There are two references to a software (distribution?) called KerTeX here on TeX.SX (one, two). Is this more an alternative to TeX Live ore more to TeX itself? Is there any benefit of using KerTeX over (whatever we use today in TeX Live 2011)?
Since I'm the author of kerTeX, I will put some precisions about the aim of kerTeX, the present state of kerTeX and the future of kerTeX.
The first aim of kerTeX is and will remain to be able, easily, to obtain what Donald E. Knuth has given us: the TeX system, that is not only TeX but also METAFONT and the fonts. Since the aim of Donald E. Knuth was to be free, to be able to write and produce his books without relying on anyone else or everything else---so that he will not hear anymore that it was impossible to produce his books the way they were done before since the fonts and the layout depended on a technology now orphaned---I find quite scandalous that this series of tools is more and more difficult to obtain and use on whatever system, because the needles are lost in a haystack.
This was the first; and this is still the main purpose of kerTeX.For having a quite confortable minimal system, not only D.E.K.'s programs are provided, but e-TeX (for right to left), MetaPost, bibtex and, of course, dvips plus AMS fonts are provided also.
This goal means too that if I need feedback to adjust the tools for systems I do not use---and I don't want to have the obligation to install all the flavours of all the existing systems because even reporting is too much an effort for people!---kerTeX, by purpose, excludes no system. The minimal requirement is a libc. For building, a subset of POSIX utilities (minimal subset). This means that by cross-compilation, almost every system can be supported. Windows can be supported via cross-compilation with Mingw with some small adjustements---I have simply not the time nor a personal need to focus on that now.
For the future, the next step will be unicode support via utf-8. But I do think that this can be done without huge changes to the core of the TeX program---for the fonts, METAFONT can stay the same, and the "support" will be an external one with tools.
Unicode via utf-8 means some changes relating to the tfm. But once more, this can be done easily by keeping a tfm a 256 glyphes subset, but using a font as a directory. (More on this later.)
What I do not want is to plague kerTeX with external dependencies that will prevent the use of D.E.K.'s programs if these external dependencies are not satisfied. This does not exclude modifications or extensions, as long as the core, the kernel is still available.
I don't want to switch from DVI to PDF natively for this very reason and for licence or copyrights reasons: I don't want to be unable some day to use the programs because some lawying gangs frighten the indirect use of PDF because of some claimed patent infringement.
I have put aside the needles from the haystack. What a huge majority of people will discover is that these needles are in 95% of the cases all they use or need. And it happens that the remaining 5% can be covered without depending on gigabytes of external things.
I hope this clarifies things. (Please don't expect me to participate a lot in threads, since I'm rather busy, with KerGIS, kerTeX and all the rest as, I hope, one can imagine...)
KerTeX is a minimal source distribution from which TeX and related tools can be built. The aims of producing KerTeX seem to be:
To that end, the source distribution does not include any pre-packed (La)TeX sources or documentation (for example no
The author of KerTeX has also stated some ideas on for example using material outside of the 8-bit range, but based on an approach using multiple fonts and file layouts rather than for example loading system fonts and using UTF-8 input. Thus the project is in some ways more closely tied to Knuth's original TeX rather than later projects such as pdfTeX, Omega, XeTeX or LuaTeX. The author has also made suggestions about a packaging approach, again based on file layout rather than a TeX Live or MiKTeX-like method.
In terms of benefit, it partly depends on what you want. The KerTeX project is very much focussed on the needs of the community the developer is in, where a small source distribution with suitable license is required. It seems unlikely that KerTeX will pick up on ideas such as direct PDF output, UTF-8 input or loading system fonts. Thus the approach is likely to appeal to users with particular needs.