Improving BT Quality Discussion

Do you have a fully fleshed idea that you think Baka-Tsuki should adopt? Post it here.

Moderators: thelastguardian, Fringe Security Bureau, Senior Editors, Senior Translators, Alt. Language Translator/Editor, Executive Council, Project Translators, Project Editors

Post Reply
Detalz
Devoted Haruhiist
Posts: 55
Joined: Thu Jul 26, 2012 5:44 pm
Favourite Light Novel:

Re: Improving BT Quality Discussion

Post by Detalz »

The discussion seems to have reached a standstill and only seems to be going around in circles now. So I'll reiterate my thoughts and provide a slightly new suggestion. Again, the main problem I see with Cloud's suggestion is the lack of people willing to take responsibility in actively regulating all of Baka-Tsuki's projects. Of course, there's way too many projects to regulate and it'd be impossible for a small group to do it all. As such, I think a bottom-up approach would be best, where we slowly work our way up to these changes.

The main point I think we should start with is this
1) A complete overhaul in the responsibilites assumed by a Project Supervisor.
  • a)I don't think creating a project page and translating only one chapter should qualify anyone as a Project Supervisor. A project supervisor should be committed to their project, and willing to commit long term. Ideally, a Project Supervisor should be someone who possesses some knowledge targeted language the novel is being translated from and have contributed some translations him/herself. However, I think a Project Supervisor should possess a minimum of a native level of English(though not necessarily better than an editor's) so he is able to properly supervise the edits being made to the project. I think it's a bit silly to name yourself as Project Supervisor and do nothing. The roles of a project supervisor will include:
    • coordinating communications between editors/translators and ensuring no excessive changes were made while ensuring optimal translated content.
    • possibly performing TLC personally; this is a must if a translator does not have the time to partcipate directly in editing/correcting these mistakes. In these cases, it'd be best if a translator personally nominated a proxy to act as the Project Supervisor(if there isn't already one)
    • recruitment of editors/TLCs. There should also be a standard set for a way to actively recruit editors. Maybe an announcement saying "This project is actively seeking editors and TLCs!" in the project page, linking to a site that explains how to register as one, and test if they meet the necessary requirements.
    b) The primary translator may assume the role of Project Supervisor at any time, especially if he feels the person in question has been inactive or not contributing to his role.
2) A system to keep track of which chapters have already undergone. Someone can make a template for a checkbox or something, that can easily be copy/pasted next to a chapter in the Registration Page, telling which chapters have already undergone edits, TLC, or both

3) Going back to a bullet point mentioned in 1), I think there should be a standard testing criteria for editors. This can be a bulk of unedited, and directly translated content, or some universal testing questions. Supervisors can also alter this testing criteria to their liking in a way that best fits the style of their project/compensate the errors of a translator's writing style.
User avatar
Hiyono
Shamisen Wordsmith
Posts: 312
Joined: Thu Jul 31, 2008 12:55 am
Favourite Light Novel: Ahouka!
Location: The Internets

Re: Improving BT Quality Discussion

Post by Hiyono »

I think your remarks on the role of Project Supervisor are entirely on point. That said, I think it comes down to responsibility. No one really wants to bear all that responsibility for what essentially boils down to little to no gain. In fact, that's why this whole argument's simply been spinning in circles - no one wants to be proactive about it because of all the responsibility implied. Stated in game-theoretic terms, we have the snowdrift problem; if everyone works together, we all benefit, but it's also easier to wait for someone else to start.

*I'm not judging anyone, this includes me as well.
"There is always an easy solution to every problem - neat, plausible and wrong." H.L. Mencken (1971)
obsidian pegasus
Project Translator
Posts: 421
Joined: Thu Jun 27, 2013 9:00 pm
Favourite Light Novel:

Re: Improving BT Quality Discussion

Post by obsidian pegasus »

One problem with Detalz suggestion is most projects on here are already well on their way with a ton of work on them already done. A system already in place resist change (sounds like interia), so Using Detalz suggestion of starting Bottoms up on a project, we need a teaser project that will develop into a full project with both a translator, project supervisor, and editor (3 seperate people) who are willing to commit long term to this project.

Does anyone reading this have a project they are about to start then? One they would be willing to adhere to this format on?

Editor test-I have a favorite editor on the boards and talk to them about whatever I do, I found them by looking at the way several people edited and its true, some people...suck. So I see the benifit of an Editor test. Perhaps taking the prologue of the work to be started and posting it here in the forums and comparing responses to see how people stack up in their edits.

Translator test- I'm all for this system being tried at least, but how do I test myself here? (I am biased after all). Also, While I do have a teaser I wanted to start soon (I couldn't be a hero so I got a job), I'm notably slow in my work and wouldn't be able to test this system within a reasonable amount of time, so it Will have to be someone else, (though I'd love to help, its taken me 1-6 months to put a chapter out, lots of guys do a chapter/day...wow).

Project Supervisor/TLC-....not sure how to qualify someone for this post, since as was said earlier, nominating someone doesn't seem right. Maybe have a main editor fill this post with a second editor as the project Editor? (note: that means 2 editors on a project, which is plenty).

Reiteration: Does Anyone have a project (and a translator and an editor) in mind which could test these new standards for the group, (getting the groups feet wet)???
黒曜石ペガサス
私の悪い文法のため申し訳ありません
User avatar
cautr
Tsuruya Gigglenator
Posts: 740
Joined: Mon Sep 16, 2013 10:41 am
Favourite Light Novel:

Re: Improving BT Quality Discussion

Post by cautr »

obsidian pegasus wrote:Project Supervisor/TLC-....not sure how to qualify someone for this post, since as was said earlier, nominating someone doesn't seem right.
To fit this role in this idea it'd be good to have someone proficient in English at least, ideally have someone who can TLC the work, which'd result in reading the raw, then reading it in English. Double the work, quarter the fun. So at most someone proficient in English could be found who then just reads the new content and tells the editors if it reads all right or not, in the end not doing anything but reading and complaining, a manager so to say. :P

I do have a project which will only just start, but I have next to no faith in that system, so I won't apply for it. I don't see why we would try to enhance every little project that starts instead of, for now, take a small step and label the projects which are great.

So, teaser -> full -> certified project. Someone with the authority has to apply for a project to be a certified project, on who then 5-10 people discuss in IRC whether or not it is up to the standards (good/great readability, formatting). We can't TLC every project anyway, so for now small steps which help improve somehow would already help. The people who decide on this could be suggested on some thread, then accept the nomination (or not) and then be put in with the rest of the nominated ppl (unless they refused) and be voted in a poll to be part of the group.

They then decide about a project which applied for the certification, if there's more than one application, from most popular to least. Certified projects could be listed on top of the sidebar and have their updates shown on the frontpage (maybe, maybe not). Also get a nice logo on their project page (or not).

The responsibility for the project to hold up this quality will then be assigned to the person who applied this project for the certification test, if this person becomes inactive (not available for 2-3 months), the certificate "runs out". Of course projects like Campione or similar stuff can't be tested as a whole. So the guys who decide on the project read different reading samples from different novels and judge by those. Oh, an odd number would help to avoid ties, I guess.

Well, this is my rough thought on the whole stuff. This could also apply to all projects, if the deciding group just goes from most popular to least popular project (only full, no teasers, though) and does the above test, if it's just not readable, the project becomes a teaser with the note of "poor english" (so cruel :D).

This all might be somewhat confusing or messy or w/e, I'm sick, be nice to me. :roll:
Translator at guhehe.TRANSLATIONS

"Life's a tale, told by an idiot." -Shakespeare
User avatar
Mystes
Heaven's Blade Successor
Posts: 15932
Joined: Thu Aug 05, 2010 6:54 am
Favourite Light Novel:
Contact:

Re: Improving BT Quality Discussion

Post by Mystes »

If we want to test this out, we can simply pick an ongoing project that's being TLed constantly and that looks reasonably well-done.
Kira0802

#campione at rizon for some #campione discussions~~ And other stuffs.
rayquest111
Reader
Posts: 6
Joined: Mon May 20, 2013 3:32 am
Favourite Light Novel:

Re: Improving BT Quality Discussion

Post by rayquest111 »

really BT IS THE BEST FOR ME

i started reading LNs with DXD
man i really love it

as for the improving BT quality is must but one must also keep in mind the limit number of excellent translators we have here ?
:P

they have there own life too and if the quality standard goes up so there work increase to who are applauded for there self-less and awesome work

secondly, as we improve the quality i would suggest that why not train community members in translating these light novel (i wish if i knew Japanese i would really help in this)

even if few members translate 3-4 pg each then it would speed up the process
which includes grammar and language check to

the bottom line

as for me i am content with both improved and current quality work done by BT community
tc :D :)
User avatar
Hiyono
Shamisen Wordsmith
Posts: 312
Joined: Thu Jul 31, 2008 12:55 am
Favourite Light Novel: Ahouka!
Location: The Internets

Re: Improving BT Quality Discussion

Post by Hiyono »

Kira0802 wrote:If we want to test this out, we can simply pick an ongoing project that's being TLed constantly and that looks reasonably well-done.
That would be entirely meaningless. The whole point of this is to "improve ... quality". Unless you choose a project which could noticeably use an improvement, there's no point. If you start off with a good quality translation, and end up with a quality translation, you've proved...what? The only way you show the effectiveness of this proposal is to take a translation which needs help and put it through its paces, and then if the end result is substantially improved, you've shown it works. Nothing else will do.
"There is always an easy solution to every problem - neat, plausible and wrong." H.L. Mencken (1971)
obsidian pegasus
Project Translator
Posts: 421
Joined: Thu Jun 27, 2013 9:00 pm
Favourite Light Novel:

Re: Improving BT Quality Discussion

Post by obsidian pegasus »

At the risk of offending someone, can you pm me a project you feel is sub par so I can have an idea of the low standard your looking at Hiyono?


side note
Rayquest, if your serious about thinking of starting to learn japanese, there's a few forums on here about ir, or pm me and I can tell you what I'm doing, (since I'm self taught/still learning). I even know some easy projects you could help on in 6 months or so... :D
黒曜石ペガサス
私の悪い文法のため申し訳ありません
User avatar
Mystes
Heaven's Blade Successor
Posts: 15932
Joined: Thu Aug 05, 2010 6:54 am
Favourite Light Novel:
Contact:

Re: Improving BT Quality Discussion

Post by Mystes »

Hiyono wrote:
Kira0802 wrote:If we want to test this out, we can simply pick an ongoing project that's being TLed constantly and that looks reasonably well-done.
That would be entirely meaningless. The whole point of this is to "improve ... quality". Unless you choose a project which could noticeably use an improvement, there's no point. If you start off with a good quality translation, and end up with a quality translation, you've proved...what? The only way you show the effectiveness of this proposal is to take a translation which needs help and put it through its paces, and then if the end result is substantially improved, you've shown it works. Nothing else will do.
I think it's quite the opposite. If we're going to choose the worst project on BT, the translator and the editor probably won't pass. And if they don't pass, there'd be no way to determine what a "standard/acceptable" BT project should be like to set a comparable.

The experiment should be something what an usual project would go through, since the intent is to certify good translation. It's not a guillotine.
Kira0802

#campione at rizon for some #campione discussions~~ And other stuffs.
IlkatSumil
Reader
Posts: 5
Joined: Sun Jan 12, 2014 10:55 am
Favourite Light Novel:

Re: Improving BT Quality Discussion

Post by IlkatSumil »

rayquest111 wrote:even if few members translate 3-4 pg each then it would speed up the process
I can't see how swapping between translators every four pages would improve translations. Not only does everyone have their own style, written Japanese is a really hard language to translate if you're not following the context closely. Even when you know what's going on, you have to make all these interpolations with pronouns, subjects, tenses, object/person numbers and indefinite vs. definite articles or you end up with unnatural and vague English. Every translator contributing a few pages would have to read the entire source in Japanese anyway, or be very familiar with the novel in another format.
User avatar
Hiyono
Shamisen Wordsmith
Posts: 312
Joined: Thu Jul 31, 2008 12:55 am
Favourite Light Novel: Ahouka!
Location: The Internets

Re: Improving BT Quality Discussion

Post by Hiyono »

Kira0802 wrote:I think it's quite the opposite. If we're going to choose the worst project on BT, the translator and the editor probably won't pass. And if they don't pass, there'd be no way to determine what a "standard/acceptable" BT project should be like to set a comparable.

The experiment should be something what an usual project would go through, since the intent is to certify good translation. It's not a guillotine.
Fine, honest opinion time. I don't buy it. I don't believe that can work, period.

Let's reexamine what happens when the quality assessment is performed. Most immediately, a binary classification of "good" and "bad" projects results. I know, I know, we're just "certifying quality", not labeling things "good and bad". Let's be honest, when you label a set of projects "good", and the other has no label, the human mind will assign the label "not good" to the remainder thereof. Now, if you're a "good" project, all is well and fine. Now let's consider what happens if a project is in the uncertified (read: not good) set. Since we're trying to "improve quality", this is the really the set we were interested in to begin with anyway. It doesn't sound pretty, but like it or not, that's the truth.

Imagine you're the translator of a project which fails to meet certification standards. You have several options. First of all, you can earnestly strive hard to improve and work toward that mark of quality, an admirable attitude. However, if a translator had this type of personality to begin with, I fail to see in what way ~The Official B-T Certificate of Verifiable Awesomeness~ would further motivate them above and beyond their original desire to do their best. No, what they need are responsible mentors that are willing to dedicate the time and effort to help them succeed. A gold star by his name won't save a drowning man no matter how badly he wants to swim.

The next option is that you fail to meet certification standards and feel too disheartened and discouraged to continue. You have, after all, already been trying your hardest, and translating's a pretty time-consuming and thankless job to begin with. Now someone's coming in and telling you you're doing a crappy job to boot, would you kindly step it up a notch and not embarrass everyone? Still feel like pouring hours into it? Probably not. (This is likely the reason for the push-back this proposal's received so far.)

The final option is to not care, and continue along as before. The readers will continue to read your work either way, and be honestly grateful for it. And hey, wasn't that why you were doing this in the first place?

---

TL;DR: If you really want to improve quality and be a good human about it, either provide a framework not in which people can succeed, but in which they cannot fail. The alternative is to exclude those who do not meet standards. You can't have no incentives and no responsibility, and expect quality. You can't have your cake and eat it too unless you're willing to bear the responsibility for the entire process.
"There is always an easy solution to every problem - neat, plausible and wrong." H.L. Mencken (1971)
User avatar
cloudii
Project Translator
Posts: 512
Joined: Mon Mar 25, 2013 4:30 pm
Favourite Light Novel: Ahouka!
Location: awkward buttface
Contact:

Re: Improving BT Quality Discussion

Post by cloudii »

^Don't we technically have to make the diagnostic tests first? At least, before we have a project volunteer to go through with it?

^^Also, overhaul of supervisor role, AKA: better define the role of the supervisor, sounds good to me.

After coming back to this conversation after a while, I'm going to point out the big-picture ideas that I like the most so far. I'm also going to put out another idea I had for implementation:


1: Regarding Tracking Edits/TLC/status
  • I still share the same position as I did in my original proposal. Projects should have a hash marking next to each unedited raw chapter. Once a chapter is fully edited by a qualified individual, the flag should be removed.

    Futhermore, once the edit is performed, the editor should add their name to the Registration page which will contain a table of checks/names like Detalz mentioned.

    NEW: I'm thinking that full projects ("high quality projects") should have their main project pages protected so that only individuals of the translator/editor/supervisor user group (on Wiki) can edit. Currently, the project pages are protected so that only registered users can edit (not anons). This idea may sound radical, but hear me out. It'll come together by the end of this post. Note that this stricter page protection wouldn't apply to teaser projects or the actual chapter pages.

    Theoretically, this stricter main page protection would bring appeal to joining the official translator/editor user group for the purposes of this challenge.
2: Regarding "qualified" editors/translators
  • NEW: I've thought about this, and I realized the best implementation of recognizing "qualified" editors/translators is actually the existing translator/editor usergroup on the wiki. In my proposal, I called them "senpai's", but it's so much easier to take advantage of the existing user group. In fact, we can drop the senpai terminology completely (hence eliminate the inferiority complex concern). If the main project pages have editing status flags, and these pages are protected, only individuals of these usergroups would be able to update the main project pages.

    Meanwhile, no one would be restricted from editing chapter pages (the existing links), and new translators would not be restricted from adding new pages to the wiki. :D No one is hurt that badly in this kind of system.

    It's an implementation technicality.

    The main page protection thing is the practical appeal of joining this usergroup. If there's no real benefits gained from taking a diagnostic test, why would anyone take a diagnostic test?
3: Regarding Diagnostic Tests
  • This has been contraversial to many people, and I've thought about this a little bit. Here's my current stance on Diagnostic Tests, because I think they're absolutely necessary to get "productive edits" to make the tracking system meaningful.

    IMPORTANT: The Diagnostic Tests should be really really really easy. As a matter of fact, they should be multiple choice. Furthermore, I think it should be an automatically graded quiz so the feedback is instantaneous, and so we won't need to dedicated a team of people to review the tests. Once an individual passes the REALLY EASY diagnostic test, they get added to the respective user group.

    Why? Let's think. What are we trying to accomplish from implementing diagnostic tests?

    In my opinion, the purpose of this is to filter out individuals who blatantly can't read Japanese/Chinese (yet want to be translators, and would probably machine-translate if we didn't stop them), and wanna-be editors who will edit a translation yet subsequently make the grammar worse.

    These are people I think we can all agree we'd prefer to discourage from contributing too much on BT before they improve slightly.

    We don't want to deny people who want to improve their Japanese or English, or people who are really dedicated and want that usergroup recognition. This is why the diagnostic test should be really easy, AND multiple choice. If it's automatically graded, a very dedicated individual can continue retaking a randomized 20-question multiple choice diagnostic. If they're ambitious enough to retake it 4-5 times, then I think I'm willing to make the assumption that they actually care about improving, and even if they may be skilled enough right now, it's evidence that they want to improve.

    I also think these diagnostic tests should be educational. Meaning, by taking the diagnostic, someone who's never translated/edited before might learn about Baka-Tsuki etiquette. That's why we'd call it a "diagnostic", not an "exam". Diagnostics give meaningful and educational feedback. I'm going to quote NanoDesu's editor diagnostic, just to make a point:
    So, you want to be an NanoDesu Translation’s editor? It may not go as easy as you think. An editor is an important compliment to the translator, and their expertise in English grammar and composition should be top notch. However, the difficulties are much more than that. For one, as an editor, you must not only worry about making the requisite changes that makes a certain passage flow more readily, but must at the same time not edit too much as to change the meaning of a sentence. This is especially relevant to translation editing, as you do not have the source material in front of you, and even when you do, you most likely will not be able to read the source language. So, even after finding an obvious mistake in the translation draft, an editor’s proposed changed must be double checked, so that each editor is sure that they are not overstepping their boundaries. The key to good edited products are therefore precision and balance; not only must he have the keen eyesight to spot errors the translator missed, but you must also have the restraint necessary to prevent over-editing in such a way as to decrease the translation accuracy, and thereby risk doing a disservice to whomever reads the work. It’s not as easy as it sounds, it definitely takes quite a bit of dedication and practice.

    tldr; Basically, the diagnostic can also double as a way to make sure that individuals understand the rules, and make it clear that editors are expected to correct sentence structure for fluency and not just uselessly spot typos like spellcheck D:< (MS word can do that for me, T___T).

    Also, if the diagnostic tests double as informational tutorials, we could also apply it for project-supervisor usergroup. I'm envisioning the diagnostic tests to be similar to the quick and easy quizzes that many companies have for their new employees. Something simple.
4: Regarding the Committee Stamp of Approval
  • Currently, this appears to be the main focal point of conversation. As much as a possible, I'd like to avoid needing a CENTRAL BODY that will need to go around evaluating whether projects are good or not.

    I'd also prefer if evaluation of this stamp to be as quantitative (not qualitative) as possible.

    Quantitative, meaning something like, ex: "All of the translators participating in this project have passed the really-easy-diagnostic-test". Or, "All of the chapters have been edited once by a qualified individual". (<---hypothetical standards)

    Easy requirements like these don't require a CENTRAL BODY to make judgements. It's also clear to an editor/translator whether his/her project meets certification standards. There's nothing to be upset about because it's a black-white evaluation.

    You might argue you can't define a high-quality project in quantitative standards. But we're dealing with an issue of practicality here. Everyone agrees that we don't have the manpower for a central board making extensive qualitative judgments. However, for the sake of practicality, I'm willing to sacrifice a thorough TLC review for a simple quantitative (yet feasible) evaluation.

    If this "stamp of approval" is such a big issue, then I'd rather scrap the idea. The stamp is a publicity idea. It's meant to attract more individuals to care about quality, and attract more people to participate in the challenge (since the original idea of the Tsuki-chan challenge was optional anyways). The real game-changer in the quality is the edit-tracking and the diagnostic tests.

    If we can drive people to work with the edit-tracking AND participate in diagnostic tests, we'll see an improvement in Baka-Tsuki no matter what.
5: Regarding Staring New Projects
  • I still really like the stricter project start-up guidelines. The workflow would go like this.

    1). Someone starts a new teaser project. Practically anyone can do this.

    2). Teaser project gets a dedicated translator. Teaser Project is promoted to "Active Project" (and is listed on the complete project listing page). However, it is not added to the sidebar yet.

    3). "Active Project" completes one full volume in entirety. At this stage, it can be included on the sidebar as a "Full Project". I know I originally designed the Tsuki-chan challenge to be optional, but it'd be really nice if there was mandatory implementation at this stage (with the page protection I suggested above).

    Take this entire post as my re-digest of the entire Tsuki-chan project. Based on the things I mentioned here, and only the things I mentioned here, does it sound better or worse? I don't believe I'm excluding anyone here, and I don't think any part of this demotivates anyone. If this system is launched properly, it should be able to run by itself. Furthermore, there shouldn't need to be a dragging weight on responsibilities.
6: Miscellaneous Resources
  • In this thread, a lot of people have mentioned things about miscellaneous resources to improve editor/translator/etc relationships and guide new members on BT. They are all very worthwhile and considerate suggestions, and I thank you all for them.

    Most of them don't appear to conflict with my general suggestion at all, so I don't see a problem with implementing a majority of the suggestions (such as better Project Guideline pages, and Getting-Started Help Page, IRC, etc).
Twitter: @cloudiirain | BT Userpage | OreShura Translator | Biblia Editor (@HereticLNT) | Clockwork Editor (@HereticLNT)
User avatar
Rage_Ender
Sailor Tsuki
Posts: 1793
Joined: Mon Jan 21, 2013 9:59 pm
Favourite Light Novel: Mondaiji; Campione; Mushoku Tensei; Black Bullet
Location: ......

Re: Improving BT Quality Discussion

Post by Rage_Ender »

You know, If you want to try it out using this, you could always try it on a project with not much people contributing to it but have at least 1 volume finished. Or test it on a side project of someone who is a regular translator with experience.
dun dun dun..>!!!!!!<
User avatar
nikomaru
Kyonist
Posts: 11
Joined: Wed Nov 06, 2013 12:38 pm
Favourite Light Novel: Tasogare-iro no Uta Tsukai (Full Text)
Location: kediri
Contact:

Re: Improving BT Quality Discussion

Post by nikomaru »

in the future, is the any posibilities to remove content images from the top of fulltext page?
Image
well, to put it bluntly those image is spoiler hence reducing the surprise effect :(
:/xO rre qejyu m.t.y.y. anw daedu/.
User avatar
krytyk
Kyonite - The Haruhi Pacifier
Posts: 108
Joined: Fri Apr 22, 2011 8:51 am
Favourite Light Novel: Ahouka!
Location: Poland

Re: Improving BT Quality Discussion

Post by krytyk »

nikomaru wrote:in the future, is the any posibilities to remove content images from the top of fulltext page?
Image
well, to put it bluntly those image is spoiler hence reducing the surprise effect :(
Possibly, however it might impact Simon's epub generator and baka-tsuki mobile app, I'll try it and talk it over with folks later.

Edit: Currently testing on one of my own series:
http://www.baka-tsuki.org/project/index ... 2:Volume_1

I'll ask around and if it isn't conflicting with any of the 3rd party apps (already know from Simon it doesn't break BTE-GEN) so all that's left to confirm if it's fine with baka-reader. Anyone?
- JAP to ENG translator for AntiMagic Academy "The 35th Test Platoon"; ELYSION virtual region; Omae wo Onii-chan ni Shite Yarouka!?; Only Sense Online -
- Retired Manga Redrawer/Typesetter. @Manga-Heaven; @Kono-Basho; @wangan-scans; &anonymous releases.
Post Reply

Return to “Proposals and Suggestions”