OmegaT v3 released (my write-up) Thread poster: Samuel Murray
|
|---|
Samuel Murray 
Netherlands
Local time: 14:32
Member (2006)
English to Afrikaans
+ ...
Here's my unofficial write-up of OmegaT v3
Translators who are happy with OmegaT will be glad to know that OmegaT v3 looks and works pretty much the same as OmegaT v2. At the same time, OmegaT is still completely devoid of toolbars and icons, and it is more than ever in desperate need of redesigned menus and dialogs. Translators who are new to OmegaT would be forgiven for having second thoughts about it when they see it.
... See more Here's my unofficial write-up of OmegaT v3
Translators who are happy with OmegaT will be glad to know that OmegaT v3 looks and works pretty much the same as OmegaT v2. At the same time, OmegaT is still completely devoid of toolbars and icons, and it is more than ever in desperate need of redesigned menus and dialogs. Translators who are new to OmegaT would be forgiven for having second thoughts about it when they see it.
A lot of hard work went into OmegaT v3, and it is important to mention this, because OmegaT is developed by a team of volunteers that keep up the good work. The User Manual is still slightly outdated, although most of the new features are mentioned in it, or in the Help > Latest Changes infobox.
A. It has always required a bit of tinkering to use some of the more advanced features in OmegaT, and that will continue to be so, but three of those are now much easier to use:
1. The tokenizers are now built-in. The word "tokenizer" is geek speak for smart handling of prefixes and suffixes when doing glossary look-ups and translation memory matching. Since the tokenizers themselves are maintained by teams of volunteers (and not by OmegaT's volunteers), this change will benefit only some languages.
2. The grammar checker LanguageTool is now built-in. Since the LanguageTool rules themselves are maintained by teams of volunteers (and not by OmegaT's volunteers), this change will benefit only some languages. If you have at least one-and-a-half a brain, you can help develop grammar checking for your own language, by joining the LanguageTool community: http://www.languagetool.org/languages/
3. It is now possible to change the keyboard shortcuts in OmegaT without the need for programmer skills. It still requires a bit of digging (and some RTFM responses on forums), but essentially you can assign any keyboard shortcut to any feature that you can access using a menu.
B. Translators who know a bit of scripting (the programming kind, not the screenplay kind) now have three separate but built-in opportunities to extend the functionality of OmegaT:
1. If you are a slightly advanced user of the command line in your computer (e.g. the "DOS" or Command prompt in Windows), you can make OmegaT perform a number of actions whenever you tell OmegaT to create the translated documents. Using this feature, you can run other programs and make those programs interact with your translations intelligently, by using scary things called variables.
Any user can access this feature from within OmegaT, but the user manual doesn't tell you how it works, and if you don't know by yourself how it works, then you're probably not one of those advanced users for whom this is intended.
2. If you can speak NetRexx, XSLT, Groovy or JavaScript, and if you can read and understand Java, then you can add features inside OmegaT using OmegaT's "scripting" plugin. At this time, OmegaT plugins are not easy to install if you're just a garden-variety translator. The scripting plugin allows scripters to write scripts that other translators can also use, but... installing such scripts also requires a bit of skill.
The upshot is that "anyone" can write extra features for OmegaT and share these with fellow translators, without any interference to or from the team that develop the main OmegaT program. An example of such a scripted feature is Find/Replace, which is not yet (gasp!) available in OmegaT itself.
3. Translators who write scripts in languages like Perl, Tcl/Tk, AutoIt and AutoHotKey can use the Export Segment feature that has been available in OmegaT v2 as well. Essentially, this saves the current segment's source text and target text temporarily to two tiny text files where clever scripts written by simple translators (or vice versa) can access them and do cool things with them.
C. Any XLIFF translator who tried OmegaT previously and found it wanting should try OmegaT again. XLIFF support is much, much improved. I'm not an XLIFF translator myself, so I don't always understand these things, but other people say that the new features are more impressive than the old ones, and you can believe anything you read on the internet.
For example, you can tweak the file filters to maintain or remove spaces in relation to tags, and you can keep or remove "leading" and "trailing" spaces (that's nerd speak for spaces before the first letter of the segment and spaces after the last letter of the segment).
Also, although OmegaT will try to reduce the number of tags by not including leading and trailing tags in the segment, you can disable that feature if your target language typically has a word order that makes such a feature annoying (several other CAT tools also have this annoyance, but OmegaT gives translators the option to disable it, if possible).
D. Like most modern CAT tools, OmegaT uses tags to indicate the formatting and other hidden elements of the source text. Working with tags break your speed, in all CAT tools, but in OmegaT the effect of tags on productivity is particularly significant. OmegaT v3 has several new features that make dealing with tags less of pain, although there is still a long road ahead before it's a "pleasure".
You can no longer accidentally break tags by careless overtyping or use of the Delete of Backspace keys (unless you disable that protection, which you can). In OmegaT, tags consist of individual characters, and previously these characters behaved like individual characters when you moved your cursor over them, but in OmegaT v3 tags behave a bit more like words, which makes it easier to copy or move them.
There is now the option to check the correctness of tags whenever you move to another segment -- previously you could only check this at the end of the job. And if you're a lazy translator, you can click a button that will automatically fix the tag errors by inserting any missing tags (this might not always be a wise thing to do, but the choice is now yours whether you want speed or safety).
E. The Find feature is improved in OmegaT v3, although you still can't do Find/Replace in OmegaT. Previously, the Find dialog could not search inside glossaries, and previously, you could not tell OmegaT to not look in the source text when doing a search, but in OmegaT v3 you can select those options.
F. In OmegaT v2, any changes you made to the file filters and segmentation rules were permanent and affected all future translations also. (File filters tell OmegaT how to open files based on the type of file, and segmentation rules tell OmegaT how to break paragraphs into sentences and how to watch out for abbreviations.) In OmegaT v3, you can make file filters and segmentation rules project-specific, which means that you can make drastic changes to these settings for the current translation, without affecting future translations.
G. And finally, there were some improvements to the "Team" feature in OmegaT. This allows a brainy project manager to set up group session during which multiple translators cooperate on a single translation project all at the same time, while benefiting from each other's translations and glossaries.
These features are all available in version 3 of OmegaT. If you want to download this version of OmegaT, look for the label "latest", not "standard", on the download page: http://sourceforge.net/projects/omegat/files/
[Edited at 2013-04-26 08:28 GMT] ▲ Collapse
| | | | esperantisto
Local time: 15:32
Member (2006)
English to Russian
+ ...
SITE LOCALIZER | Thank you, Samuel! | Apr 26, 2013 |
I find this quite well-reasoned. However, instead of:
At the same time, OmegaT is still completely devoid of toolbars and icons, and it is more than ever in desperate need of redesigned menus and dialogs.
I’d write:
At the same time, OmegaT is still completely devoid of toolbars and icons, which lets you focus on translating rather than on finding the right point for your mouse.
Anyway, thanks!
| | | | Samuel Murray
Netherlands
Local time: 14:32
Member (2006)
English to Afrikaans
+ ...
TOPIC STARTER | Menus, dialogs, and mouse work | Apr 26, 2013 |
esperantisto wrote:
However, instead of:
At the same time, OmegaT is still completely devoid of toolbars and icons, and it is more than ever in desperate need of redesigned menus and dialogs.
I’d write:
At the same time, OmegaT is still completely devoid of toolbars and icons, which lets you focus on translating rather than on finding the right point for your mouse.
Well, you would have had a point about the mouse, except:
Several features in OmegaT can only be used and accessed if you use your mouse. And for some of those features, you actually need a fairly accurate and steady mouse hand. Particularly glossary work and spell-checking in the Edit pane suffers from this. You also need mouse skills of the Excel level to edit file filters and segmentation rules.
What's more, many of the actions in the Edit pane that one would usually find on a toolbar are actually present in OmegaT as clickable menus. So it's not as if the OmegaT developer does not want users to use the mouse to use those features. It's just that clicking a single button is quicker (particularly for less used features where you don't know the keyboard shortcut) than clicking a menu and then selecting an option in it.
As for the menus and dialogs mentioned in my post, the fact is that new features in OmegaT are often simply dropped into any available space in the GUI, seemingly without plan or reason. If you know of a new feature in OmegaT, you'll find it quicker by taking all the menus from one side and checking all of them individually until you find it, than if you tried finding the appropriate dialog by reading the menu names.
It is blunderous to say that OmegaT has no toolbar because OmegaT is supposedly not a mouse-driven tool.
| | | | Susan Welsh
United States
Local time: 08:32
Russian to English
+ ...
| toolbars, icons, bells, whistles - phooey! | Apr 26, 2013 |
Samuel Murray wrote:
OmegaT is still completely devoid of toolbars and icons, and it is more than ever in desperate need of redesigned menus and dialogs. Translators who are new to OmegaT would be forgiven for having second thoughts about it when they see it.
I HATE toolbars and icons. I LOVE that OmegaT's GUI is very simple and clean. The one thing that might someday make me reconsider using it is if they were to put a lot of cr--p into the GUI.
My two cents.
| | |
|
|
|
Samuel Murray
Netherlands
Local time: 14:32
Member (2006)
English to Afrikaans
+ ...
TOPIC STARTER | Simple and clean GUI | Apr 26, 2013 |
Susan Welsh wrote: Samuel Murray wrote:
OmegaT is still completely devoid of toolbars and icons, and it is more than ever in desperate need of redesigned menus and dialogs. Translators who are new to OmegaT would be forgiven for having second thoughts about it when they see it. I HATE toolbars and icons. I LOVE that OmegaT's GUI is very simple and clean.
It is true that OmegaT's GUI is a lot simpler and cleaner than that of many other CAT tools, but it is not the absence of toolbars and icons that does that. Rather, other tools show more information, and OmegaT does not. And, other tools use more lines and borders and colours in the edit pane, whereas OmegaT does not. Saying "OmegaT should get some toolbars" is not the same as saying "OmegaT should get more bells and whistles" or "OmegaT should become more colourful".
The one thing that might someday make me reconsider using it is if they were to put a lot of cr--p into the GUI.
Of course, it goes without saying that if you don't like the toolbars, you simply deselect them, or change them to a skin that you find minimalistic enough.
| | | | Oliver Walter
United Kingdom
Local time: 13:32
German to English
+ ...
I suppose, however, that OmegaT still has the feature that means I cannot even consider using it: it insists on segmenting the whole document automatically. The user has some ability to adjust the segmentation rules but, once they are defined, the whole document is segmented according to them.
That makes it unusable for me because:
- My biggest supplier of work sends me documents to translate that contain red strikethrough text that I must leave untouched, and black text
... See more I suppose, however, that OmegaT still has the feature that means I cannot even consider using it: it insists on segmenting the whole document automatically. The user has some ability to adjust the segmentation rules but, once they are defined, the whole document is segmented according to them.
That makes it unusable for me because:
- My biggest supplier of work sends me documents to translate that contain red strikethrough text that I must leave untouched, and black text that I translate;
- While I am doing the translation, I must be able to define the character string that is the next segment to translate in the places where I don't want the automatic segmentation rules to apply. Examples of this are
- sometimes I want a segment to end at a word that is followed by a colon (:) or a semicolon (;) and sometimes I want it to end at a word later than the (semi-)colon, followed by a full stop (="period")
- Sometimes there are errors in the source text and a string that is in fact a sentence does not end with a full stop or other ESP (end-of-segment punctuation)
Wordfast (Classic) has all of these features (i.e. automatic segmentation that I can override interactively when I need to).
Oliver ▲ Collapse
| | | | Samuel Murray
Netherlands
Local time: 14:32
Member (2006)
English to Afrikaans
+ ...
TOPIC STARTER | On segmentation | Apr 27, 2013 |
Oliver Walter wrote:
OmegaT ... insists on segmenting the whole document automatically. The user has some ability to adjust the segmentation rules but, once they are defined, the whole document is segmented according to them.
Well, you can make changes to the segmentation rules at any time before or during a translation, and simply press F5 to reload the project, and then your new segmentation rules will be applied.
However, yes, adding a segmentation rule is a nightmare, because (a) you have to write it in regular expressions and (b) you can't test the rule, except by exiting and re-entering a series of dialogs for each and every adjustment. Also, any text that matches your new segmentation rule will be segmented according to it, even if you had wanted the text only segmented in one place.
Apart from Wordfast Classic, how many tools do you know of that allows self-segmentation on the fly?
| | | | Oliver Walter
United Kingdom
Local time: 13:32
German to English
+ ...
| On-the-fly segmentation | Apr 27, 2013 |
Samuel Murray wrote:
Apart from Wordfast Classic, how many tools do you know of that allows self-segmentation on the fly?
I don't know which CAT tools they are but, because this facility is important to me, and perhaps other translators, I hope we'll receive a few informative answers to this.
Oliver
| | |
|
|
|
| Impressive posting | Apr 28, 2013 |
Nice posting, Samuel. It gives outsiders like me a quick update of the current state of OT. Why is it that OT doesn't offer F/R? For me this is one of the features I use most often in my Java CAT tool (CafeTran). Doesn't development consider this to be a crucial part of any CAT tool?
About icons: CafeTran introduced the concept of smart icons, that only appear when relevant:
... See more Nice posting, Samuel. It gives outsiders like me a quick update of the current state of OT. Why is it that OT doesn't offer F/R? For me this is one of the features I use most often in my Java CAT tool (CafeTran). Doesn't development consider this to be a crucial part of any CAT tool?
About icons: CafeTran introduced the concept of smart icons, that only appear when relevant:
http://www.youtube.com/watch?v=ERMjR-TQLCY&feature=youtu.be
BTW: This is CafeTran's approach of minimalistic design:
http://dl.dropbox.com/u/15919910/MBP/Screen%20Shot%202013-04-28%20at%207.56.46%20AM.png
Hans ▲ Collapse
| | | | Samuel Murray
Netherlands
Local time: 14:32
Member (2006)
English to Afrikaans
+ ...
TOPIC STARTER | On-the-fly is actually a misnomer, and OmegaT's status in this regard | Apr 28, 2013 |
Oliver Walter wrote: Samuel Murray wrote:
Apart from Wordfast Classic, how many tools do you know of that allows self-segmentation on the fly? I don't know which CAT tools they are but, because this facility is important to me, and perhaps other translators, I hope we'll receive a few informative answers to this.
1. Well, now that I think about it, most CAT tools that I have seen do have it, in some or other way. They may visually pre-segment the entire file but that doesn't mean you can't resegment at [almost] any particular point in the file, and that would give you the same satisfaction as non-presegmented on-the-fly segmentation.
One reason why I like non-presegmented on-the-fly segmentation is that it helps me to see the untranslated section of the file as a large blob of text and not as chunks that have already been processed. This makes it easier for me, mentally, to decide when to change the segmentation. But I suppose if a CAT tool offers a seamless, easy way to change the segmentation at any given point, then that would be sufficient for most people who prefer on-the-fly segmentation.
2. OmegaT can fake this, but ultimately it depends on how complex the segmentation requirements of the document is. In OmegaT, you might think that you're translating documents (i.e. if you don't know the inner workings of the program, you it would seem to you as if you're translating the files individually), but in reality you're just creating a translation memory, with some segmentation rules.
When you press the button that causes the translated files to be generated, OmegaT actually re-translates those files against the translation memory that you had created for that project. This means that you can't add an exception in one place if the same sequence of text occurs in another place. Since a recent version, OmegaT does offer in-context matching to help with this problem, but it only works if the unique context is in the immediate vicinity.
If you have a source document that consists of ten identical pages, but you have to translate page 3 and page 7 in a slightly different way, then you can't do that with OmegaT... because the in-context matching context doesn't stretch that far, and the segmentation rules can't rule about anything that goes beyond a paragraph mark.
Wordfast Classic can do this because you're translating the actual source/target file. CAT tools that use a bilingual file format can [usually] also do this because you're translating an actual file, and the final target file will reflect the bilingual file's segments in the same sequence as they are in the bilingual file, even if identical segments are translated differently. OmegaT can't do that (yet).
| | | | Samuel Murray
Netherlands
Local time: 14:32
Member (2006)
English to Afrikaans
+ ...
TOPIC STARTER | Lack of find/replace in OmegaT | Apr 28, 2013 |
trhanslator wrote:
Why is it that OT doesn't offer F/R? For me this is one of the features I use most often in my Java CAT tool (CafeTran). Doesn't development consider this to be a crucial part of any CAT tool?
I can only speculate, but I don't think it is fair to say that the developers don't consider it crucial. There are a limited number of developers with limited time resources, and they focus on things that they consider important (and fun to code, I suppose). Some things consume huge amounts of coding time even though the visual effects are minimal, e.g. the new XLF filter (which I don't even use).
Another thing that I can say (speculatively) is that in OmegaT there has been an attitude that it's better for the developers to focus on features that you simply can't do in other free tools, and give a lower priority to features that you can do in other free tools. For product development projects with scarce resources, this attitude makes sense.
In this line of thinking, find/replace can be done in a TMX editor... although you would have to close the project before you perform the find/replace action, and reopen it again afterwards, and you wouldn't be able to see the stuff you find/replace in a visually pleasing way.
| | | | There is no moderator assigned specifically to this forum. To report site rules violations or get help, please contact site staff » OmegaT v3 released (my write-up) | Trados Business Manager Lite | Create customer quotes and invoices from within Trados Studio
Trados Business Manager Lite helps to simplify and speed up some of the daily tasks, such as invoicing and reporting, associated with running your freelance translation business.
More info » |
| | Wordfast Pro | Translation Memory Software for Any Platform
Exclusive discount for ProZ.com users!
Save over 13% when purchasing Wordfast Pro through ProZ.com. Wordfast is the world's #1 provider of platform-independent Translation Memory software. Consistently ranked the most user-friendly and highest value
Buy now! » |
|
| | | | X Sign in to your ProZ.com account... | | | | | |
Login to reply/comment 
