[Journal] Translation of the journal

[Paul Kelly]

On Mon, 28 May 2007, Tyler Mitchell (OSGeo) wrote:

[...]
> Hi Paul, thanks for speaking up.  This make sense to me and is sort of what I 
> thought, but the I'm pretty new to the whole text encoding world.  Yves, I 
> guess using this method for accented characters is likely the most complex 
> thing a translator would have to do (other than maintaining other

I was just thinking (again speaking from how I understand LaTeX works) 
that perhaps also different hyphenation dictionaries will need to be 
loaded for each language. LaTeX has a dictionary containing most common 
words along with suitable places in them to place hyphens, if it needs to 
break words over lines. I'm not sure if dictionaries for languages other 
than English are included by default: I would guess it would depend on the 
specific LaTeX distribution used. Just that it's one other thing to 
consider (this time not for the translators I guess, but for whoever 
generates the PDF copies). Regardless of whether the dictionaries are 
included by default or need to be downloaded and installed, I would 
imagine the language used would need to be specified in the preamble to 
the document so that LaTeX knows which hyphenation dictionary to use.

> This is one of the reasons why we chose the Creative Commons, By Attribution 
> and No Derivatives licence.  So that we would really have some measure of 
> control over the translation process.  Of course, I don't expect all the 
> initial editorial team to be able to handle the additional workload - having 
> more people involved will really help, as long as they engage with our 
> Journal translation process.

Ah OK I see that now. So when authors submit articles they are explicitly 
giving permission for the OSGeo journal editorial team to publish 
translations, and no one else. That would seem to tidy that up nicely.

>>> And that would also make it easy for all the typeset final versions in 
>>> different languages to all be generated by the same person
>
> Could you elaborate on what you mean here?

Actually by "generated by the same person" I really meant using the same 
workflow on the same machine with the same versions of software tools 
installed etc. - I was thinking perhaps that was necessary to ensure a 
standardised consistent layout between different translations. And also 
making the point that because all LaTeX source is in plain ASCII text, it 
shouldn't need a variety of different set-ups for generating the different 
language versions - a standard LaTeX installation should be fine (although 
given the point about hyphenation dictionaries above).

But now thinking again, I see that an obvious follow-on from that is that 
if everything is clearly specified and documented with regard to which 
LaTeX packages are required to be installed, which PDF generator to use 
etc., *anybody* should be able to process the LaTeX code and generate the 
same final PDF copy of the journal. And the version downloadable from the 
OSGeo site can be regarded as the only "official" version. Perhaps md5sum 
signatures could even be included for the full PDF of each volume (and 
each translation) so people can verify (if necessary) that they have the 
orginal unmodified version?

Paul