It is time to start rolling out the Forerunner (aka Preview) version of CafeTran Espresso 2019. This update and the subsequent builds this year will make up the CTE 2019 version to be released in December/January.
The update can be performed via Drag and Drop as follows:
1. Run CafeTran.
2. Download 20181210_update.zip file from here and place it on your desktop. Do not unzip or rename the file after downloading.
Note: On Mac OSX system, Safari web browser may unzip the file automatically after downloading it. Before you download the update file, uncheck the following Safari option:
Safari > Preferences > General > Open "safe" files after downloading.
3. Drag and Drop the downloaded 20181210-update.zip file anywhere in CafeTran's initial screen - the Dashboard.
Alternatively, you can install 20181210_update.zip file via the "Install Update" button in the Help > About panel instead of dragging and dropping.
Important! Please complete all your translation projects in your current CafeTran version before updating.
What's new in this update:
Is there an option to tell t he assembly function to give the glossary a precedence over what it isolates on its own from the TM?
Specifically the glossary, I don't think so, but you can use two settings to finetune the auto-assembling suggestions in general:
- Use the context option "Keep out of auto-assembling" on any resource (TM, Glossary) that you don't wish to use for auto-assembling. So, you can keep out of auto-assembling all but your glossary, if that's what you want.
- Set the priority of each resource (TM or Glossary) to High, Medium or Low, according to the priority you want to give. Please note this affects both fuzzy matching and auto-assembling results. This is also found in the resource context options.
Hm. The „Keep out of auto-assembling“ can be a nice tool, but I think Piotr may speak of scenarios where this is not an option (e.g. false fragments or hits from the primary TM).
And unfortunately the High, Medium or Low setting does not always help.
There is still some room for improvement.
There is always room for improvement, but there are also other settings that come into play, depending on the scenario. To avoid making assumptions, Piotr, can you specify the usage scenario you have in mind? Is this a support question or an improvement request?
Another thing is to set the Matching workflow to Fuzzy instead of Fuzzy & Hits in TM options (to remove the calculation of hits from a specific TM). Yet another is to raise/tweak the thresholds found in Preferences > Memory.
@tre I agree. Cafetran could become adaptive and learn from the user. For example if it “translated” a word or phrase as some nonsense, and “saw" that the user corrected it, it could “make a note" that this is a better translation and use it in the future.
I think CafeTran already attempts to tackle this, although I'm sure it can be further refined.
Do you have "Automatic fragments adjustment" enabled in Preferences > Auto-assembling?
Introduced in 2017 Yeddi Update 15 (20170627), this a new and experimental feature picks fragments/terms from the current segment and updates their priority on the fly. It can also adjust to some slight changes in the fragments (e.g. plural forms) creating virtual fragment pairs if they do not exist in your resources. It also detects term’s synonym usage adjusting its priority automatically.
If you use Auto-assembling with this feature enabled, maybe some specific examples in the form of a support ticket might help the developer improve it.
...but we still can't see a proper segment comparison between TM and current segment
Do you mean customizing the way what's been added too is highlighted? Currently, by colour and strikethrough we can clearly see (only) what's been subtracted. Yes, it would be nice to have both.
The new update (2018121001) of CafeTran 2019 Forerunner is available with the following features:
Please download and apply the update file as described in the first announcement post to this thread. Thank you!