> I still believe adding matching with phrase-TMs would be a great feature.
If your feature "just" consists of a separate repository of phrases that
are suggested (as with current glossary terms), I don't see any great
advantage over using the existing glossary function for this purpose.
The restrictions of the glossary function (such as the lack of keyboard
shortcuts for selecting and inserting terms/phrases) could be eliminated
by adding those functions to the glossary function. I can envisage that
being conceivable and not particularly expensive, though I may be wrong.
Where the sub-sentence matching feature becomes particularly valuable is
when it is used for automatic composition of a segment. That is when,
for instance, you have a segment that is made up of four different
parts, each of which is identical or almost identical to another
segment, and the CAT tool attempts to compile a translation from them.
That, though, is something quite different, especially in terms of the
> I won't pay anyone for coding the feature, but I will give you an
offer: I promise to translate the user manual of OmegaT to Swedish, if
anyone is willing to code phrase-matching with sub-strings of the
current sentence (as I have elaborated in previous posts).
There you have the problem. That's not the way the project works. I'm
only the co-ordinator: I don't develop features or decide what features
other people develop. That's the task of the development team, and even
they only develop what an individual developer wants to develop. If you
want a feature and there is no developer who wants it for themselves,
you will have to make an attractive offer to persuade them. I have paid
our developers several times for features myself. You can try tempting
them with a Swedish translation of the user manual, but from my
experience they prefer dollars or euros. ;-)
On 04/03/2012 09:00 AM, Per Tunedal wrote:
> Hi Marc,
> I still believe adding matching with phrase-TMs would be a great
> feature. I might add it as a RFE and wait. Maybe I am ahead of my time
> or maybe phrase matching is already passé.
> I won't pay anyone for coding the feature, but I will give you an offer:
> I promise to translate the user manual of OmegaT to Swedish, if anyone
> is willing to code phrase-matching with
> sub-strings of the current sentence (as I have elaborated in previous
> Per Tunedal
> On Mon, Apr 2, 2012, at 15:57, Marc Prior wrote:
>> A few points.
>> The OmegaT project is already short of volunteers even to complete the
>> core tasks.
> -- snip ---
>> With regard to OmegaT's development: any feature requiring more than
>> minimum coding is added either because the developers see the benefit
>> for themselves, or because someone pays them to add it. Some features,
>> it's true, start as RFEs and ultimately are implemented, typically
>> because if a number of users express interest, sometimes over a period
>> of years, a developer begins to see the benefit to themselves. But
>> thinking that something would be a good feature is a long way from
>> actually getting it implemented.