Please make sure to update to WPML 4.3.5 and check our list of Known Issues before reporting

Hi, Amit here, I am the WPML Support Manager, our current ticket queue is high, update your WPML plugins and make sure you meet the minimal requirements for running WPML before reporting an issue please - many tickets are resolved doing that

Please look at our updated list of Known Issues and you can also use our support search to find helpful information and of course review our documentation before opening a ticket.

If you do need to open a ticket please make sure to provide us with all the needed information as described in this page

This is the technical support forum for WPML - the multilingual WordPress plugin.

Everyone can read, but only WPML clients can post here. WPML team is replying on the forum 6 days per week, 22 hours per day.

This topic contains 3 replies, has 2 voices.

Last updated by Bruno Kos 2 weeks, 4 days ago.

Assigned support staff: Bruno Kos.

Author Posts
November 13, 2019 at 12:02 pm #4938659

daneelM

Hello,

I am just wondering, we can't seem to get the translation memory to work properly.

Some of our posts have repeating paragraphs and even when manually translated 2 / 3 times, the translations do no seem to automatically appear from the "translation memory."

We currently do not have any "automatic translation" credits, so I was wondering if the translation memory requires the credits on order to automatically translate from memory?

If not, what could be the cause of the memory not working as expected.

Any help or information would be much appreciated.

Thanks.

November 14, 2019 at 7:46 am #4944423

Bruno Kos
Supporter

Languages: English (English )

Timezone: Europe/Zagreb (GMT+01:00)

Hi,

I received feedback from our advanced translation editor team and this is what is behind the scenes:

The TM process is as follows

- job is send to ATE
- job is translated, all segments are saved in TM
- when a new job is sent to ATE, after segmentation we check each segment if it is already in TM
- we apply translation from TM for any matching segment
- in your case however, at the time of segmentation there was nothing in TM, entry in TM was created later

For the above job, this is the case (2310554)-> Entry in TM was created after job was opened & segmented.

Our ATE team is currently developing a mechanism to apply TM to jobs already in progress when new TM is applied, but that is still in progress.

Regards,
Bruno Kos

November 18, 2019 at 2:32 pm #4965869

daneelM

Hello,

Sorry for the late reply.

So is the solution for this to send only one copy of jobs that require memory initially.

Then once these are completed, send in the rest of the jobs that will make use of previously translated texts.

Or how do I go about making sure that that the entries in TM are created on time / accordingly?

Many Thanks.

November 19, 2019 at 8:39 am #4969467

Bruno Kos
Supporter

Languages: English (English )

Timezone: Europe/Zagreb (GMT+01:00)

Hi,

So is the solution for this to send only one copy of jobs that require memory initially.

That sounds exactly how ATE team explained it as well. So the more strings and paragraphs you add as you will translate the content, the more of that will end up in translation memory, but with that exact order of precedence. So once the job has been completed, all these strings will end up in translation memory and you can reuse them later for new translations. Important thing is that they need to be exactly the same, otherwise TM will not be able to find them.

Regards,
Bruno Kos