[liblouis-liblouisxml] Re: A New LibLouis
- From: "John J. Boyer" <john.boyer@xxxxxxxxxxxxxxxxx>
- To: liblouis-liblouisxml@xxxxxxxxxxxxx
- Date: Mon, 28 Oct 2013 09:12:39 -0500
Keith,
I agree that this is a very bad time to start talking about a massive
overhaul or even starting over. It is distracting me from fixing the
bugs in emphasis and Nemeth translation.
It is not true that everyone shudders at digging into the liblouis and
liblouisutdml code. Several programmers besides myself have made
significant contributions.
I think that a lot more thought needs to be given to the needs of the
various stakeholders and to the algorithms that will achieve those goals
before we even consider changing programming languages.
John
On Mon, Oct 28, 2013 at 12:37:34PM +0000, Keith Creasy wrote:
> Hi folks.
>
> First of all thank you everyone for the great discussion. I am still reading
> and trying to get my head around what everyone is saying. You guys got very
> busy while I was out on vacation.
>
> Now, having said that, this is a very bad time to be spending resources
> taking what is in reality an excellent tool and starting over. In my opinion
> there is no more powerful and cross-platform language than C and C++. It is a
> bit more difficult to learn and write well in but anyone talented enough to
> write code for LibLouis should not find it to be a great challenge. What we
> need urgently is for someone to help write up-to-date documentation. That
> isn't fun and very few people enjoy it but it is the one thing that makes
> working with these libraries difficult. Please consider contributing there.
>
> I do understand that we need some changes and improvements. First on my list
> is separating translation, style, formatting, and pagenation. These are
> entirely different procedures and should not be mingled.
>
> Some sort of support for regex might be useful.
>
> The native or default encoding for LibLouis should be UTF-8 and UTF-16.
>
> I'm not so sure about a scripting language. Basically a scripting language is
> a way to call various functions with a definite syntax. Better I think is to
> determine what we need a scripting language for and then decide how to solve
> the problem It might lead to a scripting language but it might just lead to
> some new functionality in LibLouis to fix the problem.
>
> Finally, I'd like to see LibLouisUTDML be entirely XML based. It should just
> use the low-level LibLouis functions to do translation. Then, depending on
> the settings perhaps use other LibLouis functions for styling. It should use
> namespaces properly and use the namespace and doctype to determined what
> semantic rules to apply rather than just using the file extension or root
> element except as a fallback. Semantic action files should be correct and
> based on the DTD or schema for the document type to which it refers and never
> on a sample document.
>
> I'd like to see LibLouis and LibLouisUTDML migrated to C++. This offers some
> of the advantages of higher-level languages and improves modularization and
> encapsulation. For large applications C gets very messy very quickly if great
> care is not taken.
>
> Remember that we are aiming for a March, 2014 release of the new design of
> BrailleBlaster. We have a relatively short list of things that must be fixed
> in LibLouis and LibLouisUTDML in order for that to happen. When it does
> happen I believe the interest and excitement surrounding BrailleBlaster,
> LibLouis, and LibLouisUTDML is going to explode in the U.S.
>
> Those are just my initial thoughts.
>
> Thanks.
>
> Keith
>
--
John J. Boyer; President, Chief Software Developer
Abilitiessoft, Inc.
http://www.abilitiessoft.com
Madison, Wisconsin USA
Developing software for people with disabilities
For a description of the software, to download it and links to
project pages go to http://www.abilitiessoft.com
Other related posts:
