August 9, 2013

Pre-release version

August 1, 2013

CafeTran gets you connected

July 20, 2013

CafeTran's review appeared in the 224th _Tool Box Newsletter_ by Jost Zetzsche

Walking the Shire with a Coffee in Hand

This heading refers to Igor Kmitowski, the guy (and in this case, it really is the guy) behind the translation environment tool CafeTran. He loves his hobbits (he internally names his different software versions after Tolkien characters) and he loves his coffee (the product is actually called CafeTran Espresso). Oh, and he also calls himself "semi-autistic" ("over the last two or three years I have become a coder more fluent in the Java language than in other forms of communication").

I can attest to finding some truth in that last statement: neither CafeTran's documentation nor Igor's communication skills in general are particularly good — it took a bit of patience to coax enough adequate information out of him about his product.

But that does not make the tool itself any less impressive.

I wrote about this tool a little more than five years ago (in edition 108 for all Premium subscribers who have access to the archives). Here are some of the highlights of that review:

• CafeTran is small, it's cheap, it's easy to learn and maintain, and it's definitely not built for corporate use!
• It's a tool built for the freelance translator that aims to give as much of a true translation environment as possible.
• It's written in Java, so it's platform-independent.
• It supports text, XML, HTML, MS Office XML, and OpenOffice formats.
• It provides an easy integration of queries to Internet-based resources (sort of like IntelliWebSearch).
• It has an auto-complete feature of already-typed text (I particularly like this).
• It offers an extraction feature for frequent terms for glossary-building purposes.
• All this for a price tag of only 40 euros!

A few of these details have changed (it now costs 80 euros, for instance — still shockingly inexpensive), though the gist is still the same. The tool itself, though, is much more than it used to be — so much so, in fact, that I would no longer call it an "easy tool." For the laughable purchase price you get a full-fledged tool that supports many more formats than it used to (it now also supports all MS Office 2007 and higher files, InDesign, FrameMaker, a lot of software development file formats, AutoCAD DXF files (!), and a great number of bilingual formats coming from other TEnTs, including Transit, Trados, Wordfast, memoQ, and Déjà Vu) and offers additional features that are very, very clever.

Let's take a step back, though. Igor told me that "it's a challenge to rival the teams of excellent programmers from SDL, Kilgray or Atril and I made a rule to avoid examining other CATs not to get overwhelmed." I totally get that and it really shows. What we see so often between the other vendors is a mutual catch-up game — which in many cases works great for us, the consumers. Someone like Igor with a "few hundred customers" and the hope that the tool will "one day become my only source of income" does not really need to worry about the competition but can completely focus on his customer base and their requests. This leads to solutions that are unique to this tool, and it also leads to more and more features.

"More features" is great, of course, but it also can make a tool more and more complex. CafeTran appears to be caught in that Catch-22, but this doesn't mean that it's not workable anymore — it is just no longer a tool that you can simply sit down with and start using (very much like the "big boys" — Trados, memoQ, etc.).

When you open the tool you're presented with the Project Manager and a plethora of options that are initially difficult to understand. For instance, you can choose "TMX Memories" as well as "External Databases" for both segments and terms. The difference? The flat text TMX memories require more resources and can at some point be sent to a true database format with fewer resource requirements. And terminology is indeed processed in the same formats as true TM data — even though it's now also possible to do this as tab-delimited text files. (Terminologists will cringe at the no-concept-based terminology databases.) To me, these look like a case where the tool started out with one feature and then added a parallel feature because of user requests ("need more memory" or "our glossaries are in text format") without questioning whether the original feature was still needed. I would recommend some pruning.

Still, here are some of the features that really make the tool into a productivity hub:

• Automatic assemble of segments from the various parts found in records from glossaries, translation memories, and automatically extracted subsegments from translation memories
• Automatic suggestion based on the first few keystrokes and content in glossaries, TMs, and machine translation engines (the pre-configured possibilities here are Google Translate, Bing Translator, and MyMemory)
• Automatic extraction of proposed terms for every segment you work on as candidates for the glossary
• Very easy and intuitive way to enter terms into the glossary (highlight source, highlight target, click button or press key)
• Context matches that differentiate between having the same context on one side of the segment or on both (cleverly named 101% vs. 102% matches)
• Possibility to use images as reference material
• Clipboard-based workflow for unsupported applications and environments
• Smart handling (i.e., deletion) of unnecessary formatting tags in scanned and otherwise badly formatted documents
• The translation is done internally via XLIFF, and a fully translated XLIFF file is automatically generated at the end of the project (aside from the actually translated file)
• The interface is completely configurable as far as what kind of panes are displayed in what manner
• QA checks include all the standard checks also found in other tools
• For files that originated in Trados, it supports bilingual Word docs, TTX files, and SDLXLIFF and SDLPPX files (and, as mentioned above, various other intermediate files from other TEnTs)

This only scratches the surface. If you've received the impression that this is "just a tool for beginners" because it only costs 80 euros, think again (and accept my apologies for poorly communicating what I tried to say). If you're looking for something new and think this might be a real option, especially because it also runs on Mac (approximately 20% of CafeTran users use it on Mac) or Linux (less than 10%), you really might want to try it out. Plus, Igor has promised as his next great project to spend time on the documentation (right now he's working on enabling the tool to export comments on translation segments).

Kołobrzeg, June 29, 2013

And so with the sunshine and the great bursts of leaves growing on the trees, just as things grow in fast movies, I had that familiar conviction that life was beginning over again with the summer.
― F. Scott Fitzgerald, The Great Gatsby

Dear CafeTran users,

The new version of CafeTran Espresso – with code name Aragorn – is officially released.

Happy Summer Holidays!

Igor

Exciting New Features of CafeTran Aragorn

• Boost your productivity by inserting alternative target terms via the context menu

• Enjoy dynamic auto-assembling: CafeTran remembers which target term you selected from the context menu and will use it in the rest of your translation session.

• Profit from QA Word lists, allowing you to check for word presence or absence

Improvements in CafeTran Aragorn

• The Grid now displays character formatting (bold, italics, underlined). The Document tab is no longer created in the tabbed pane.
• Improved Déjà Vu XLIFF filter, keeping surrounding tags out of segment content
• Improved DOCX filter to get rid of some unnecessary tags
• Improved Transit NXT filter, handling non XML conform tags in language pairs
• Improved SDLXLIFF filter, better handling of segments that only contain tags
• Improved TTX filter, QA tags now checks tag positions strictly (like Tag Editor itself), propagation now works
• Improved Statistics, enabling calculations in spreadsheet applications
• Improved QA tasks
• The Task menu has been taken out of the Translation menu, allowing quicker access to important Translation tasks
• The QA menu has been taken out of the Edit menu, allowing quicker access to important Quality Assurance tasks
• Simultaneous Web Searches are possible now.
• Find and Replace dialog can stay open, while editing the target text box
• Second Surround with characters function, allowing two different sets of surround characters
• Stemming is enabled in tab-delimited glossaries too (enter piękn|y (beautiful) as a source term and pięknego, pięknych etc. will also be recognised)
• The New glossary entry dialog can now be closed with Ctrl+Enter
• Two new toggles to skip display of the New glossary entry dialog: Translation | Options | Skip new term window (keyboard shortcut and icon)
• Pressing the Ctrl key while clicking the Add terms icon or pressing the Add terms key will call the New glossary entry dialog when both toggles (above) are active
• Edit | Target Segment | Adjust end punctuation (default keyboard shortcut: Ctrl+]) now removes wrong trailing normal spaces (and non-breaking ones) too
• Change case (default keyboard shortcut: ) now works in dialogs (New glossary entry, Find and Replace) too
• When inserting tags with Cmd+Click (Ctrl+Click on Windows), the correct tag number is calculated
• The Library | Internet menu has been renamed to Library | Web, to refer to CafeTran's web searches
• The Add segment to memory icon has been moved to the second position on the icon bar and renamed to Add segment to memory and go to next segment (the latter is true for the corresponding menu item in the Translation | Action menu too)

Small Changes Can Make Big Differences

• You can now vertically flip (reverse) the position of the source segment toolbar and of the target segment toolbar. This is especially handy when adding source and target segment selections to the glossary with the mouse: position the Add term to database or glossary icon at the top of target segment – between the source and target box – to reduce your mouse mileage:

Fixes in CafeTran Aragorn

• When working with 3 glossaries, the New glossary entry dialog will save to the correct (write-enabled) glossary
• Unfolding of tags at case change is fixed.
• While inserting the first match (Control+Alt+1) the auto-suggestion dialog was displayed: fixed
• Transit hyphen issue in MIF projects created with Transit XV is fixed with a tag, for now.

Kołobrzeg, March 28, 2013

Hi all,

The new version of CafeTran Espresso - the code name Arwen is officially released.

The most striking improvement is in the instant access to the Find function without a need to switch constantly a scope of search. You will notice new buttons upon selection of a phrase or word to search wherever you wish. You can watch a short video.

The Auto-complete (Auto-suggest) function is first letter case insensitive now to avoid entering twice the same phrase (term) to the TM or glossary for this purpose.

Merging memories into one (see the Memory | Merge memories… menu).

QA Terms/segments inconsistency check list inconsistent translation of whole segments (For segments TMs) and terms (For terms TMs).

One click QA spell checking of all segments for both internal Hunspell spell checker and external Libre(OpenOffice) spell checker.

You can examine the Unicode number of a character at the cursor. The Unicode number is displayed above the source text box:

Defining non-standard space Unicode characters to help correct recognition of words (see Options | Memory tab). The Unicode numbers in the field should be comma delimited (e.g U+00A0,U+2009).

Improvements to the SDLXLIFF filter and MS PowerPoint filter.

Support for memoQ XLIFF.

Support for TXML projects.

Improvements in the Documents alignment interface to make it more intuitive.

Okay, I think I did not miss anything. Enjoy it!

Happy Easter,
Igor

What's brewing?

Igor is working hard on getting CafeTran ready for the Spring Update. New features include:

• Improved SDLXLIFF filter
• TXML support
• Unicode number indicator
• Handling of user-defined space characters
• Power Concordance

Click here for a Sneak Preview.

CafeTran Espresso 2013 - AKA Frodo

Kołobrzeg, January 31, 2013

Dear CafeTran fans,

I am pleased to announce the next update CafeTran Espresso 2013 called Frodo. It features some useful translation tasks.

1. Inserting all Exact and Context Matches from Translation Memories to the Project in one step.
2. Transferring source segments with no letters to target segments in one step.
3. Filtering and transferring the filtered source segments to target segments in one step.
4. All the tasks (see the menu Translation | Tasks) can be performed on the segments filtered through QA, View or Search.
5. Filtering out Context and Exact Matches, Segments with letters and reversing the filters (no Context Matches, Segments with no letters).
6. Removing target segments based on the filtering criteria.
7. Removing TMX units (TUs) based on the filtering criteria in the TMX edit mode.
8. Removing TMX units with the same source segments in TMX edit mode.
9. Presenting the Exact and Context Match status in the grid.
10. Jumping over "No letters" segments.
11. Applying Checked status to the filtered segments.
12. Selecting Memories for Pretranslation.
13. Other small fixes and improvements (adding new words to spelling dictionary doesn't require confirmation, tags aren't expanded when modifying adjacent words, QA checks are now more detailed/can be selected more individually etc.).

The above functions make it possible to filter out any group of segments based on any criteria, present these segments for translation or exclude from translation.

Full version users with valid technical support are welcomed to request for the update at the support email address (moc.liamg|troppus.nartefac#moc.liamg|troppus.nartefac)

Cheerio,
Igor

CafeTran Espresso 2013 - AKA Bilbo

Kolobrzeg, January 13, 2013

Dear CafeTranslators,

Today, I updated CafeTran to include the Context Matching feature discussed recently here. The program analyses the preceding and the following segments against the memory segments to determine the accuracy of the Context Match, and reports the result in the memory tab. Moreover, Context Matches have priority over Exact Matches and are chosen for target segments. This is especially handy for shorter identical source segments having various translations. Read more about Context Matches.

Since Context Matching seems to be a useful function, I have decided to change the program's code name for this update. The current version is CafeTran Espresso 2013 - Bilbo (still in the Hobbit mood :)

Full version users with valid technical support are welcomed to request for the update at the support e-mail address (moc.liamg|troppus.nartefac#moc.liamg|troppus.nartefac)

Cheerio,
Igor

CafeTran Espresso 2013 - AKA Hobbit

Kolobrzeg, December 27, 2012

CafeTran is updated frequently. New builds and new versions are released very often. The build number can be found in the About dialog, it consists of the build date and a trailing number: YYYYMMDD##:

As of Christmas 2012 CafeTran uses version names to differentiate between new versions. The first version will be called 'Hobbit'.

CafeTran Espresso 2013 Hobbit offers the following new features and improvements:

1. Custom formatting

CafeTran can add additional styles to the current segment (e.g. bold, underline, italics). It can also cancel the existing source segment styles. The formatting tags insertion is as easy as it is with the regular tags transfer.

For example, just type b and press Esc to start the word in the bold style while typing /b and pressing Esc ends the style. The function is also available via the context toolbar and works on the selected text. It works with Ms Word, Ms PowerPoint and the OpenOffice/LibreOffice documents.

More info here: Custom formatting

2. Enhanced clipboard workflow

The powerful translation via clipboard workflow has been enhanced to enable applying styles (e.g. italics) to target segments. Before, the feature was limited to pure text translation which meant that all formatting was lost. Now you can segment and translate formatted source documents and transfer them back to the original application.

The image translation workflow can benefit from the new style tagging feature as well. The translated text is exported along with styles to open it in your favorite editor.

3. Mouse tag insertion

Hold a CTRL (Windows) or Apple (Mac OS) key and mouse click in the target segments to insert a tag.

4. Inserting non-breaking spaces and non-breaking hyphens

Via shortcuts (see menu Edit | Target segment)

5. Surrounding a phrase with defined characters

Set to insert French quotes, by default. via a shortcut (available also through the context toolbar). The mouse selection of text is not necessary if you surround only one word. Read more.

6. Direct translation of SDLPPX packages

Including re-injecting the translated SDLXLIFF file into the package. Read all about it.

7. Importing SDL memories from SDL packages and Apple glossaries

Both included sources are automatically converted to the TMX format.

8. Saving CafeTran projects as compressed packages

The new, compressed CafeTran project format, to send project packages easily across the Web.

9. New eye-friendly color themes

New themes can be selected in the menu Edit | Appearance | Themes | Image background - transparent dark and transparent bright.

For example, after selecting the Image background - transparent dark theme, you will notice a slider and a camera icon. The slider lets you adjust/adapt the brightness of the background to the ambient light while clicking the camera icon enables to change the background image to your personal preference. It is a great UI feature that saves your eyes.

Many smaller improvements, including:

• Green arrow in the F/R dialog to clone a selection/the complete Find text to the Replace field.
• CafeTran ignores words containing numbers during spell-check
• Jump settings are temporarily deactivated when performing QA. Handy for partially pre-translated Transit and SDL files.
• Transit segment status can be set to Translated via Translation | Task. See: Translating Transit files
• SDLPPX packages can be re-packed with the translated SDLXLIFF files, with segment status set to Translated, for opening in Studio and perform a final verifying (to be on the save side before sending your work to the client).
• Up to 20 Recent Projects are kept in the Program Manager now. CafeTran saves the assigned TMs (for native CafeTran projects).

December 2012 update

Kolobrzeg, December 26, 2012

The upgrade is almost ready and will offer more functions than you'll expect.

Updating the Wiki to cover all those new goodies will be quite a job, since it will require some new screenshots etc.

Stay tuned for an update / detailed descriptions of:

• Translating SDLXLIFF files / SDLPPX packages
• Handling Transit NXT files

Kolobrzeg, December 11, 2012

Custom Character Format Tagging

The imminent update of CafeTran will bring the long-awaited custom character format tagging, as requested by many users on Proz.com and in the CafeTran Google Group.

Click here for an explanation of how to add custom tags ...

Dear Cafetranslators,

The time has come for the October update with following new features, improvements and changes:

1. Refreshed Find and Replace panel.
2. Multiple segments filter allowing to perform multi-step filtering both of Project and TM segments.
3. QA filtering of spelling mistakes.
4. New Auto-assembling panel replacing the previous pop-up window.
5. Enhancement in the Auto-assembling feature.
6. Single-click mouse tag placement (activation via target window context menu).
7. Deciding on the Project location when creating a CT project.
8. New filter for Transit XML files and improvements in the SDLXLIFF filter.
9. No more segment breaks on format change (bold, italics, underline) in the Excel filter.
10. Import of TBX terminology.
11. Some changes in CT Terminology (Auto-assembling replaces Autotranslation, Fragments term is now used for exact subsegment matches)

Current CafeTran users do not need to download and install the new version. Please contact me to for the update instruction.

Cheers,
Igor