CafeTran

Hi!

I'm pretty new to CafeTran, been using it intuitively for some months which was fine, but now I have a big project for which I need to be able to use it more efficiently. I tried hard to find the info I need (knowledge base, video tutorials, forum etc) but haven't been successful. I'll be extremely grateful if someone can help me out...

I'm translating a book (a patient's guide). A good portion of the text was compilated from existing documents that have been made accessible to me. Also, two other translators working on the project have shared their TMs with me. After working on a couple of chapters, I realised that I have everything set up wrong, because the only things that seem to be working are the online ressources and MyMemory - the TMs aren't being refered to.

- On one of the forums I read about the "Big mama" memory concept, which appeals to me - I'd like to put all of my memories into one, that keeps growing as I go along, but in spite of racking my brains for days I haven't figured out how. 

These are some of the problems / questions I have:

- How do I set up a main, basic "Big Mama" into which I incorporate the rest?

- When I imported my collegues TMs, I used the "import TMX" button in the total recall column, which I now realise is wrong - they should be in the "Translation memories (TMX)" column, right? How can I now move them into my Big Mama?

- I want to align the bilingual source documents (the ones that the authors based the book on, a series of files on various topics), and first make one TM out of all the aligned files to share with the rest of the team, and then incorporate that into my Big Mama. I used YouAlign to align them, then edited them in CafeTran, but then I got lost.

- How can I create a TM out of the chapters I already translated so I can add that into the BM? Until now I had "Project Memory" and "Project Glossary" selected, which apparently was wrong as that created a separate memory for each file (my chapters are segmented into a series of sub-files, so I think a separate "Project memory" was created for every sub-file, and wasn't even being refered to for the subsequent files within the same chapter).

- How do I set my project up so that it is one project for the whole book, not for each file, amd so that the TM and Glossary grow as I go along?

Many many thanks in advance...

Hello,

Below you will find some answers to your questions.

- On one of the forums I read about the "Big mama" memory concept, which appeals to me - I'd like to put all of my memories into one, that keeps growing as I go along, but in spite of racking my brains for days I haven't figured out how. 

In case other users wonder what a Big Mama translation memory is, it is a general/global TMX that holds all (or many) individual TMs from your past projects and is either updated on the go or used as read-only and updated manually (by adding new TMX files to it) for more finetuned usage.

I a way, a Total Recall database can be used a kind of Big Mama TM. Either single global TMX or Total Recall can be used in that case, but TMX may have some advantages for relatively small general TMs (and bigger TMs can still be used as read-only with the preliminary memory matching feature).  

These are some of the problems / questions I have:

- How do I set up a main, basic "Big Mama" into which I incorporate the rest?

Copy your existing TMX files in a specific folder, preferably each file named in a way that makes sense to use. You want to be able to reuse these TMX files in the future.

The inside a CafeTran project with the correct language pair, use Memory > Open memories folder.

Then Memory > Save that memory as. Closer the project, and drag and drop the resulting TMX into the Dashboard. Select that memory when creating new projects. To add TMs to it, you can then use the Memory > Join memory menu action.

- When I imported my collegues TMs, I used the "import TMX" button in the total recall column, which I now realise is wrong - they should be in the "Translation memories (TMX)" column, right? How can I now move them into my Big Mama?

Total Recall is a database translation memory solution inside which you can create various database tables and import TMX data. You can then selectively recall all or a pertinent subset of this data (based on the source text content of your project) in the form of a single ad-hoc TMX file by enabling a Total Recall table in the Dashboard or via the Total Recall menu during a project (View > Show Total Recall menu if it is hidden).

It seems you have misused/misunderstood Total Recall.

You can use Total Recall as a Big Mama (especially if your assembled TM is big), or just use the Big Mama approach to TMs.

Either way, for your Big Mama TM, you will need to use the TMX data you have imported.

In the Big Mama approach, keeping an archive of your TMs is important, to be able to recreate your global TM at will, in case any issue occurs.

For a quick and dirty, but patchy solution, click on the gear icon in the Dashboard under the Total Recall column and deselect Recall in context. That way all the table content will be recalled in the form of a TMX. Make sure this TMX is NOT set as read-only (deselect read-only). Then Memory > Save memory save as. Then import that TMX into your big mama.

- I want to align the bilingual source documents (the ones that the authors based the book on, a series of files on various topics), and first make one TM out of all the aligned files to share with the rest of the team, and then incorporate that into my Big Mama. I used YouAlign to align them, then edited them in CafeTran, but then I got lost.

CafeTran Espresso offers an internal workflow to align documents, but in my experience this best works for small documents, For aligning projects, I prefer the LFAligner.

Just import the resulting TMX into you Big Mama TM as explained earlier.

- How can I create a TM out of the chapters I already translated so I can add that into the BM? Until now I had "Project Memory" and "Project Glossary" selected, which apparently was wrong as that created a separate memory for each file (my chapters are segmented into a series of sub-files, so I think a separate "Project memory" was created for every sub-file, and wasn't even being refered to for the subsequent files within the same chapter).

Do you create a separate CafeTran project for each chapter? This is suboptimal. I would either add all chapters (files) in a folder and create a multiple document project, or add new chapters to the existing project.

Project with Multiple Documents - https://cafetran.freshdesk.com/support/solutions/articles/6000112837-project-with-multiple-documents

To add new source files to an existing project (even single document): Project > Add document.

Warning: do NOT edit or move the source text documents that are copied inside each project folder that CafeTran creates. They are needed for exporting the translations.

Using "Project Memory" (and maybe "Project Glossary") is recommended and can be used in combination with a Big Mama TM or General glossary. In that case, you would prefer to keep your Big Mama TM as read-only and add any new project TM (Memory > Join memory) once the project is finished.

In your case, do you trust each Project Memory has all the project segments? If yes, copy each TMX to the big mama folder location. Otherwise, open each project and use Project > Export and Exchange > To TMX to export a chapter/project TMX, that you can then reuse.

- How do I set my project up so that it is one project for the whole book, not for each file, amd so that the TM and Glossary grow as I go along?

See link above (Project with Multiple Documents).

---

I hope this helps. Please note I won't make any follow-ups to this post, but other CafeTranslators might wish to add to this information.

If you need further personal assistance, please note I offer CafeTran consulting/training services, not affiliated with the official support options. You can look me up in ProZ.com.

Jean Dimitriadis

Hi Jean, wonderful! Many thanks for your thorough answer!