3 ms·

LaTeX3 Roadmap

39 points by gnosis 15y ago

jessriedel 15y ago
I honestly believe that the most cost-effective way to advance math and hard science research in the United States would be for some person or organization to put down $1 million to hire some developers to produce an enterprise-quality version of LaTeX, with an editor. The amount of grad student, post-doc, and professor time that is wasted dealing with LaTeX is staggering. There's just no reason grad students should need to spend, e.g., an hour to figure out how to get columns working.
- Anon84 15y ago
  I tend to use LyX (www.LyX.org) whenever I can get away with it (most of time). When it's time to submit, I export to LaTeX and check(fix) that everything is fine.
- sciurus 15y ago
  The next time I need to create that kind of publication, I plan to try ConTeXt to see if it is any better than LaTeX. "LaTeX was created with the idea of separating content and presentation to such an extent that the typical author would write their content and then use a style file created by someone else to provide the visual presentation. Thus, the proliferation of style files and related things. ConTeXt, on the other hand, retained the idea of separating content and presentation, but was created with the idea of being used for books, where each book tends to have a different layout, and so the expected 'end user' is the person doing all the layout. Thus, it's designed to provide a vast amount of flexibility for layout in a way that can be fairly easily defined without needing to write a package (or go find one that's already written)." "ConTeXt is a macro package that’s far more advanced than LaTeX. You can enhance LaTeX with third party packages, but not all macro packages work together with each other. ConTeXt is an integrated, powerful and flexible macro package for which you seldom need third party packages." http://wiki.contextgarden.net/Main_Page http://wiki.contextgarden.net/Main_Page
  - mturmon 15y ago
    The Context graphics have always looked excellent. Problem is, if you need to submit your publication to a journal, then you would have to convert it to ordinary LaTeX, and that would suck.
- Argorak 15y ago
  Beyond the fact that I worked with "enterprise-grade"-Software that wasted much more than an hour on columns... Let me make a daring statement: in some regards, (La)TeX is the most "enterprise-ready" document writing solution out there. The community is extremely committed to backwards compatibility - if you had something working in LaTeX, there is a high probability it wont break for the next 10 years (my files from 2000 still compile properly). It is reasonably fast for PDF generation, which makes it good for batch use (also, it still knows what "batch use" is, being from the 70s and all), making it a good solution to generate things like invoices. Newer Tex compilers like XeTeX and LuaTeX also eliminated most of the biggest wrinkles (most glaringly the inability to use any modern Font format like OpenType). Also, for interested programmers, it comes with a huge toolchain to manipulate and stitch together PDF, which I often use. The problem is the bar of entry: LaTeX is a _typesetting_ system, which means that all terminology comes from typesetting, not from programming, a fact that many programmers hate.
  - nvictor 15y ago
    exactly. it's a type system for which one has to invest some time. and (for the first comment) no, i didn't waste my undergraduate time learning it. it's precious knowledge to me even though i don't use it today.
ggchappell 15y ago
How seriously should we take this? The LaTeX 3 project is almost twice as old as Perl 6. It was in development for years even before the first announcement of Duke Nukem Forever. That's a long time, folks. Certainly, LaTeX could use an update. And I applaud those working on it. But really, is it going to happen?

39 points by gnosis 15y ago

I honestly believe that the most cost-effective way to advance math and hard science research in the United States would be for some person or organization to put down $1 million to hire some developers to produce an enterprise-quality version of LaTeX, with an editor. The amount of grad student, post-doc, and professor time that is wasted dealing with LaTeX is staggering. There's just no reason grad students should need to spend, e.g., an hour to figure out how to get columns working.

Anon84 15y ago
I tend to use LyX (www.LyX.org) whenever I can get away with it (most of time). When it's time to submit, I export to LaTeX and check(fix) that everything is fine.
sciurus 15y ago
The next time I need to create that kind of publication, I plan to try ConTeXt to see if it is any better than LaTeX. "LaTeX was created with the idea of separating content and presentation to such an extent that the typical author would write their content and then use a style file created by someone else to provide the visual presentation. Thus, the proliferation of style files and related things. ConTeXt, on the other hand, retained the idea of separating content and presentation, but was created with the idea of being used for books, where each book tends to have a different layout, and so the expected 'end user' is the person doing all the layout. Thus, it's designed to provide a vast amount of flexibility for layout in a way that can be fairly easily defined without needing to write a package (or go find one that's already written)." "ConTeXt is a macro package that’s far more advanced than LaTeX. You can enhance LaTeX with third party packages, but not all macro packages work together with each other. ConTeXt is an integrated, powerful and flexible macro package for which you seldom need third party packages." http://wiki.contextgarden.net/Main_Page http://wiki.contextgarden.net/Main_Page
- mturmon 15y ago
  The Context graphics have always looked excellent. Problem is, if you need to submit your publication to a journal, then you would have to convert it to ordinary LaTeX, and that would suck.
Argorak 15y ago
Beyond the fact that I worked with "enterprise-grade"-Software that wasted much more than an hour on columns... Let me make a daring statement: in some regards, (La)TeX is the most "enterprise-ready" document writing solution out there. The community is extremely committed to backwards compatibility - if you had something working in LaTeX, there is a high probability it wont break for the next 10 years (my files from 2000 still compile properly). It is reasonably fast for PDF generation, which makes it good for batch use (also, it still knows what "batch use" is, being from the 70s and all), making it a good solution to generate things like invoices. Newer Tex compilers like XeTeX and LuaTeX also eliminated most of the biggest wrinkles (most glaringly the inability to use any modern Font format like OpenType). Also, for interested programmers, it comes with a huge toolchain to manipulate and stitch together PDF, which I often use. The problem is the bar of entry: LaTeX is a _typesetting_ system, which means that all terminology comes from typesetting, not from programming, a fact that many programmers hate.
- nvictor 15y ago
  exactly. it's a type system for which one has to invest some time. and (for the first comment) no, i didn't waste my undergraduate time learning it. it's precious knowledge to me even though i don't use it today.

ggchappell 15y ago

How seriously should we take this? The LaTeX 3 project is almost twice as old as Perl 6. It was in development for years even before the first announcement of Duke Nukem Forever. That's a long time, folks. Certainly, LaTeX could use an update. And I applaud those working on it. But really, is it going to happen?