codes, I have published on our
website at:
http://saomaicenter.org/en/blog/accessible-formats/rules-and-guidelines-vietnamese-braille
For the Braille tables for Myanmar and Khmer languages, I got the
Braille signs from the colleagues: Myanmar (Myanmar Assistive Tech
Research & Dev Center and the Myanmar Association of the Blind) and the
Khmer from the Krausao-mei school in Cambodia.
Please let me know for any further info.
Best,
Phuc
On 1/7/2019 10:47 PM, Bert Frees wrote:
Hello Phuc,
Thanks! That sounds great. I have made a pull request from your files:
https://github.com/liblouis/liblouis/pull/683. I have also added some
observations. Please have a look.
I have some questions regarding the vi-south-g2.ctb table. Is this a
variant of the other Vietnamese braille system, or should it be considered
a completely different system? Is the difference explained by a difference
between the dialects, or is it purely historical and could both regions
use
the same braille system? Can the regions where this system is used be
pinpointed exactly? Is it fully contracted braille or should it be called
partially contracted, in accordance with vi-vn-g2.ctb?
Some official documentation would help. I am looking for the best way
for a computer to distinguish the braille systems. If the regions where
they are used can be identified exactly, we can use that. Otherwise we
need
to give the systems (or variants) distinct names.
Op vr 14 dec. 2018 om 13:15 schreef Dang Hoai Phuc <
danghoaiphuc@xxxxxxxxx>:
Dear Bert, Christian and all Liblouis Developers,
I would like to contribute new translation tables for Burmese, Khmer
and
Vietnamese languages which I have created recently. Please give me a
hand to check and commit to the Github. The link to download tables:
https://www.dropbox.com/s/972hy2yypjgxx6o/Braille%20Tables.zip?dl=0
Fyi, it contains the following tables:
For Khmer (Cambodia): kh-g1.ctb (uncontracted Braille)
Burmese for both uncontracted and contracted: mm-g1.ctb and mm-g2.ctb
Vietnamese for uncontracted, partial contracted, full contracted and
South contracted version (used by many adults in the Southern part of
Vietnam): vi-vn-g1.ctb, vi-vn-g2.ctb, vi-vn-g3.ctb and vi-south-g2.ctb
So the total is 7 new tables. I haven't checked much with backward
translation so your help should be highly appreciated.
With best regards,
Phuc
