 XLIFF TC Meeting (Conference Call) |
|
| Name | XLIFF TC Meeting (Conference Call) |
| Time | Tuesday, 26 May 2020, 08:00am to 09:00am EDT
(Tuesday, 26 May 2020, 12:00pm to 01:00pm UTC) |
| Description | New dial-in link at this action item. https://www.oasis-open.org/apps/org/workgroup/xliff/members/action_item.php?action_item_id=3921 |
| Minutes | Agenda I. Administration Attendance: David F., Rodolfo, Lucía, Bryan. Quorum achieved. D: I can change the current video conference system to Microsoft Teams R: yes, that would be a good idea. D: I will do it next time.
A. Approve 28-April meeting minutes https://lists.oasis-open.org/archives/xliff/202005/msg00000.html R: I second. D: Meeting minutes approved.
B. Meeting availability in light of COVID-19 II. Where do we go next? D: This has already been discussed. It can be closed. R: Agreed.
B. Rendering Requirements for XLIFF 2.2 - David R: I have a request. I remember years ago, maybe 20 years ago. The whole idea of XLIFF was to separate the text from the formatting. Let the translators concentrate on the text and forget about the rest of the code. D: this is a different thinking. The rendering practices have been developed since, the tabular display, etc. I agree on the original idea of XLIFF, but it does not mean that the translator has to work without context. R: I understand the translator needs context. But, that is independent from xliff. There are tools that allow you to have a nice preview in the original format. So whenever the translator makes the change in the segment, the change is displayed in the sidepanel. It is being done. XLIFF does not prevent this. D: I am just trying to get to the picture. [David shows the paper written on rendering requirements]. It describes what you need to do for the canonical representation, it has several examples, such as GUI mockups. It is not only about displaying the context in relation to the target format, but it gives you structure information, where it is in the xliff. This is designed to ordering. R: [Rodolfo shares his screen and shows the preview option in a CAT tool, the full rendering is done in the tool] D: How does it work? R: I have the original file, a skeleton is created on the fly when the change is made and the preview is displayed. D: I do not mean that XLIFF should have those capabilities. The other thing is how to display the translation environment. You are displaying source/target, inline content. You are an experienced developer, you do not need to be told how to do it. The goal is to target the other developers that do not know how to render xliff. R: Ok. D: For previews with absence of the original format. You can use tool specific. I think what is important is to explain how to represent elements like placeholders. R: All tools that support xliff support placeholders. D: There are different approaches. There is also with pseudo span, you still need to interpret it as span. So guidance for that would be helpful. R: for whom? D: for both. The translator needs to see that code span. R: the translator can see them live. [Rodolfo shares again his screen and shows how translators could see the content of the placeholder] D: And what happens if you violate the order of the tags? R: yes, you will get an error. CAT tools work like this. D: You need proper extraction. This is the idea that the paper has, to know how to display it. R: Yes, CAT tools have graphic representation to represent them. Rendering information for translators is good. D: It is all in the paper. It can be used by the main developer, so the translator has all the information relevant. [David is referring to reordering] R: You need to know why? Because when you try to implement it you realise it is a bad idea. When you start to do reordering, translators will not like it. If they use sentence segmentation, they will end up with mismatches in the TM. D: An example, for example, could a table with cities in alphabetical order. R: Exactly, you could do the reordering outside XLIFF. D: We have this order attribute. R: It is not implemented because it is a bad idea. D: Translators cannot use it because is not implemented. R: You have a fact, it is not implemented by CAT tools, and that tells you that it is a bad idea. D: I do not agree, maybe they do not implement it because they need guidance on how to do it. There should be a push from the client side. R: Are you going to target the client? D: We have an on and off pulse. R: I am asking all these things, because I want to work on XLIFF 2.2 on things that are relevant. D: Another thing is the hotkeys representation, ITS categories display. R: Terminology is displayed in CAT tools in a dedicated panel. A translator tool has different panels for different elements: comments, notes, terminology, etc. B: I assume whether we are discussing now if we should have the rendering as a new feature. R: the original goal of XLIFF is to separate text from formatting. If you are interested in doing it, there are tools already implementing it. So, I do not see why we need to include it XLIFF. L: So, what happens if that XLIFF file is transferred to another tool different from the one that has the same preview option? R: If the preview option is also implemented, it could be displayed as well. D: the rendering guidance does not need to change anything on XLIFF. It just presents guidance on how to render XLIFF. It will also help to support features that are available. XLIFF 2.O was also designed to be an ecosystem-oriented standard. That can travel through roundtrip and add useful information that might come from different tools. Rendering goes in this line. To allow translators to work with whatever TMS they like. R: But do you think that companies are interested in doing this? D: I do not think vendors might be interested because they interoperability is not in their best interest. R: But who will implement it then? D: the problem is that translators do not have power to push for this. I have been talking to Alan Melby, and he might be interested in coming back to TC to represent translators.
G. Features, Steven Loomis D: Another topic. I was going to contact Steven. Rodolfo, I know you are more interested in SRX exchanges. Steven is interested in different segmentation methods in relation to ULI. I am not sure how this is compatible. If SRX are expressible enough. R: they have been useful for many years. D: Ideally, this feature needs to do both. A method for representing segmentation methods. R: Ok. Let’s see what Steven has to do say on this.
D: I think the technical solution was sound, but it does not have support, so I am sceptical. Another interesting feature is being able to display semantic information about placeable. Specially for <ph>, it can be very semantically meaningful, like a variable for product name. Semantic description would be displayed to the translator instead of the placeholder.
B: We have reached the hour, apologies for my late arrival. D: We have had a good discussion. |
| Agenda | I. Administration A. Approve 28-April meeting minutes https://lists.oasis-open.org/archives/xliff/202005/msg00000.html B. Meeting availbility in light of COVID-19 II. Where do we go next? B. Rendering Requirements for XLIFF 2.2 - David C. GALA TAPICC XLIFF Extraction and Merging Best Practice v1.0 Allison D. New comment from Yves E. Invalid sample files identified by Rodolfo Raya F. XLIFF calls, Steven Loomis G. Features, Steven Loomis H. FragID prefix registration, Rodolfo III. Subcommittee and sister TC reports IV. New business |
| Submitter | Bryan Schnabel |
| Group | OASIS XML Localisation Interchange File Format (XLIFF) TC |
| Access | This event is visible to OASIS XML Localisation Interchange File Format (XLIFF) TC and shared with
|