Re: [trans-ws] Interest in trans-ws

[Arle Lommel <arle@lisa.org>]

My interest in the Trans WS activities is two-fold:

(First off, please note that "translation" here includes localization.)

1. From the LISA/OSCAR angle I am interested in making sure that any new
standards in the translation sphere are compatible with the OSCAR standards
of TMX and TBX, as well as with XLIFF. (By compatible I mean that, at the
least, they can be used in tandem without causing problems with each other.)
I would like to see that TransWS standards also provide a mechanism for the
transportation (or reference to) of translation resources where appropriate
and that this mechanism be intelligent enough that this integration is
smooth and painless.

2. A few years back I was involved in a project (that really didn't go very
far) to define what we called Translation Request Packets (TRPs), a
standardized XML form in which a requester of translation could provide
details about a translation job that would be usable for bidding purposes
and would also allow for the inclusion of all job specifications in a single
TRP. In this idea job specifications would be transmitted in a single file
to a production team, rather than via various channels with the inherent
possibility of loss or corruption this entails. The TRP would build
transparency into the translation process if it were accessible to both
client and vendor and allow for assurance that processes and requirements
were held to. The model for this was very much the traditional job ticket,
but as an XML file rather than a paper tracking sheet.

At the time (1999) that the group I was in looked at this subject we came to
two conclusions: (1) the existing technology did not really allow for this
idea to work well; (2) the market demand at the time was minimal. Most
vendors were just beginning to examine how web portals could be used for
purposes of job acquisition, but the portals tended to be very simple and
opaque - i.e., a user could upload a file and get a quote, but would have
absolutely no idea what that quote really meant or how the translation would
be carried out. This might be fine in the context of an existing
relationship of trust between client and vendor, but without that the bids
meant very little.


I think that TransWS could greatly benefit the translation community if the
standard is broad enough to allow for multiple uses and needs. At the least
it needs to be usable for transmission of "job tickets" for both bidding and
actual translation jobs, and needs to allow for the transmission of
sufficient metadata about a translation that jobs can be handed off and
"plugged-in" to a vendor's workflow without requiring substantial manual
intervention and negotiation of technical details. If this job is only half
done I think any standard would be seen as yet another burden to comply with
rather than a benefit to the GILT community. If done right, I think that the
potential benefits to the GILT world are substantial.

-Arle