I have a few observations on areas where CafeTran could potentially make life easier for new users:
A general observation - most functions should be opaque to the user - he should not have to understand how the glossary system or Total Recall works in order to use them. (I'm sure someone is going to tell me this is impossible, but it definitely isn't.) This does not appear to be the philosophy at present, but it should be.
For “power users”, there should be a button labelled 'Advanced' on the relevant Options tab, which should pop up a further set of more advanced configuration options. I think it would be helpful to move quite of few of the options on most tabs into 'Advanced' pop-up panels. Unfortunately, since it's unclear what many of the options do and there are no tool-tips (hint: add tool tips), I can't say which options are useful to basic users and which should be reserved for advanced users.
When I saved my first glossary, As well as having to specify the location for the txt fie, I also had to specify a location for a 'resource'. I still don't know what this is and I have no way of knowing what an appropriate location is. There should be a global setting for resource location (with an explanation of 'resources') – I shouldn't have to enter this when I save a glossary.
I probably want to sync this 'resource' between my home and office computers - or that could be disastrous – there's no way for me to know.
I love the idea of this, but in practice it seems
quite useless, producing garbage substitutions. It seems to be
pulling lots of entirely spurious fragment matches from my tmx files.
Am I doing something wrong?
There's been a lot of chat on the forums that it's better to use tmx files for glossaries, rather than the flat text files used by the glossary system. If that's the case, this should be the default option. (The glossary system should use tmx files by default.) And this should be opaque to the user – this really isn't something the user needs to worry about.
I love the idea of this, but in practice it seems quite useless, producing garbage substitutions. It seems to be pulling lots of entirely spurious fragment matches from my tmx files. Am I doing something wrong?
You are not satisfied with the auto-assembled translation, there are several things you can do to improve it.
1. If your glossary has an entry that contains two or more synonyms, place the best one in the first position.
2. Assign high priority to the glossary or TM which seems the best, and assign medium or low to others.
3. If you don't want to use a certain glossary or TM for auto-assembling, right-click on that pane and choose "Keep out of auto-assembling." Then, the matches are shown normally but are not used in auto-assembling.
4. Go to Edit > Preferences > Auto-assembling and choose "High priority for glossary" if you want CT to consider glossary matches before TM matches.
5. If you don't want the auto-assembled translation to be automatically inserted into the target segment, go to Edit > Preferences > Memory and lower "Auto-assembling insert threshold (%)."
amos: …he should not have to understand how the glossary system or Total Recall works in order to use them
Agree. However, a minimal, non-technical explanation would be useful, I think. Like:
TM: The only resource you cannot do without. To be used for segments, can (and should) also be used for phrases/fragments, and single words
Total Recall: A database, to be used to search large resources very fast. Can also be used to create a TM with segments that are relevant for the current job from the database, or for additional information (synonyms, alternative translations, other languages, etc.).
Glossaries: For use with regular expressions only. Not suitable for beginners
For more on CT’s resources, see here
…I can't say which options are useful to basic users and which should be reserved for advanced users
Neither can I, but then again, I’ve been using CT for only six years now. I think that, apart from that Igor, nobody can.
When I saved my first glossary, As well as having to specify the location for the txt fie, I also had to specify a location for a 'resource'
It makes sense, but it’s highly unintuitive. Another reason not to bother newbies with glossaries.
Auto-assembling:…Am I doing something wrong?
Undoubtedly. Start with seeing AA not as something that will be inserted in your target language pane, but as an automatic way to select the best possible term/phrase from your resources. Then decide whether you want it (automatically) inserted or not. For an approach, see here
The glossary system should use tmx files by default
Masato: Glad to see this kind of nice topic!
You won't be glad for long.