Appendix C: Maintenance guidelines
Tips and tricks on how to perform maintenance on your resources
Translation work can get complicated from time to time. There is one thing that can definitely help you make it easy and more structured again - order in your resources.
When using your resources, make sure to check if all the below aspects are tackled, so that your work would be faster, easier, and safer.
Keep your resources in the right place
When planning where to store your resources, keep in mind that memoQ is not compatible with network drives and cloud storage services like OneDrive or Dropbox.
When you have a TM attached to a project, memoQ is actively using it in the background during your work. Unfortunately, external drives can lose connection or start synchronizing when you are translating and, in the end, corrupt your files. Therefore, keep your projects and resources in a local folder on your computer drive.
Want to learn more? See the documentation topic about location settings.
Clean up translation memories regularly
A translation memory's size depends on the number and the length of its entries (translation units) and some of its properties.
When your TM gets too big (around 1 million entries), it can cause issues like resource corruption or an occasional software crash.
Additionally, when translating big or many similar files, you can end up with repeating entries across different documents. These are called duplicates.
You can create duplicates in your TM by:
confirming a lot of identical translations to it, using different contexts
using a translation memory that allows multiple translations for the same source segment and the same context
importing TMX files into the translation memory with many identical source segments
importing a TMX file more than once
confirming a lot of non-text segments (segments made up from numbers, dashes, tags, etc.)
The Filtering and sorting window opens. Here, you can apply filters and sorting steps to get to the entries you need. When you are done, click OK.
The TM editor opens, showing segments based on the filters you chose. Then, on the Translation memory editor ribbon, click Remove Duplicates. The Filter for duplicates window opens - that’s the place where the magic happens. You can tell memoQ here what to treat as a duplicate, and what to do with the duplicates once they are found.
Things that can have an impact on the TM editor's performance:
how many entries are in the TM
how many entries are duplicates or contain only numbers or tags
if the TM folders are stored on an SSD (Solid-State Drive)
how many applications are running in the background
if your TM folders are added to your antivirus app as “exceptions” so it won’t block memoQ from accessing the TM
What to do with TMs you no longer need?
Only delete a translation memory if you are absolutely sure you do not need it anymore. Otherwise consider these options:
Shrink it by merging or deleting duplicate entries (see above).
Remove entries older than a certain date.
On the TM editor's ribbon, click the Filter/Sort button.
Choose a Before date, and export the entries created or modified earlier into a TMX file.
Press Ctrl+Shift+A to select all visible segments.
On the ribbon, click the Delete button, then the Save changes button.
Close the TM editor.
On the Resource console, import the TMX file into another translation memory.
Now you can use a smaller TM in your projects, and add the other one (with the old entries) whenever you need it.
To delete a TM, right-click it on the Resource console or in Project home, and choose Delete from the context menu. To confirm, click Yes under the confirmation message.
After project wrap-up, confirm and update all segments to the master TM
After project wrap-up, delete the working TM.
Keep your term bases nice and clean
The size of your term base depends on the number of entries, number of languages, and images added to it.
While TMs have a universal format (TMX) that transfers well between different CAT tools, TBs have multiple ones (for example, comma-separated CSV files, Excel sheets or XML files). Also, TBs in different CAT tools have different fields and data types. Therefore, it is easy to grow your TB's size or create duplicates when working on it.
You can create duplicates in your TB by:
importing entries from a file
using the Quick add term command while adding terms from the translation editor.
What are the best practices for working with term bases?
keep them below 2 GB
delete the ones you no longer need
add only small images to term bases
clean up duplicates regularly.
How to clean up duplicates in your TB?
In the Project home window, click Term bases.
Right-click a term base, and in the context menu, click the Edit button.
On the Term Base Editor ribbon, click the Show Duplicates button, and start checking your entries.
Want to learn more? Read the documentation topic about filtering for TB duplicates.
Check your LiveDocs
If you use LiveDocs corpora to align files and distribute reference material, keep track of how fast they grow.
Aligning two 900-KB DOCX files would result in a corpus of 2 MB. After editing the alignment and marking it as finished, the size grows to 2.4 MB.
What are the best practices for working with LiveDocs?
Abandoned corpora can take up a lot of space. Delete unused ones or those you created for a specific project - after that project is finished, consider exporting them before deleting.
Sometimes LiveDocs corpora get corrupted. You can restore your work by importing your alignment pairs and aligned documents into a freshly created corpus.
To prevent corruption, export your corpus to a TM after you finished working on it.
Do you need more help or got more questions? See this page for more information.