Feature to draw more attention to posts about bad TLs in chapter threads. [Rejected]

Dex-chan lover
Joined
Jan 18, 2019
Messages
544
I don't like only noticing someone's valid complaint about mistranslations after I'm finished the chapter and am viewing the chapter thread where it's mixed with plot discussions.

For example, the "Report" function for forum posts could have an option which is: "Report Major Translation Criticism". That report would be forwarded to mods who know a little Japanese, at least enough to judge at a glance whether the report is credible and not bogus, and whether there are multiple egregious mistakes rather than just a single mistake or only inconsequential mistakes. There would also be an explanation about what is meant by major translation criticism, to make it clear that just because you don't, say, like the Frieren TL's style doesn't make it a mistranslation.

The result of an accepted report could be either "pinning" the criticism to the top of the thread, and/or placing some kind of flag/marker like a red asterisk next to the "55 Comments" button text on the right side in the chapter viewer, so people who want to stay away from bad TL can notice that and maybe even decide to take a quick look at the substance of the complaint.

Note, this should not be used to report self-reported machine translations, but some of those are fairly accurate. Not as good as manual TL, with tons of nuance lost, but a good editor and a little bit of Japanese knowledge or very simple grammar can mitigate any serious error. Only posts to the effect of "this is what it says in the raw here and here and here; it's completely different" should be reported.
 
Upvote 4
Dex-chan lover
Joined
May 18, 2019
Messages
3,747
That report would be forwarded to mods who know a little Japanese, at least enough to judge at a glance whether the report is credible and not bogus
That is a lot of work for something most people don't care about, much to my chagrin. Also, MD's stance is to not moderate translation quality.

Your best course of action is to wait if some kind soul points out the TL is complete dogshit in the comment section, then you can block the uploader/group.
 
Group Leader
Joined
Jan 8, 2023
Messages
386
to make it clear that just because you don't, say, like the Frieren TL's style doesn't make it a mistranslation.

so people who want to stay away from bad TL can notice that and maybe even decide to take a quick look at the substance of the complaint.

Only posts to the effect of "this is what it says in the raw here and here and here; it's completely different" should be reported.
I'm having difficulty understanding some of the points the OP made. I guess it is bad English. :meguuusad:
 
Last edited:
Dex-chan lover
Joined
Dec 25, 2019
Messages
2,151
Tags and reports wouldn't really work for reasons discussed above (tags would have to torpedo actual quality MTL works where the translator cleans up the machine's work to make it readable and of sufficient quality. Reports would create significant mod workload and would probably be backlogged in a hurry in addition to being subjectivity masquerading as something authoritative)

The only thing I can think of that might actually function somewhat properly but would still probably be fraught with problems would be some sort of user-led translation quality rating. Allow users to vote on translation quality independently from the series' content quality and display that. The problem is that almost any way to implement that would come with at least one significant hurdle.

I wish there was a way to handle this better because reading chapter comments often comes with spoilers and can be random as to whether or not you actually get translation judgement in any given chapter. Putting translation comments in the series' overall discussion works better but some series have hundreds of posts and pages of comments to sift through while others have almost none as nobody really translates on the whole series when they can just post comments to each chapter (I have at times used the series thread to provide a short review of a series I've read, but I don't even keep that up with regularity because I forget). And I'm like a couple of posters above in that I really wish people cared more about translation quality but there are a lot of works with total garbage-tier translation quality that get defended to death by people who think you should be happy to get anything that remotely resembles English even if it's a jumble of word salad with enough errors to burn through a whole pack of red pens. Unfortunately the expectation of at least some level of competence and pride in work is not considered important for a portion of the scnalation community or the reader-base.
 
Dex-chan lover
Joined
Jan 25, 2023
Messages
9,820
I remember a TL here "Shiina Complex", which was pulled out. Maybe that should be the baseline for awful TL.

Every single chapters would then marked by a red asterisk.

And the mods are not here to judge the quality of a translation...
Hmmm, it is free after all, ( which means WE ARE THE PRODUCTS ).

There is none. You can follow the whole shit grammar/shit translation debacle on e-hentai forums to see why it didn't really work out. In the end, it's just way too much work for something the majority of consoomers don't give a shit about.
I thought E-hentai is only about OOOHHHs and AHs and "NOOOOOO!!!" This is the reason I stopped reading. At least here we need legit plot-understandable English.
 
Dex-chan lover
Joined
Jan 18, 2019
Messages
544
@HelperHand Thanks for the link to those old threads; I already took a look at them before I started the thread (and so, I'm also not surprised people see problems with this feature now, saying similar things in this thread).

I agree that the issue of mod interest/workload mentioned by @ieatass69 ("a lot of work for something most people don't care about") @OisE ("mods are not here to ...") and @Darkshade13 ("Reports would create significant mod workload and would probably be backlogged in a hurry") is absolutely an issue with the feature. Honestly, I arrived at the idea of a report feature after I thought a little about potential alternate proposals and concluded that clear guidelines followed by mods would be a good way to prevent misuse of a feature that increases awareness of problematic TLs, and I think it's possible that the amount of workload may be less than people think, but nonetheless, it's true that any feature which involves increased moderation would need to seriously justify that.

A few other miscellaneous points I saw people made which I can provide counterpoints to, if anyone's still reading this thread. Spoilered because I don't particularly need further responses and I feel like these points detract from the main issues anyway.
  • "MD [mods] shouldn't be the judge of TL quality": 1. Making certain posts more visible to people, by pinning them to the top of a comment thread, isn't really a harsh judgment or condemnation. It won't be visible unless people are in the chapter's reader view or the comment thread itself. People can still take it or leave it. 2. If the implication here is that mods would need some kind of special Japanese skill to judge these reports: not really. If the guidelines for what qualifies as major translation criticism were properly formulated and understood, and the mods were used to the seeing these types of post, they'd be able to make quick decisions for the vast majority of cases without knowledge of Japanese or even looking at the raw. Naturally, if there was any doubt, the report could be rejected.
  • "Every chapter would be marked with a red asterisk": That wouldn't happen. Even changing the hyperbolic phrase "every chapter" to something less unrealistic like "10% of chapters", it still wouldn't happen. It's not impossible to distinguish between TL nitpicking and multiple egregious errors that show the translator seriously failed to grasp the meaning of the sentence. If people think this, it's possible they think so because they heard some JP->EN TL on Twitter say something like "TL is an art form and there are too many people who criticize non-literal translations" and formed some kind of mistaken "everything's relative" view about TL quality, but that's not true.
  • "A tag for this has already been suggested": To be clear, I think a "MTL tag" is a bad idea. At the core, this feature is just pinning a post in a comment thread, whereas a MTL tag would be subjective and dubious since there are plenty of MTLs that aren't self-reported, or self-reported MTLs that are edited with care or only deal with simple grammar and thus aren't actually problematic enough to be able to receive a "major translation criticism".


The only thing I can think of that might actually function somewhat properly but would still probably be fraught with problems would be some sort of user-led translation quality rating. Allow users to vote on translation quality independently from the series' content quality and display that. The problem is that almost any way to implement that would come with at least one significant hurdle.

I wish there was a way to handle this better because reading chapter comments often comes with spoilers and can be random as to whether or not you actually get translation judgement in any given chapter. Putting translation comments in the series' overall discussion works better but some series have hundreds of posts and pages of comments to sift through while others have almost none as nobody really translates on the whole series when they can just post comments to each chapter (I have at times used the series thread to provide a short review of a series I've read, but I don't even keep that up with regularity because I forget). And I'm like a couple of posters above in that I really wish people cared more about translation quality but there are a lot of works with total garbage-tier translation quality that get defended to death by people who think you should be happy to get anything that remotely resembles English even if it's a jumble of word salad with enough errors to burn through a whole pack of red pens. Unfortunately the expectation of at least some level of competence and pride in work is not considered important for a portion of the scnalation community or the reader-base.
Great points... While there's no way to perfectly note problematic TLs, my general feeling is that you'd gain a lot even if you only drew attention to the most blatant ones. And I feel like, as long as you're conservative in the way you design a feature, you should be able to design one that at least does that.

For example, with the voting system you just described, which is certainly worth thinking about since people don't like the idea of mods doing more work: if the voting system's core wasn't specifically about what counts as a serious TL problem, and was simply along the lines of "Does this post provide useful info about the TL?" and the red asterisk was changed to a more neutral green color, so the result of the vote was just "let's bump this post about TL quality to the top of the thread and add a green asterisk after '55 comments' in the reader view"... it's hard to imagine that system resulting in serious abuse, as long as the proper parameters were set for threshold required to be promoted like that, perhaps factoring in thread views and time since release. I think that even the people who "defend TLs to death" wouldn't necessarily have a reason to brigade/abuse the system (unless the TL criticism truly was subjective/nitpicking) because highlighting TL issues doesn't mean condemning the scanlation outright.

Let me also say: it's perfectly find to love a TL even if it's "garbage", because people will often take what they can get, and people [like me] who know Japanese have no right to say these TLs shouldn't exist. (And when I first learned Japanese yet tried to read stuff beyond my level, I wouldn't say my experience was necessarily much qualitatively different from a MTL reader's.) It's true that at the end of the day, many people enjoy these TLs, and no one is suggesting hiding them or censoring them. But if there are multiple TLs, it would help to have a chance to know in advance which one is garbage, and even if there's only 1 TL and its problems aren't easy to see, it would be nice to give people the chance to pay attention and judge for themselves if they want to accept its problems. That's my two cents.
 
Dex-chan lover
Joined
Dec 28, 2018
Messages
368
The problem isn't just that people are sniping or posting bad translation. It's that they almost exclusively do it to series that are popular, in an attempt to become the "main" group for a series and get eyes on their "donations to pay for manga to scan"coughbullshitcough crowdfunding pages.

However, trying to implement a quality control system in MD would require fully honest participation, almost completely subjective standards, and would be a fools errand. Better for the userbase to start adapting best practices like hopping to a title's main page for a second to confirm the group is the same one you've been following, and block groups that put out bad translation. Done for long enough, and eventually we'll have a better site once all the scammers are gone or ignored.

This can work. For example, almost at the very beginning of the ability to block groups, the troll group/user whose final chapter uploads were a big reason the functionality was put in place immediately almost completely stopped posting. They still try from time to time, but so many people blocked them that they can only catch newcomers following new series.
 
Dex-chan lover
Joined
Jan 18, 2019
Messages
544
jeezits, that's a thoughtful post. FYI, this thread's suggestion isn't a "quality control system" in any way, though.

What you're saying is that trying to stop or discourage people from sniping series and posting bad translations is difficult. Sure, and I agree with your reasoning, which even builds on experiences you've had that have informed you much more deeply than me on that issue. But trying to do that doesn't interest me anyway, and is unrelated to this suggestion.

...

What some of you guys have shown me in this thread is that the current discourse related to "translation problems" is polluted by controversy over things that aren't directly related to the modest feature being proposed here. It would be nice if one day people could take a more critical look at the idea of increasing visibility of commentary on translation problems.

I know from experience that I'm the only one on MD who likes to read forum posts that inform me about problems with TLs, even if I will read those TLs regardless. The only potential implementation problem is whether "moderation" could be burdensome, but even then, there are other ways to implement the feature (such as with a user vote system like I described 2 posts up).

That said, it's not trivial to code a feature like this either, and I'm sure that other features are a higher priority.
 
Last edited:
Contributor
Joined
Oct 1, 2020
Messages
385
1. More reports for mods to read = less time for them to fix actual problems. More development time spent on some kind of qc feature = less time for actually useful features/fixes. (Even if you dont call it a qc feature, its effectively a qc feature.)
2. You would need to moderate more than just JP to EN. There are 14 languages with five figure+ chapter uploads to this site. If you tried to accommodate JUST the biggest 14, There would need to be mods for AT LEAST every major source/target language pair, which is nowhere even close to practical/reasonable/necessary.
3. Drawing the line between simple mistakes and shit that ruins the reading experience will be a constant source of uncertainty and inconsistency. ESP if the mods doing the evaluations speak the langs but aren't translators with an understanding of what binary/non-binary errors other translators might make, and when exactly to draw the line, especially in a way that's fair and consistent across uploads, over time, between manga and translators. Not to mention theres potential conflict of interest issue with people who are active making scanlations while also doing moderation work.

The only thing I can think of that might actually function somewhat properly but would still probably be fraught with problems would be some sort of user-led translation quality rating. Allow users to vote on translation quality independently from the series' content quality and display that. The problem is that almost any way to implement that would come with at least one significant hurdle.
I agree with everything else in this post, but beyond the dev time setting this up, readers won't be able to speak the source languages (or else they'd be reading the source material) and will be in no position to evaluate the quality of the translation except as it's presented in English. Since the accuracy of the translation is the most important part of a translation, the vast majority of the people evaluating the quality of the product will not be qualified to evaluate it.

There is none. You can follow the whole shit grammar/shit translation debacle on e-hentai forums to see why it didn't really work out. In the end, it's just way too much work for something the majority of consoomers don't give a shit about.
^^^^

"MD [mods] shouldn't be the judge of TL quality": 1. Making certain posts more visible to people, by pinning them to the top of a comment thread, isn't really a harsh judgment or condemnation. It won't be visible unless people are in the chapter's reader view or the comment thread itself. People can still take it or leave it.
It may not seem harsh to you, but adding a big note in the comments or wherever from a mangadex mod calling out the people who provide the site it's content is a really bad idea from the site's perspective. MD goes out of its way to avoid becoming a player in scanlation disputes. Why would a scanlator allow their scans on a site that lets you opt out if the website itself criticizes you for things you as a scanlator might not consider to be a big deal. Mangadex needs the people putting their scans onto the site, the people who use the site do not need mangadex, this would just be self harm as a service.
2. If the implication here is that mods would need some kind of special Japanese skill to judge these reports: not really. If the guidelines for what qualifies as major translation criticism were properly formulated and understood, and the mods were used to the seeing these types of post, they'd be able to make quick decisions for the vast majority of cases without knowledge of Japanese or even looking at the raw. Naturally, if there was any doubt, the report could be rejected.
If you're going to criticize someone else's work you need to, at a minimum, know what you're talking about and have a leg to stand on, or at least look at the source material before you judge them. Truly absurd. Is this a joke?
"Every chapter would be marked with a red asterisk": That wouldn't happen. Even changing the hyperbolic phrase "every chapter" to something less unrealistic like "10% of chapters", it still wouldn't happen. It's not impossible to distinguish between TL nitpicking and multiple egregious errors that show the translator seriously failed to grasp the meaning of the sentence. If people think this, it's possible they think so because they heard some JP->EN TL on Twitter say something like "TL is an art form and there are too many people who criticize non-literal translations" and formed some kind of mistaken "everything's relative" view about TL quality, but that's not true.
What makes you think that? Where did you hear this from? There is a big difference between errors made because of a fundamental lack of language knowledge (binary errors), and suboptimal translations (non-binary errors). It sounds like you're after the binary errors, but you haven't articulated this clearly. If you are indeed after binary errors, it is even more important to be identifying them accurately, which is something only someone who speaks both languages can be trusted to do consistently!!!
How big of an error gets a chapter marked? or, a certain number of errors per chapter? It may seem like a small thing, but knowing how and where exactly to draw the line in these types of situations (i.e. how much of a fuckup is too big of a fuckup) is very important because if you're not suggesting a well defined set of standards that set of standards is completely worthless at best and realistically an active detriment to the site/readers. Look at how that turned out on e-h for porn and not shit with actual story people might care to dig deeper into. Insanity.
For example, with the voting system you just described, which is certainly worth thinking about since people don't like the idea of mods doing more work: if the voting system's core wasn't specifically about what counts as a serious TL problem, and was simply along the lines of "Does this post provide useful info about the TL?" and the red asterisk was changed to a more neutral green color, so the result of the vote was just "let's bump this post about TL quality to the top of the thread and add a green asterisk after '55 comments' in the reader view"... it's hard to imagine that system resulting in serious abuse, as long as the proper parameters were set for threshold required to be promoted like that, perhaps factoring in thread views and time since release. I think that even the people who "defend TLs to death" wouldn't necessarily have a reason to brigade/abuse the system (unless the TL criticism truly was subjective/nitpicking) because highlighting TL issues doesn't mean condemning the scanlation outright.
Most readers dont give a fuck, a lot of readers are ESL who wont even realize if the grammar is shit, let alone recognize a potential failure on the part of the actual translation itself. If you let readers decide how quality a translation is, the best you''ll get is a popularity contest in the comment section, and brigading and more useless drama at worse. Letting people like/dislike a chapter based on the quality of the scan in general would be less problematic, but still too stupid. How many people will use this feature? Enough to justify all of this garbage? No! Ratings will not be consistent across titles/translators/time. If you only let scanlators vote (i.e. members of a group, group leaders, contributors, ect) it'll just be the same issues as with the readers because literally anyone can be a scanlator. If you have the site/mods select certain scanlators/translators to vote then that would be a worst of both worlds scenario since it's still the site selecting people, AND there will be a limited pool of people/resources able to make evaluations, just like if the mods did the evaluations.

jeezits, that's a thoughtful post. FYI, this thread's suggestion isn't a "quality control system" in any way, though.
even if you dont call it a quality control system, thats literally what it is, and it's how people will see it. https://www.dictionary.com/browse/quality-control

What some of you guys have shown me in this thread is that the current discourse related to "translation problems" is polluted by controversy over things that aren't directly related to the modest feature being proposed here. It would be nice if one day people could take a more critical look at the idea of increasing visibility of commentary on translation problems.

I know from experience that I'm the only one on MD who likes to read forum posts that inform me about problems with TLs, even if I will read those TLs regardless. The only potential implementation problem is whether "moderation" could be burdensome, but even then, there are other ways to implement the feature (such as with a user vote system like I described 2 posts up).

That said, it's not trivial to code a feature like this either, and I'm sure that other features are a higher priority.
it would be nice if you understood the practical limitations of what you're suggesting. We still dont even have chapter/title views working. No author rework, or any number of other actually useful features, and even then, a system like what you're suggesting would still not be worth pursuing for the other reasons I've described above. "not trivial" and "higher priority" are understatements, since this is always going to be a terrible return on investment even if dev time is infinite.
And what's more, the MD people have already gone on record multiple times to say that they will not be doing any quality control of scanlators, so no matter what exactly you suggest, its not even gonna get taken serious.
 
Dex-chan lover
Joined
Jan 18, 2019
Messages
544
it would be nice if you understood the practical limitations of what you're suggesting. We still dont even have chapter/title views working. No author rework, or any number of other actually useful features, and even then, a system like what you're suggesting would still not be worth pursuing for the other reasons I've described above. "not trivial" and "higher priority" are understatements, since this is always going to be a terrible return on investment even if dev time is infinite.
And what's more, the MD people have already gone on record multiple times to say that they will not be doing any quality control of scanlators, so no matter what exactly you suggest, its not even gonna get taken serious.
Yeah, as you quoted, in the post above you, I said "it's not trivial to code a feature like this either, and I'm sure that other features are a higher priority" so I'm well aware that not all features are created equal. I also, for example, expect to appreciate the redone MDLists (one day) way more than I would have this feature.

I think that you and I both understand that the main thesis of your post is stated in your first paragraph, and restated in this last paragraph: the simple phrase "actually useful features". You don't think this feature is useful, and that's your freedom. Given that, I'm not sure that it makes any difference if I share my thoughts on the other reasons you listed that made you judge that this feature is a bad idea. Probably not, but I'll do it anyway.

1. More reports for mods to read = less time for them to fix actual problems.
You're pretty much just restating your main thesis with the word "actual problems" here, but yeah, I've said that "mod interest/workload ... is absolutely an issue with the feature". Darkshade13 and I even discussed alternate implementation styles that would require less moderation involvement. But in the first place, if you say that mods shouldn't spend time on it, you shouldn't even want devs to spend time to create the feature. And indeed, you don't.
2. You would need to moderate more than just JP to EN. There are 14 languages with five figure+ chapter uploads to this site.
I wasn't that the MangaDex constitution banned features that didn't apply to every language as discriminatory. Is it not worthwhile to have a feature that may be used in a large % of chapters? I'm in no way ruling out other languages, by the way, I just want to highlight what you're saying for argument's sake.
3. Drawing the line between simple mistakes and shit that ruins the reading experience will be a constant source of uncertainty and inconsistency.
(I'll quote this now just to say that I plan to respond to all your "TL issues are a clusterfuck to judge" lines at once further down in my post here, since you mention this multiple times.)

It may not seem harsh to you, but adding a big note in the comments or wherever from a mangadex mod calling out the people who provide the site it's content is a really bad idea from the site's perspective. MD goes out of its way to avoid becoming a player in scanlation disputes. Why would a scanlator allow their scans on a site that lets you opt out if the website itself criticizes you for things you as a scanlator might not consider to be a big deal. Mangadex needs the people putting their scans onto the site, the people who use the site do not need mangadex, this would just be self harm as a service.
As I said before, it's clear to me now that the discourse on "translation problems" is polluted. So I don't know if there's a point for me to explain this in detail.

Try to envision this: this feature is implemented, and if a scanlation group does not like the idea of commentary on their translation being pinned to the top of their thread, they have full veto power and can just instantly un-pin it. If you're thinking, "But how would this feature even be useful if bad scanlators can stop people from criticizing them??" then, it's just that you are conceptualizing a war of egos between scanlators and criticism, when the person who came up with this feature (me) was conceptualizing scanlators who maybe don't know they made mistakes and don't mind the fact others point them out, or may even be appreciative. I know I've been thanked, many times, and have thanked others. The kind of translator who is proud of their work and who doesn't accept criticism of their TL choices is the kind of translator who is actually paying enough attention to not make massive mistakes, or only make them incredibly rarely. You could even throw in a rule saying that outright insults to the translator or scanlation group would disqualify a post for being pinned, if you're so scared of any controversy.

If you're going to criticize someone else's work you need to, at a minimum, know what you're talking about and have a leg to stand on, or at least look at the source material before you judge them. Truly absurd. Is this a joke?
What makes you think that? Where did you hear this from? There is a big difference between errors made because of a fundamental lack of language knowledge (binary errors), and suboptimal translations (non-binary errors).
Most readers dont give a fuck, a lot of readers are ESL who wont even realize if the grammar is shit, let alone recognize a potential failure on the part of the actual translation itself.
In a world of ego wars where people are scheming to fake knowing the source language just to discredit people, or where overconfident idiots are the kind of people who take the time to point out TL issues, it would truly be "absurd". I probably can't change what you think you've experienced, I can only say that basically 100% of posts I've seen that went unchallenged and highlighted significant TL Issues were not unfounded. It would indeed be absurd to think that my proposed feature expects that people personally TL check every post about TL issues before they can make any kind of judgment on whether it seems correct. By all means, show me all the posts on MD which seem to highlight significant TL issues but it's actually founded upon the poster's bad Japanese/Chinese/etc. And yes, I'm aware that people often nitpick over phrases that are non-literal and translate based on intent, and I've made clear that is not what "significant TL issues" refers to. And if you think that people can't even judge when someone is nitpicking, you're vastly underestimating the extent to which people have been cultured to detect this.

even if you dont call it a quality control system, thats literally what it is, and it's how people will see it. https://www.dictionary.com/browse/quality-control
So quality control consists of just putting a note on the product's website saying "a customer reported a problem", huh? OK!

I'm surprised you made such a long post in a thread about a feature that fundamentally doesn't interest you. However, I won't get my hopes up that you have an interest in continuing this discussion further.
 
Contributor
Joined
Oct 1, 2020
Messages
385
So quality control consists of just putting a note on the product's website saying "a customer reported a problem", huh? OK!

I'm surprised you made such a long post in a thread about a feature that fundamentally doesn't interest you. However, I won't get my hopes up that you have an interest in continuing this discussion further.
there is a difference tho. When I downvote a youtube video, that's my opinion. If youtube the site "downvotes" a video (demonitizes/tanks search hits ect) that is the service itself endorsing the viewpoint that vid is bad. In this case the product/service youtube makes is the website itself, not the youtube video, so it is effectively a method of quality control on youtube. When youtube demonitizes a video it's because they're trying to make their service better, not the video. Same logic applies to MD since the service is the website itself, not the content hosted on it, just replace youtube with MD and demonetize with applying a red asterisk to the chapter, pinning a critical comment, or other things suggested in this thread. I am very confident at least a sizable portion of regular readers would see a pinned comment or red asterisk in the same way as folks see demonetization on youtube.

In a world of ego wars where people are scheming to fake knowing the source language just to discredit people, or where overconfident idiots are the kind of people who take the time to point out TL issues, it would truly be "absurd". I probably can't change what you think you've experienced, I can only say that basically 100% of posts I've seen that went unchallenged and highlighted significant TL Issues were not unfounded. It would indeed be absurd to think that my proposed feature expects that people personally TL check every post about TL issues before they can make any kind of judgment on whether it seems correct. By all means, show me all the posts on MD which seem to highlight significant TL issues but it's actually founded upon the poster's bad Japanese/Chinese/etc. And yes, I'm aware that people often nitpick over phrases that are non-literal and translate based on intent, and I've made clear that is not what "significant TL issues" refers to. And if you think that people can't even judge when someone is nitpicking, you're vastly underestimating the extent to which people have been cultured to detect this.
the vast majority of people who read a chapter on md do not leave comments, it would not surprise me if they also didn't even bother looking in the comment section at all, but i dont have data on that. For every person that upvoted or commented on the reddit post you linked, my suspicion is that there are dozens if not hundreds who didnt notice or dont care. keep in mind this is the internet, there are a hundred thousand idiots liking ai generated images on facebook this very second. Most people want to eat their grass in blissful ignorance, if i have to confront them and get into an argument about it, i'm not gonna waste my time telling them how far the rabbit hole goes because they dont want to hear it no matter how i feel about things.

3. Drawing the line between simple mistakes and shit that ruins the reading experience will be a constant source of uncertainty and inconsistency.
(I'll quote this now just to say that I plan to respond to all your "TL issues are a clusterfuck to judge" lines at once further down in my post here, since you mention this multiple times.)

the point i was trying to make wasn't even that it's impossible to determine, just that it will be prohibitively difficult to justify as a site policy.

I wasn't that the MangaDex constitution banned features that didn't apply to every language as discriminatory. Is it not worthwhile to have a feature that may be used in a large % of chapters? I'm in no way ruling out other languages, by the way, I just want to highlight what you're saying for argument's sake.
thing is as time goes by the devs have been focusing more of their efforts into making the site accessible to other languages so that MD can occupy a similar position that it occupies for the english fanTL market. If they stopped focusing on implementing features like changing the interface language, and instead focused on efforts that would go in the opposite direction it would be doubly contrary to their goal of language inclusivity.

I think that you and I both understand that the main thesis of your post is stated in your first paragraph, and restated in this last paragraph: the simple phrase "actually useful features". You don't think this feature is useful, and that's your freedom. Given that, I'm not sure that it makes any difference if I share my thoughts on the other reasons you listed that made you judge that this feature is a bad idea. Probably not, but I'll do it anyway.
I wish there was some viable, fair way to evaluate scans on quality without compromises/drawbacks but i'm not convinced there is, so I recognize that it's not useful. The more alternatives i consider the clearer it is to me that the only solution to bad scans is educating scanlators, and not educating the average reader, because the average scanlator is going to be putting some sort of effort into what they do, even if it's very little effort, and the reader is just reading. There are way fewer scanlators than readers, so if we educate scanlators better they'll be more impact for less effort. It's way easy to contact a guy uploading solo scans, than it is to create consensus in the minds of the public.
The best idea I liked for marking MTL'd chapters would be a completely optional check box on the upload page to allow the group making the upload to mark their post as MTL and have a little tag show up somewhere just so folks know. Even this totally voluntary self reporting was not acceptable to the MD people, so I can say with high confidence that more complex ideas around quality control do not stand an ice cube's chance in hell of getting implemented, for better or worse.
 
Dex-chan lover
Joined
Jan 18, 2019
Messages
544
there is a difference tho. When I downvote a youtube video, that's my opinion. If youtube the site "downvotes" a video (demonitizes/tanks search hits ect) that is the service itself endorsing the viewpoint that vid is bad. In this case the product/service youtube makes is the website itself, not the youtube video, so it is effectively a method of quality control on youtube. When youtube demonitizes a video it's because they're trying to make their service better, not the video. Same logic applies to MD since the service is the website itself, not the content hosted on it, just replace youtube with MD and demonetize with applying a red asterisk to the chapter, pinning a critical comment, or other things suggested in this thread. I am very confident at least a sizable portion of regular readers would see a pinned comment or red asterisk in the same way as folks see demonetization on youtube.
YouTube demonetizing videos is a major consequence that enforces behavior on videos and so it does affect their content, albeit I'm still not sure quality control is the term you're looking for. Making a reader's comment/reply more visible/noticeable is a basic feature that's far from quality control. The only statement that such a feature implies is that MD respects the desire of readers to more readily learn about and communicate to one another TL mistakes in the chapters they read. You say you're "very confident" that people would see these two features (YouTube demonetizing videos, and this thread's feature) the same way. I'm not, and fail to even see any meaningful similarities between them. Instead of picking YouTube's demonitizing, you would have been better off trying to compare this feature to something like X's community notes (which people love), though there are still major differences.

the vast majority of people who read a chapter on md do not leave comments, it would not surprise me if they also didn't even bother looking in the comment section at all, but i dont have data on that. For every person that upvoted or commented on the reddit post you linked, my suspicion is that there are dozens if not hundreds who didnt notice or dont care. keep in mind this is the internet, there are a hundred thousand idiots liking ai generated images on facebook this very second. Most people want to eat their grass in blissful ignorance, if i have to confront them and get into an argument about it, i'm not gonna waste my time telling them how far the rabbit hole goes because they dont want to hear it no matter how i feel about things.
You're replying to my key point explaining why there is nothing fundamentally impractical about mods/people being allowed to judge/vote to pin a post about TL issues. And what you're saying is that aside from the people who upvote on Reddit, there are countless more people who are quiet and ignorant, and they exist on MD too. I agree, but how does that matter? This feature wouldn't rely on universal civic duty voting. Maybe I've missed your point.

the point i was trying to make wasn't even that it's impossible to determine, just that it will be prohibitively difficult to justify as a site policy.
I never said that you said it was impossible. The phrase "clusterfuck to judge" there doesn't mean outright impossible, it basically means just what you said, prohibitively difficult.

thing is as time goes by the devs have been focusing more of their efforts into making the site accessible to other languages so that MD can occupy a similar position that it occupies for the english fanTL market. If they stopped focusing on implementing features like changing the interface language, and instead focused on efforts that would go in the opposite direction it would be doubly contrary to their goal of language inclusivity.
I'm not actually opposed to wanting more inclusivity. My later statements (the "In a world of ego wars" paragraph) explained why I don't believe it's necessary for people to know the source language before they can promote/upvote/moderate posts about TL issues, so it doesn't matter.

I wish there was some viable, fair way to evaluate scans on quality without compromises/drawbacks but i'm not convinced there is, so I recognize that it's not useful. The more alternatives i consider the clearer it is to me that the only solution to bad scans is educating scanlators, and not educating the average reader, because the average scanlator is going to be putting some sort of effort into what they do, even if it's very little effort, and the reader is just reading. There are way fewer scanlators than readers, so if we educate scanlators better they'll be more impact for less effort. It's way easy to contact a guy uploading solo scans, than it is to create consensus in the minds of the public.
The best idea I liked for marking MTL'd chapters would be a completely optional check box on the upload page to allow the group making the upload to mark their post as MTL and have a little tag show up somewhere just so folks know. Even this totally voluntary self reporting was not acceptable to the MD people, so I can say with high confidence that more complex ideas around quality control do not stand an ice cube's chance in hell of getting implemented, for better or worse.
Just for the record, this feature doesn't involve evaluating anything as a whole, and it doesn't involve commenting on scans in general, just their translations. Educating scanlators isn't mutually exclusive with this feature, but is always welcome for those who wish to learn; I remember the days when MangaHelpers was popular. My experience has been that many readers also appreciate knowing about major TL mistakes in a work they enjoy. As for the MTL self-reporting idea, I guess in your opinion it's at least better than this thread's idea which you believe involves MD bullying scanlators. It does seem feasible, and some people who hate all MTL would appreciate the chance to reduce their exposure to it, and it would help people make quick decisions between chapters with multiple scanlation choices in cases where one is MTL, though you'd have to weigh that against the controversy that comes from the fact many people will refuse to report it because they have pride in their MTL. I still consider this thread's feature more useful since it gives people the opportunity to have a concrete sense of what a chapter's TL issues may be, and even distinguish between high quality MTL and low quality MTL.
 
Contributor
Joined
Oct 1, 2020
Messages
385
YouTube demonetizing videos is a major consequence that enforces behavior on videos and so it does affect their content, albeit I'm still not sure quality control is the term you're looking for. Making a reader's comment/reply more visible/noticeable is a basic feature that's far from quality control.
Regardless of if you personally consider it quality control, mangadex sees taking a stance on the quality of a chapter as quality control. I have heard more or less that exact phrasing from the horses mouth multiple times consistently through the years. You may not realize it, but if mangdex itself weighed in on the quality of a chapter that would be a very major consequence for the chapter/group/scanlator akin to being demonetized on youtube or otherwise censured on other social media sites.
You say you're "very confident" that people would see these two features (YouTube demonetizing videos, and this thread's feature) the same way. I'm not, and fail to even see any meaningful similarities between them.
If you can't understand or accept the obvious similarities between your suggested system, which effectively censures user created content, and the example I provided (youtube demonetization,) which ALSO censures user created content, you are out of touch with the impression this gives most mangadex users and the scanlators which provide the site content.
which, incidentally is why you missed this point;
You're replying to my key point explaining why there is nothing fundamentally impractical about mods/people being allowed to judge/vote to pin a post about TL issues. And what you're saying is that aside from the people who upvote on Reddit, there are countless more people who are quiet and ignorant, and they exist on MD too. I agree, but how does that matter? This feature wouldn't rely on universal civic duty voting. Maybe I've missed your point.
If practically no one (on the scale of a manga hosting website) gives a shit about the feature, then developing it and maintaining it is not worthwhile, esp considering the numerous other concerns I've touched on in previous posts. The mangadex developers are not making their site for the 1% of people who want to prattle on in the comment section about a mistranslated sentence, they're making the site for the average readers and the scanlators who post their content to the site, and they would be at best alienating their target market, if not actively subverting the people the site relies upon for content, which is it's whole reason for existing. Your stated position fundamentally misunderstands the purpose and goals of the website you want to change, while ignoring the practicalities of the issue in favor of your own convenience.
I'm not actually opposed to wanting more inclusivity. My later statements (the "In a world of ego wars" paragraph) explained why I don't believe it's necessary for people to know the source language before they can promote/upvote/moderate posts about TL issues, so it doesn't matter.
You seem to have made no meaningful consideration for potential abuses, and the inevitable interpersonal drama and needless loss of trust between the site and the scanlators that use it. There are already oodles of groups sniping series and involved in drama wars, mangadex the site does not need to waste resources to become a participant in any way with those pointless disputes! There are so many downsides to your suggestion, and the upside is "it will make it easier for me to know about TL errors before i read the chapter + read the comment section, assuming the system works correctly in the first place" and as a result the proposition will not be worthwhile for the site's owners!
I'm not saying wanting some method of quality control is wrong, I have thing's i'd like too, as i mentioned in my previous post. What I AM saying is that advocating for changes when they're at odds with the core principals of the service trying to be changed is just irresponsible. And so, I dont advocate for the changes I wanted, because I recognize there's no point advocating for them since they wont be implemented, and even if implemented could probably do more harm than good. The same applies with your suggestion regardless of if you acknowledge the issues. Nothing is certain in life, but I can practically guarantee the people who make the site aren't gonna do what you're asking them to do, so theres no point continuing to advocate for it as things stand.
 
Dex-chan lover
Joined
Jan 18, 2019
Messages
544
Regardless of if you personally consider it quality control, mangadex sees taking a stance on the quality of a chapter as quality control. I have heard more or less that exact phrasing from the horses mouth multiple times consistently through the years. You may not realize it, but if mangdex itself weighed in on the quality of a chapter that would be a very major consequence for the chapter/group/scanlator akin to being demonetized on youtube or otherwise censured on other social media sites.
It's not "taking a stance", it's facilitating the visibility of a comment. And it's not "the quality of a chapter", it's one aspect of the translation of a chapter. I would assume that the features that the horse's mouth you mentioned (MD staff) were referring to when they said "quality control" were, indeed, quality control, like the threads HelperHand linked to. This feature is completely different because it's just improving the visibility of a post about TL issues in a chapter's translation, and I've already outlined multiple ways to mitigate any theoretical negative impact. You might as well say that MD is "taking a stance" because they allow forum comments. You're characterizing this feature as something far more dramatic than it actually is.
If you can't understand or accept the obvious similarities between your suggested system, which effectively censures user created content, and the example I provided (youtube demonetization,) which ALSO censures user created content, you are out of touch with the impression this gives most mangadex users and the scanlators which provide the site content.
"effectively censures" isn't close to accurate. you're using the word "effectively" as a translation for "in my imagined scenario, through a sequence of imagined events". But I've already deconstructed that multiple times. In no way does this feature insult or judge scanlators. In no way does it force them to do anything. In no way does it even holistically comment on their chapter's quality. It facilitates reader awareness of translation issues, and those issues could be ignored or de-pinned at the leisure of the scanlator.

Since you brought up the "impression this gives" and said that you think I'm "out of touch with" that, I wonder if we can try to objectively consider who might be more out of touch with how people react to posts about TL issues. I know a common source language (JP) and have posted/commented on JP->EN TL issues on MD, but I'm not sure if that is true of you. Or at least, have you been active in many MD threads where people point out/discussed translation errors? If so, I would have expected you to realize it's not a taboo to casually discuss this subject, and I've never seen a scanlator who dropped a series because they couldn't handle the fact someone non-rudely pointed out their mistake in a MD thread. There are already countless posts in countless forum threads which point out a TL issue without any kind of scanlator drama, and as I said before, translators are more often thankful than offended.
which, incidentally is why you missed this point;
If practically no one (on the scale of a manga hosting website) gives a shit about the feature,
I made a key point explaining (to counter your criticisms) why there is nothing fundamentally impractical about mods/people being allowed to judge/vote to pin a post about TL issues. In response, you said that "not everyone upvotes". I said, "I agree, so what? Did I miss your point?" Now you're telling me, "your feature would be unpopular". I guess my response now would be to just say... okay?
You seem to have made no meaningful consideration for potential abuses, and the inevitable interpersonal drama and needless loss of trust between the site and the scanlators that use it. There are already oodles of groups sniping series and involved in drama wars, mangadex the site does not need to waste resources to become a participant in any way with those pointless disputes! There are so many downsides to your suggestion, and the upside is "it will make it easier for me to know about TL errors before i read the chapter + read the comment section, assuming the system works correctly in the first place" and as a result the proposition will not be worthwhile for the site's owners!
I'm not saying wanting some method of quality control is wrong, I have thing's i'd like too, as i mentioned in my previous post. What I AM saying is that advocating for changes when they're at odds with the core principals of the service trying to be changed is just irresponsible. And so, I dont advocate for the changes I wanted, because I recognize there's no point advocating for them since they wont be implemented, and even if implemented could probably do more harm than good. The same applies with your suggestion regardless of if you acknowledge the issues. Nothing is certain in life, but I can practically guarantee the people who make the site aren't gonna do what you're asking them to do, so theres no point continuing to advocate for it as things stand.
I've made considerations for things like excluding posts which contain insults, or giving scanlators veto power. But you included the word "meaningful" here as a qualifier, so I suppose that's just not meaningful to you. I'm aware of the reality of scanlation drama and competition and how TL issues can become debate fodder. Virtually all comments on MD about TL issues are not born from scanlator drama, but rather people who notice a TL issue and casually decide to point it out, and it leads to no drama whatsoever. As for whether there's a "point" that I "advocate" for a feature I proposed: the existence of this subforum means that MD staff aren't afraid to let people talk about what kind of features they like, and I saw no reason not to respond to your criticisms. Of course, whether any feature is implemented depends on MD developer interest and to an extent whether users have an interest as well, and I don't have expectations this feature will be implemented in the near future, but it's because I love the site and think that this would be a nice feature that I don't hesitate to share my views.
 

Users who are viewing this thread

Top