Improving BT Quality Discussion

[enn_are]

Wow, this thread is exploding. Might as well throw in my 2c.

Firstly, I applaud your desire to improve translation quality. I think both translators and readers benefit from having a good editor go over their work, and if this initiative results in more of that then good job.

Having a "certified" stamp as a means of signalling to the community that certain projects are higher quality is probably the least disruptive way of doing so. However, if the community* cannot understand that BT is not a translation group, will they understand the difference between a "certified" project and one that is not? In reality, it'll probably be a spectrum - some will understand, some won't, but, hey, since we're judging collectives by their lowest common denominator Even if the community doesn't understand, IMO the project will still benefit from it, so I see it as worth doing regardless.

Personally, I'd rather see a positive representation of translation quality, eg. something like having a ☆ after a chapter link to show edit / ☆☆ to show edit + TLC by "certified" people, instead of the red symbols proposed. There's a couple of reasons for this:
1) Stars are typically associated with quality / rating, so should be immediately understandable.
2) It provides differentiation from existing projects without requiring an update to all 119 existing projects.
3) If it made sense and didn't clutter things too much, the stars could be links to wiki user pages / have hover-over tooltips to provide permanent and visible recognition to editors / translation checkers.
4) Personal preference for positive reinforcement

Anyway, good luck to you, and I hope you can achieve the improvements you're seeking.

* I don't really understand what the community that various people refer to is, or why they're trying to appease them. FWIW I'm a translator on a niche project who does this in their spare time for a bit of fun, so the audience of people that I care about is comprised of exactly one person: me. But I recognize that not everyone's motivations for participating are the same as mine.

[TheCatWalk]

I don’t know actually know how ND, Heretic, and other outside groups operate with their editors, but I assumed the editors work more closely and discuss more with the translators than is common on BT. The fact that the editors and translators have a more formal arrangement that encourages that to happen is, I assume, one factor that enables them to have more consistent quality.

well, irc helps a lot in this case and they both have them, ND being specially active in this translator-editor interaction while BT has.............an irc group with lot of member?

And to emphasize something that might have been overlooked: This would be optional. A project chooses to follow these rules and get certification if it wants to.

exactly my point.

[cloudii]

Alright, so I've put together a formal project proposal.

EDIT: (http://notepad.cc/share/DPCJv7RrRn) Please see Simon's post below for a prettying rendition of this ugly plain text.

I think... I'm going to roll over a collapse now. |D

xD It seems like feedback has mostly been... semi-positive, neutral, and skeptical. No flat-out no's. I think sometime I'll try and email TLG, but does anyone know how I can find his email?

Apart from that, I'd like to eventually get a poll from the Translators, Editors, and Supervisors about whether they're interested in this effort as a whole. My suggestion involves a small bureaucracy of people to review diagnostic tests (primarily involving individuals who currently have access to the Administration Forum). I'd like to know how many of you would actually be willing to participate in that. See the formal proposal for the details of how it'd work if it did happen.

Otherwise, this thing would generally run by itself without any supervision except the by the individual Projects that implement it.

[Simon]

As.. the other links looks like a mess(windows can't handle only \n).

Here: Improving BT Quality Discussion

To destroy is easier than to create. The hardest part will be, making the 8th layer understand what this is.

On other hand:

Spoiler! :

Big boobs are filed with dreams.
But small once are filled with hope.

So it should be given a try. That's what I think.

[onizuka-gto]

Alright, so I've put together a formal project proposal.

http://notepad.cc/share/DPCJv7RrRn

I think... I'm going to roll over a collapse now. |D

xD It seems like feedback has mostly been... semi-positive, neutral, and skeptical. No flat-out no's. I think sometime I'll try and email TLG, but does anyone know how I can find his email?

Apart from that, I'd like to eventually get a poll from the Translators, Editors, and Supervisors about whether they're interested in this effort as a whole. My suggestion involves a small bureaucracy of people to review diagnostic tests (primarily involving individuals who currently have access to the Administration Forum). I'd like to know how many of you would actually be willing to participate in that. See the formal proposal for the details of how it'd work if it did happen.

Otherwise, this thing would generally run by itself without any supervision except the by the individual Projects that implement it.

Please create a page on the wiki and publish it there, then link it back here.

I will review it before the end of this month, if there is sufficient discussion for me to consider it valid, i will allow it to go to a community wide vote on the forum and wiki.

[Teh_ping]

Oni, Cloud has sent an email to Big Boss, CCed to you and me. Please review that instead.

[krytyk]

While I agree that with the overall quality of translations is not the highest, except for a few major projects ran by senior translators (cheers folks). I completely disagree with policy of having to get certified in any way. Translators are doing everything in their own free time, out of their own will. Trying to restrict them further would be counterproductive, and would be against what baka-tsuki itself is. Saying that out of my own personal viewpoint on the baka-tsuki as a whole, I would not post any of my translations if it were so, it does not sit with me for someone to look over my arm and check me, I'm student no longer - don't need people to grade me. Like someone said before, we should be happy with the people willing to translate anything at all and putting it on the wiki, putting artifical barriers there would be like putting a middle finger towards everyone who used their free time to contribute.

As for translation quality itself, I do agree that the translations are often pretty poor, but the way to improve the translations quality should be different. First, translators should be encouraged to do more extensive research on terms in the LN. Second, translators should make specific project guidelines that includes those terms, explanations of how and why they are used. Third, new translators should have an easy way of contacting, consulting terms they are unfamiliar with senior translators (cheers folks) to improve the quality - especially with machine translations that are currently a major part of the TL contributions.

So yeah, while I agree to the notion of improving translations quality, I completely disagree with most of the cloud's proposals as they go against what Baka-Tsuki is.

Instead, I think we should look forward towards more fitting ways of TL quality improvement, such as encouraging translators themselves to improve, help them out with their translation methodology, and create environment for them to consult and discuss translations with senior translators. Improvement of translators skill inevitably leads to better quality of translations.

~This post is based on my personal feelings as a reader and editor, which I've been for years now, and not on my feelings as a recently-started translator.

[Nurin]

Hmm... From what I understand of clouds proposal, it's not like you need to get certified to translate. Like, it's not every project that is going to take this on, and for the projects that take it, new translators can still contribute, but their translation will pass through TLC, case that might not happen if you are a certified translator and bla bla bla(Read clouds proposal for more details)...

[krytyk]

Hmm... From what I understand of clouds proposal, it's not like you need to get certified to translate. Like, it's not every project that is going to take this on, and for the projects that take it, new translators can still contribute, but their translation will pass through TLC, case that might not happen if you are a certified translator and bla bla bla(Read clouds proposal for more details)...

Even so, the other part still stands.

That proposal hardly includes anything that will actually improve quality. If there's no willing TLCheckers, then there will be actually nothing at all in all of that. You improve quality by helping translators to improve their translations, not by rating them nor dividing them betwen certified or uncertified, or testing them. I don't see anything that would help a below average translator to improve his own translation quality, however I do see things that would discourage them and make them think they're worse than the certified people. Also I'm not seeing hordes of willing TLCheckers waiting to get to work. The people who have qualifications to do it are working on their own projects already, and while I think they might help out here and there, I'm not seeing them dedicating themselves to TLC.

[Nurin]

Hmm... From what I understand of clouds proposal, it's not like you need to get certified to translate. Like, it's not every project that is going to take this on, and for the projects that take it, new translators can still contribute, but their translation will pass through TLC, case that might not happen if you are a certified translator and bla bla bla(Read clouds proposal for more details)...

Even so, the other part still stands.

That proposal hardly includes anything that will actually improve quality. If there's no willing TLCheckers, then there will be actually nothing at all in all of that. You improve quality by helping translators to improve their translations, not by rating them nor dividing them betwen certified or uncertified, or testing them. I don't see anything that would help a below average translator to improve his own translation quality, however I do see things that would discourage them and make them think they're worse than the certified people. Also I'm not seeing hordes of willing TLCheckers waiting to get to work. The people who have qualifications to do it are working on their own projects already, and while I think they might help out here and there, I'm not seeing them dedicating themselves to TLC.

Ok, you do have a point, but I also think that it could bring some new translators that would like the new system... How they call it? Two sides of a coin? Double-edged knife(Or sword whatever)? Well, but I would also say, that the actual B-T, is not able to make Cloud's propose real, that's because only a very few people would join. Like you said, the more reasonable would be create a environment where we could help new translators... But there's also a problem... If you look around the forums, there's loads of people willing to help, but the problem is, most of people that need help don't give a f**k about it... So, Well... I won't say more than this...

[EusthEnoptEron]

As for translation quality itself, I do agree that the translations are often pretty poor, but the way to improve the translations quality should be different. First, translators should be encouraged to do more extensive research on terms in the LN. Second, translators should make specific project guidelines that includes those terms, explanations of how and why they are used. Third, new translators should have an easy way of contacting, consulting terms they are unfamiliar with senior translators (cheers folks) to improve the quality - especially with machine translations that are currently a major part of the TL contributions.

Regarding terms and guidelines: I think those are covered pretty well already. Hell, there's even a dedicated forum for them. I completely agree that there should be an easy way of communicating and exchanging knowledge, though. In my own projects, I've often noticed that there's a lot to discuss, imperatively.

However, I'm looking at it from a grammatical / stylistic perspective, since I agree with larethian et al that TLC is awfully unproductive. Also I assume that translators know their stuff and are able to read and interpret most of the original work correctly; I'm more bothered by the low level of English in the translations, be it because of a lacking command of the language or trade-offs made in order to save time. Fresh translations tend to be in a very raw state -- and that's entirely OK and cannot be avoided because of the frequent (language) context switches -- but as things stand, those "drafts" often stay that way without significant changes (correct me if I'm wrong). No one really bothers to improve them because when the translation is done, the demand for moar is satisfied and the core consumer group is happy. In teams where translators and editors work hand-in-hand, I think, the translations often don't even get released until both parties are content with their work, but that's not how projects are usually handled on B-T. If we had a seal of quality of sorts, there would be a target for editors to aim for that would take the translations out of the pool of ambiguity and incompleteness they are drowned in.

Now, it's entirely possible (and likely) that this wouldn't change anything, and the constant lack of editors (and translators, for that matter) would persist, leaving us with a myriad of pages awaiting an edit pass, but I'm generally open to the idea. I guess it depends largely on the support / reception from the community (old and yet-to-come). Oh well, so much for my 2 cents. I should get down to doing something that's actually meaningful. :D

[cloudii]

If we had a seal of quality of sorts, there would be a target for editors to aim for that would take the translations out of the pool of ambiguity and incompleteness they are drowned in.

I really like the way EusthEnoptEron worded this. <3 Let me just respond to a few things.

CLAIM: Translators would feel demotivated if their translations are "graded".

• Misconception 1: Individual translators are not graded. Instead, individual projects/chapters are recognized for outstanding quality. This is similar to how Wikipedia implements a a "Good Article" Recognition system, which ties into their system of recognizing "Featured Articles".
• Misconception 2: Chapters will be flagged with a binary system. An initially translated chapter either has been EDITED, or has not. This is no indication of translator skill or quality.
• Misconception 3: The TLC-senpai or Editor-senpai (mentioned in the proposal) would not be running around "grading" translations. They will continue to act as regular editors and TLC-ers and work on what they want. They will not be PM-ing translators and saying: "Your translation for Chapter 4 is a B+". They're just going to work on a flagged chapter and directly make the changes, just like any other editor. These "-senapi's" only have the right to remove the binary flags.

CLAIM: This system would divide translators into two classes, etc. People would be unhappy if they are part of the lower class.

• Misconception 1: The diagnostic system is a mechanism that will recognize/promote individuals to something along the lines of a "senior editor". This is not meant as a system to ostracize non-certified editors. It is merely a system to recognize individuals who excel.
• Analogy: Are regular soldiers in the military supposed to be unhappy if they do not receive a "Medal of Honor"? The Medal of Honor is a high distinction. Becoming an "editor-senpai" or "translator-senpai" should be thought of as similarly.
• Analogy: Larethain (not to name any names, sorry!) is in the "Senior Translator" Usergroup by the definition of the Baka-Tsuki Forum. If you are a translator, are you supposed to be upset that Lare-tan is a "Senior Translator", while you are not? Lare-tan worked hard for his "Senior Translator" status, and is absolutely qualified for it. I don't think any translator would get upset over this, especially since we're all so used to it. I'm just proposing a mechanism in which members can be promoted to this higher status (which I've called a "-senpai" in the proposal).

CLAIM: This system doesn't help translators improve at all!

• Misconception 1: This statement is true. However, this is only because this system is targeted at improving chapters, not improving individuals. This system only provides a organized means of attracting skilled individuals to eventually edit/TLC an initial raw translation.
• Why not directly encourage translators/contributors to improve? Because this is aggressive, and can be the very de-motivational thing many feared to see. Why? Because as a novice translator myself, I don't want people nagging at me to improve. I'm trying my best already to slowly improve my language skills, and it'd be annoying if people repetitively tell me to improve. On the other hand, I personally have no problem if a skilled Editor/TLC-er eventually checks over my translation. At least they won't tell me I suck. They just make the changes. No communication necessary. As a matter of fact, my moral conscience would prefer it if my translations are eventually checked (hence why I want a flagging system). I get nightmares thinking my mistranslations might stay as mistranslations forever... on the internet!
• Why not provide resources to help translators improve? Honestly, we have them already. There's a "General TLC Thread" in Aux Brigades. Does it get used? Not really. We have IRC. Does it get used? ...well, really only for unproductive things. None of this is bad, but it barely causes an impact on the overall quality of translations. I'm looking for a system that might improve the quality of Baka-Tsuki's translations over time.

CLAIM: I don't think this will change anything b/c we don't have enough staff!

• YOU ARE ABSOLUTELY CORRECT! If no members are interested in taking the diagnostics to become "-senapi's", nothing will change at all!
• But then again, what do you have to lose?

[hayashi_s]

Setting aside the obvious problem of where the TLCs are going to come from, who is going to set the criteria over what makes a "quality" translation? It's very easy to say you're going to have a gold standard, it's less easy to determine what the requirements for a gold standard is. In fact the two problems are linked, people who TLC are more likely to be able to judge whether a translation qualifies, and while this proposal looks good on paper, the lack of qualified TLCs might turn this into a white elephant.

[krytyk]

Setting aside the obvious problem of where the TLCs are going to come from, who is going to set the criteria over what makes a "quality" translation? It's very easy to say you're going to have a gold standard, it's less easy to determine what the requirements for a gold standard is. In fact the two problems are linked, people who TLC are more likely to be able to judge whether a translation qualifies, and while this proposal looks good on paper, the lack of qualified TLCs might turn this into a white elephant.

Yeah well, it's basically that.

Also, I'm not sure if the topic's name is fitting contents as it is. Currently it's one person pushing and defending his/her proposal, and other discussing it, there's no other proposals. It's more like an audit over one idea rather than a real discussion about how to improve the quality itself.

Like I said before, I agree with notion of improving translation quality, however if "discussion" is reduced to this state - of trying to push one proposal and desperately defending it, then I have to excuse myself from participating.

This is just a remainder that certain someone named the topic "Improving BT Quality Discussion", thus I expect it to be more of a discussion than an audit. Especially that this topic is worth discussing seriously and is something that is necessary.

[pudding321]

I have gone through the points and the plans, and I hope someone can answer me some questions.

1. There won't be enough people to support this, for most projects. For ones with enough people, I don't think there's really a need for this to improve the quality anymore.

2. Most project supervisors and editors know the 'golden quality' of the translations and what is best for the project. Giving it a seal RESTRICTS its standards.

3.Where even basic editing or typo-filtering is lacking, asking for people to do it more systematically is over-demanding and unrealistic.

It just looks more like showing viewers what's qualified or not as a bid to raise BT's reputation, which I don't really care. I don't find any special trick in this that will improve the quality of the translations themselves at this point. I believe editors or translators jump in to projects mostly because of their interest in the project itself, not because of its translation quality or whether it needs a stamp of approval.