LinuxQuestions.org
Help answer threads with 0 replies.
Go Back   LinuxQuestions.org > Forums > Non-*NIX Forums > General
User Name
Password
General This forum is for non-technical general discussion which can include both Linux and non-Linux topics. Have fun!

Notices


Reply
  Search this Thread
Old 02-24-2019, 04:33 AM   #1
rblampain
Senior Member
 
Registered: Aug 2004
Location: Western Australia
Distribution: Debian 9.2
Posts: 1,127

Rep: Reputation: 50
google translate document


For anyone interested.
I have found that the claim Google can freely translate documents of a size of 1MB is wrong (misleading advertising?). I have an English document (text only) of a length of 133KB and, tested in the first 4 languages in Google list of available languages, Google systematically translates the first 300 lines (lines terminated with nl - ASCII 10) which is under 36KB and then only translates some other lines if they contain only 1 word - which indicates Google keeps looking at each line following line number 300. Otherwise, Google consistently returns the original English text from line 301.

Last edited by rblampain; 02-24-2019 at 05:46 AM.
 
Old 02-24-2019, 02:26 PM   #2
RandomTroll
Senior Member
 
Registered: Mar 2010
Distribution: Slackware
Posts: 1,169

Rep: Reputation: 188Reputation: 188
Quote:
Originally Posted by rblampain View Post
I have found that the claim Google can freely translate documents of a size of 1MB is wrong (misleading advertising?).
Where did you see this claim advertised? I've had to break docs up to get Google to translate all of them.
 
Old 02-24-2019, 03:57 PM   #3
l0f4r0
Member
 
Registered: Jul 2018
Location: Paris
Distribution: macOS, Slackware
Posts: 803

Rep: Reputation: 281Reputation: 281Reputation: 281
Quote:
Originally Posted by RandomTroll View Post
Where did you see this claim advertised?
https://support.google.com/translato.../6306382?hl=en
 
Old 02-25-2019, 11:49 AM   #4
RandomTroll
Senior Member
 
Registered: Mar 2010
Distribution: Slackware
Posts: 1,169

Rep: Reputation: 188Reputation: 188
Quote:
Originally Posted by l0f4r0 View Post
This is 'Translator Toolkit'. Is this the same as Google Translate? Which do you use?
 
Old 02-25-2019, 03:15 PM   #5
l0f4r0
Member
 
Registered: Jul 2018
Location: Paris
Distribution: macOS, Slackware
Posts: 803

Rep: Reputation: 281Reputation: 281Reputation: 281
@RandomTroll: Hummm...Google Translator Toolkit is not Google Translate indeed. A priori, Translator Toolkit is a way to edit and collaborate around translations handled by Google Translate.
Quote:
Originally Posted by https://support.google.com/translatortoolkit/answer/6306366?hl=en
How is Translator Toolkit different from Google Translate?
Google Translate is an instant, online translator. Translator Toolkit is an online editor that lets you modify and collaborate on translations generated by Google Translate.
@rblampain: Regarding Google Translate, upon reflection, I'm not sure there is a file size limit of 1MB after all.
However there is a limit for sure of 5000 characters when one uses the text translation service (there is even a countdown at the bottom right corner of the text area). It doesn't seem to apply for URL based translations though. I wonder if this limit still applies when one updates documents (if somebody wants to test that use case, feel free )
 
Old 02-25-2019, 09:25 PM   #6
RandomTroll
Senior Member
 
Registered: Mar 2010
Distribution: Slackware
Posts: 1,169

Rep: Reputation: 188Reputation: 188
Quote:
Originally Posted by l0f4r0 View Post
It doesn't seem to apply for URL based translations though.
It did when I used it - though that was last in 2002. I translated a book of short plays from French then submitted the originals to Google Translate. I had to break them up. Then the author got a deal for an English translation and their lawyers sent me a letter. I thought mine were so crummy they couldn't object to them - I forgot about lawyers, who don't know crummy when they read it.
 
Old 02-26-2019, 01:09 AM   #7
ondoho
LQ Addict
 
Registered: Dec 2013
Posts: 11,046
Blog Entries: 7

Rep: Reputation: 2780Reputation: 2780Reputation: 2780Reputation: 2780Reputation: 2780Reputation: 2780Reputation: 2780Reputation: 2780Reputation: 2780Reputation: 2780Reputation: 2780
Quote:
Originally Posted by RandomTroll View Post
It did when I used it - though that was last in 2002. I translated a book of short plays from French then submitted the originals to Google Translate. I had to break them up. Then the author got a deal for an English translation and their lawyers sent me a letter. I thought mine were so crummy they couldn't object to them - I forgot about lawyers, who don't know crummy when they read it.
2002?
google translate did not even launch until 2006.
something's fishy...
 
Old 02-26-2019, 03:44 PM   #8
RandomTroll
Senior Member
 
Registered: Mar 2010
Distribution: Slackware
Posts: 1,169

Rep: Reputation: 188Reputation: 188
Quote:
Originally Posted by ondoho View Post
2002?
google translate did not even launch until 2006.
something's fishy...
Hmmmm.... maybe it was Babelfish; I had an invitation to G-Mail from an industry insider - could I have had special access?
 
Old 02-27-2019, 12:56 AM   #9
l0f4r0
Member
 
Registered: Jul 2018
Location: Paris
Distribution: macOS, Slackware
Posts: 803

Rep: Reputation: 281Reputation: 281Reputation: 281
Quote:
Originally Posted by RandomTroll View Post
It did when I used it - though that was last in 2002. I translated a book of short plays from French then submitted the originals to Google Translate. I had to break them up.
Of course if you meant Babelfish the discussion is over but, additionnally, are you really speaking about URL based translations? Your description brings me to mind about document upload based translations...or maybe you are implying that you first uploaded your content on a webserver and then you provided the corresponding URLs to Google/Babelfish?

Last edited by l0f4r0; 02-27-2019 at 12:57 AM.
 
Old 02-27-2019, 04:10 PM   #10
RandomTroll
Senior Member
 
Registered: Mar 2010
Distribution: Slackware
Posts: 1,169

Rep: Reputation: 188Reputation: 188
Quote:
Originally Posted by l0f4r0 View Post
are you really speaking about URL based translations? Your description brings me to mind about document upload based translations...or maybe you are implying that you first uploaded your content on a webserver and then you provided the corresponding URLs to Google/Babelfish?
The latter. I had a personal website to which I uploaded the originals, then pointed Google/Babelfish to them.
 
Old 02-27-2019, 05:10 PM   #11
l0f4r0
Member
 
Registered: Jul 2018
Location: Paris
Distribution: macOS, Slackware
Posts: 803

Rep: Reputation: 281Reputation: 281Reputation: 281
^ Okay

I've read somewhere (cannot remember where exactly) that URL based translations were not limited in size.

More, the following sentence from Google FAQ:
Quote:
Originally Posted by https://support.google.com/translate/answer/6142478?hl=en&ref_topic=7011659
You can translate up to 5,000 characters. To translate longer pieces of text, cut and paste the text in blocks, or upload an entire document.
seems to imply that there is no limit as well regarding uploaded documents...

Anyway, all of this can be easily verified if someone has a little free time
 
Old 02-27-2019, 11:34 PM   #12
RandomTroll
Senior Member
 
Registered: Mar 2010
Distribution: Slackware
Posts: 1,169

Rep: Reputation: 188Reputation: 188
I checked my archives. It was indeed Babelfish. I discovered in 1998 it had a limit of 4K characters.
 
Old 03-03-2019, 03:30 AM   #13
rblampain
Senior Member
 
Registered: Aug 2004
Location: Western Australia
Distribution: Debian 9.2
Posts: 1,127

Original Poster
Rep: Reputation: 50
In my case, it was a ".txt" file of 133KB ASCII characters only - browser default: UTF-8 but no UTF-8 (2 byte or more) character in file.
In browser:
Code:
--> Google Translate --> Document --> Choose a document
Upload a .doc, .docx, .odf, .pdf, .ppt, .pptx, .ps, .rtf, .txt, .xls, or .xlsx
Browse your computer
Navigate to file (after selecting English to Afrikaans/Albanian/Amharic/Arabic) -->
Code:
Translate
And the first 35KB of the English text were translated in each case (precisely 300 lines).

My main objective is to encourage those companies to use clarity and honesty in their advertising rather than finding work-around which is obviously to split a big file. Although I could be wrong (found only once), it seems that Google kept trace of my translation attempts and reduced considerably the amount translated on the second try, that is when I tried to translate into the same language the text following the first 35K/300 lines (they claim you can upload up to 1GB per year which implies some sort of record-keeping on their part). Their objective is obviously to make you pay for it but I am personally reluctant to pay for something and discover that the paid version has just the same or similar annoying "features" than the free version. I found exactly that in a paid email service but in Google case, using the translation paid version involves a fair bit of work.

Last edited by rblampain; 03-03-2019 at 04:43 AM.
 
Old 03-03-2019, 02:30 PM   #14
l0f4r0
Member
 
Registered: Jul 2018
Location: Paris
Distribution: macOS, Slackware
Posts: 803

Rep: Reputation: 281Reputation: 281Reputation: 281
Quote:
Originally Posted by rblampain View Post
In my case, it was a ".txt" file of 133KB ASCII characters only - browser default: UTF-8 but no UTF-8 (2 byte or more) character in file.
In browser:
Code:
--> Google Translate --> Document --> Choose a document
Upload a .doc, .docx, .odf, .pdf, .ppt, .pptx, .ps, .rtf, .txt, .xls, or .xlsx
Browse your computer
Navigate to file (after selecting English to Afrikaans/Albanian/Amharic/Arabic) -->
Code:
Translate
And the first 35KB of the English text were translated in each case (precisely 300 lines).
Strange, I can reproduce your behavior indeed.
See attached:
  • file_to_be_translated3.txt: original ASCII english content (135 972 bytes)
  • file_translated3.txt: translation into UTF-8 Afrikaans by Document based Google Translate (137 454 bytes)
Translation is done until line#63 (character #23819). After that, translated content is equal to the original english content...
Attached Files
File Type: txt file_to_be_translated3.txt (132.8 KB, 2 views)
File Type: txt file_translated3.txt (134.2 KB, 2 views)
 
  


Reply


Thread Tools Search this Thread
Search this Thread:

Advanced Search

Posting Rules
You may not post new threads
You may not post replies
You may not post attachments
You may not edit your posts

BB code is On
Smilies are On
[IMG] code is Off
HTML code is Off



Similar Threads
Thread Thread Starter Forum Replies Last Post
LXer: Translate Shell A Tool To Use Google Translate From Command Line In Linux LXer Syndicated Linux News 0 11-30-2017 07:05 PM
Translate a PDF document to English? Xeratul General 3 07-25-2016 10:08 PM
[SOLVED] Weird Firefox behavior when using google.translate.com corp769 Linux - Software 8 07-02-2011 08:43 PM
BASH/No X: Using google translate to convert TXT files (translate) frenchn00b Programming 10 09-13-2009 10:55 PM
to translate or not to translate HTML rblampain General 2 07-05-2007 09:04 AM

LinuxQuestions.org > Forums > Non-*NIX Forums > General

All times are GMT -5. The time now is 04:06 AM.

Main Menu
Advertisement
My LQ
Write for LQ
LinuxQuestions.org is looking for people interested in writing Editorials, Articles, Reviews, and more. If you'd like to contribute content, let us know.
Main Menu
Syndicate
RSS1  Latest Threads
RSS1  LQ News
Twitter: @linuxquestions
Facebook: linuxquestions Google+: linuxquestions
Open Source Consulting | Domain Registration