I present to you the greatest manga from the soon-to-be-shut-down Manga Library Z. No other Japanese comic has such an exquisite use of language, which is presented in its original untranslated form, and plot twists quite so unforeseen.https://www.mangaz.com/book/detail/140401
>>272899176Yep, this is great.
Is this yuri?
>>272899290It's yuriyurishii.
Bump
Amen.
Have you ever been challenged by Christians?
Namu Amida Butsu.
>>272899260Is the bubble order right to left even though the text is left to right?
Yuri can shift the time.
>>272899576I think so...? But who can truly say.
And thus ends the yuri-yuri-shiest manga to ever grace this godforsaken planet, the second coming of which would surely save all of us from the circulation of suffering.
>>272899303Long neck is long.
>>272899764thank you for the dump op, i definitely cant wait see what else this author has stored for us
>>272899526This page is funny
nani the fuck
>>272899526Flat is justice, do not lose to the Christians cultural colonization!
>>272899438That took a weird turn
>>272899176Double the yuri!?
>>272899764Thank you for the dump, this is the funniest thing I've read in a long time
>>272900091Unfortunately, it looks like they haven't been drawing anything for the better part of a decade.https://x.com/nobiematsuura
Someone repost the manga-library Z scripts some anon wrote, we have to back it up or a lot of these will become lost media
>>272872112
Also working to archive some of this stuff, hopefully script anon gets around to implementing that wrapper, I've got a couple home servers I can dedicate to running the script for a few days. Unfortunately I'm not made of storage, so a couple terabytes is all I can offer
>>272904625There you go, mzrip v4 archive with the wrapper, CSV made by anon in the last thread included in the archive already. If you haven't already, install NodeJS. Run "npm install" in the script directory in any case for good measure.https://drive.google.com/file/d/1D6dGBcukWcatRaEvYk3MDXrPMxJRfkmI/view?usp=sharingTo download all series listed in the attached CSV from lines `from` to `to` (which default to 1 and line number respectively):node mzrip_csv.js series.csv --from=1000 --to=1999 --threads=10 --out=DownloadsI have, emphatically, not debugged this well, so it will crash and burn sometimes, and ChatGPT was used to bootstrap much of this code to save time and headaches. I'm legitimately surprised at how well it can write basic code these days, though it has outdated library definitions and JS libraries like to break compatibility.
If there are more anons, with this script you can also in theory coordinate which line number range you're saving. But seems like there's just one anon. And to be honest I have no idea how we could merge the libraries afterwards, given the sheer size of them.
>>272906446Important - forgot to mention one more thing to do before you start ripping. The image library I'm using to load and edit images has a nasty issue, and it sometimes fails to load very big images (think 8k). This is fixable in a hacky way.After you run "npm install", go to file "node_modules\jpeg-js\lib\decoder.js", find the line "maxMemoryUsageInMB: 512" (in the current version line 1106) and change that to "maxMemoryUsageInMB: 2048". Unfortunately, the Jimp library is not overriding this jpeg-js parameter correctly, so you have to change it in the jpeg-js code itself.Example where this is a problem: book ID 187162, すごい人になりたい恵子ちゃんと邪魔する初美ちゃん (the example image is so large I had to 50% it to upload here).
>>272906531I genuinely don't know what's happening and I don't have any hard drives anyway, I'm dirt poor and can barely afford food
Damn, I was testing this on a very fast SSD, and when running it on my external HDD that actually has some free space it's noticeably slower. Maybe when using oxipng I should also use a temporary directory for downloading and optimizing that can be on a faster drive... but to be honest I'm getting a headache from scripting already.
>>272906833>>272906446Great stuff anon, thank you very much.Grabbing the first 100 IDs then I'll get it set up on my spare machines.Just a heads up to anybody else that may be archiving, I noticed earlier but forgot to mention; some of the IDs I scraped earlier aren't valid, every instance of this appears to be -1 in the id number (e.g. 腕子さん has the ID of 191271 in the spreadsheet but 191272 on site)Not terribly important as I'll pipe the terminal output to a log file and check it later to pick up anything it missed manually.
I think the real issue is there's no way to know what's already been archivedSomeone needs to set up a google sheet or something to organize this unless some hero archives the whole thing
>>272907568My script should be able to handle that. This is because series IDs are not always the same as the IDs of the first or the only volume. But the script goes to the series page and gets the volume IDs from there. Sometimes a series page redirects directly to /book/details/some_id, the script should handle that gracefully too. Though maybe there are some cases I'm not yet aware of.Also, books 9901/9902/9903 are not real, so they won't be downloaded. These are fake covers that are displayed in the mobile app that basically say "mobile app is dead, go to the website", with no actual book content inside.
>>272907805Extremely based.>>272907605Definitely, if more people get interested in archiving this stuff.
>>272907568I have just realized that I haven't added any retry logic, I now have (5 retries with some incremental back-off). Please download again and run "npm install" to install the newly required libraries... I haven't debugged the retry part of it yet, I'll keep an eye on the logshttps://drive.google.com/file/d/1D6dGBcukWcatRaEvYk3MDXrPMxJRfkmI/view?usp=sharing
A lot of this stuff really is kinda garbage though, got to admit.
>>272907568By the way, I'm looking at the search page and it gives me "作品一覧：6674件". Could it be, by any chance, that your CSV doesn't have the R18 stuff maybe?
>>272908867If it's paid, no.
>>272908983Do include those as well, because the script can actually get all of the paid stuff. They have not hidden the files in the API, the web viewer itself limits you to the first view pages if you're not premium.
>>272899075this looks like a terrible deviantart comic
>>272899075Kawaiikochans spin-off?
>>272899388Confess now, anon.
Are there currently any other efforts to back up the entire site? If it's backed up, where will be posted and shared? A lot of the manga are not available anywhere else, so they'll become lost media if they aren't posted elsewhere.
I've updated the script to handle double redirects better, to deal with r18 series in a more stable fashion.https://drive.google.com/file/d/1D6dGBcukWcatRaEvYk3MDXrPMxJRfkmI/view?usp=sharingI've also added series_2.csv to the archive, which contain ~2k of the series IDs missing from series.csv but present in search - many of them were also not r18. This still doesn't cover unlisted titles, but first things first...
>>272899438God told us to be fruitful and multiply, but then denied so many women their most important 'fruits'. What did he mean by that?
>>272910315Supposedly here as well, this is some web archival group that specializes on this sort of stuff: irc://irc.hackint.org/#mangoesHowever, I haven't seen any discussion yet and the chat archive website is, ironically, down.
How much has been archived so far, and how is it currently planned to be distributed or handled? Because if it's just gonna be downloaded to hard drives without any form of secondary backup or showing, then they could be lost if the drives go bad.
>>272899487>>272899526This gag is actually really good.
>>272910840That's the tricky part, because nobody here has enough hard drive space to actually save a lot. If I had, I would've made a torrent out of it, but it's doubtful that my free ~3TB will amount to much.Sucks ass big time that we can't download source files, which would've made the whole process faster and take up much less space.
>>272910937No one has made any progress in trying to get source files? There's 22 days left.
Not getting the source images is not ideal, but honestly, if the alternative is no back up at all, as long as the pictures are of an acceptable resolution, I don't think it would be a major issue.
What I meant to say is, the important thing is we have stuff backed up at all.
>>272911011My only lead was decompiling the old and delisted Android app, as it most likely accessed binary encrypted source files, but it didn't amount to anything useful as all I got from the mobile API was the same result as on the web right now (i.e. only scrambled, not encryped images). Feel free to try, my meagre skills are at the limit already. I really wish one of the experts from the scanlation threads took a shot at this, they helped me with rippers before.I'm not sure if anyone else is even trying to get the source files at the moment.
>>272911138I really wish I could help out, but I just don't have the tech knowledge, so the best I can do is spread the word and try to get people onboard. Have you tried asking the Scanlation Thread or other places like IRC for help?
>>272911070The resolution is the same, but there is an extra layer of JPEG compression caused by the scramble, and the de-scrambled files are saved as bloated PNGs. The visual difference should be very, very minimal, though since we don't have the pre-scramble source files we can't tell for sure.
>>272911227I'm not a techie, is there a reason why they need to be turned into PNGs instead of just JPGs?
>>272911170There was no scanlation thread yet after the shutdown news have dropped, and I'm not sure which IRC could I even go to. I did intend to bug people in the thread once it appears, which might happen tomorrow if its schedule is the same as usual.
>>272911300Try making a scanlation thread and put the Manga Library Z closure in the OP, maybe that would raise interest. It's already Friday in most of the world so it's the usual day to make it even if it's not the usual time.>IRCTry ArchiveTeam or Mango. Apparently they're already working on it, but it would be good to do follow ups and the like.
>>272911267To avoid extra JPEG compression basically. The source image has been compressed to JPEG once, scrambled image added another JPEG compression, de-scrambled image would have added another if it were saved as JPEG. And each time you compress something to JPEG, you create more and more visual artifacts, as there is no such thing as lossless JPEG. So I'm saving the de-scrambled files as PNG to avoid that last extra compression.Same kind of quality rot as demonstrated here, though he does it with lossy video compression instead: https://www.youtube.com/watch?v=JR4KHfqw-oE(though as an anon mentioned in the last thread, there are ways to re-compress JPEGs to JPEGs without creating more compression artifacts, but they don't apply in this particular scenario)
>>272911471By an estimate, how much storage could be saved by having compressed JPEGs instead of the uncompressed PNGs? Because if the difference is negligible and won't be discernable by the average person...
>>272909129Dozed off, sorry. I'll get on it now
>>272911267This is what applying JPEG over and over again does to an image, so the less times it has to be done, the better.
>>272911581I've got the missing stuff already >>272910627 - though if we're using different methods, we might have some different results and you might get something I didn't. It also has the updated script, which should be able to handle some of the r18 ids better (handle /series/ to /series/ redirects more gracefully).
>>272899685Goddamn Satan turned the anime girls to cardboard!
Bloody hell, a 50 volume motorcycle manga that alone has about 10k pages to download.https://www.mangaz.com/series/detail/117761
Largely replicating the other anon's work, but here's a csv with the ids and names of all works that can be read for free or in trial, around 6500 of them:https://files.catbox.moe/cf66d2.csvI left out the ones that were greyed out and didn't have a button at the bottom. If the script can access those too, I'll get them for you.
>>272911919The script won't crash on them, but from what I've seen so far there's no data to get there. The ones I've seen are either those mobile app placeholder books, or "waiting for approval" books.
>>272911774https://files.catbox.moe/1nqus8.csvHere's an R18 scrape anyway if you'd like to compareShould be all 865 objects listed on https://r18.mangaz.com/title/index?category=r18I'm heading to bed now so if not I'll compare in the morning. I've had the script running this entire time, it appears to be working flawlessly so far. I'll update and resume before bed. Side note, I probably won't be archiving any of the R18 stuff, there's a fair amount of loli stuff and I don't want the bobbies at the door. >>272911919With this many title scrapes I can't imagine we'd be missing anything that's listed. Good stuff
>>272908759Basically yeah. The platform was mainly for works that have dead licenses or were totally out of print. That doesn't mean its going to be a collection of classics. A lot of good and bad mixed together.
>>272899260The nature - it is so TINY!
>>272911267PNG is way superior.>>272911227The PNGs can be hugely debloated easily by compressing them and cutting the bit depth to 8 or 4.
>>272912102I run them through oxipng, a lossless optimizer, but I don't want to do any lossy optimizations automatically out of principle. Only if there will be no other way to save everything, and even then people can run lossy compression over already saved images if needed.And to be honest oxipng-optimized images aren't that much bigger than JPGs in the end, since most of the pages are black and white and easily turned into limited palette PNGs. Color pages are still super bloated, but there's usually a grand total of one of those per volume.
>>272912458Compression on a PNG doesn't harm the quality, it just makes it take longer to open (irrelevant on anything that isn't 25 years old). Bit depth is the big space saver. The site isn't really hosting that many modern series and the scan quality isn't phenomenal so 4 bit depth for anything that isn't a color page is good. But 8 is safer if you're worried. Most of the site's old and kinda shitty quality color pages can be handled with 8 bit depth (256 colors) just fine JPG particularly sucks for BW images I find. Basically useless.
>>272912556So what can be done then?
According to someone on IRC the archive team will start archiving the website tomorrow, so hopefully whatever we're doing here is just a backup.
>>272912833Do you have the message? That's good news if true, but multiple backups is a good idea to make sure there's no single point of failure. Redundancy is always a good thing for this sort of thing.Did they say that they'll have their own methods, or could they also get help? Collaborating and sharing the load might be a good idea.
>>272912949The channel isn't exactly brimming with activity, so I have no answers for any of that, just someone mentioning the fact. Wrong time zone maybe. Chat log server is down too.
>>272911887Which is available in at least three separate editions on Kindle, one of them being an official Akita Shounen release, and isn't going anywhere. The page on MangaZ even lists the publisher it's commercially available under that contributed it to the library. Please inform yourself on the matter before shitting your pants over every single series.
>>272914554And so? It's still worth backing up, just in case.
>>272899487>>272899526I kneel.
Shit, they just basically closed melonbooks to the rest of the worldhttps://x.com/melonbooks/status/1854776176574955538https://www.melonbooks.co.jp/news/index.php?news_id=914Get a Japanese IP if it shows 404 for you
>>272917812How trustworthy is American Express? They dropped DLsite shortly after the others.
>>272917812>they just basically closed melonbooks to the rest of the worldOld news, it wasn't opening without JP IP for many months already.
>>272899764This was great, thanks OP
Can we put it on The Eye?
>>272917812>>272917956I don't even think they allowed foreign cards to begin with. I have an account with them and they refused to accept my Visa, even though other Japanese sites accepted them just fine (this was before the crackdowns). The site works for me though. I have a virtual Amex card but I don't know if they'd accept it. Piccoma stopped allowing foreign visitors a while back and I have to use a VPN to access it, but while I can open the site with a VPN it won't accept any of my cards, and I've tried Visa, Mastercard, and Amex.
>>272910627I had to manually install axios-retry for the new version to work. I guess it's not listed in required modules so "npm install" skips it.Also, Oxipng doesn't work at all for me and throws "Error: Command failed" at every single page. I'm not sure it's a Windows 7 issue this time around.
>>272918446I bought a doujinshi e-book from melonbooks with a Mastercard about 2 years ago.
>>272899176Akarin!
>>272918560You must be lucky because I was trying to buy something from there about a year ago (a digital doujin) and it would not accept my card. It was bullshit since it was the only place where it could be bought digitally.
>>272912833Fantastic news
>>272917904I got an Amex to use for Melon and it actually worked on Melon for a digital purchase, but I cannot for the life of me understand how does credit repayment work, and their card does not have debit support altogether. I have a 3 euro debt (europoor here, yes), and they sent me a bill which says that I have to pay 0 euros until next week, with the whole details on where and how to pay 0 euros. Either I don't get something, or their system is wacky.On that note, I wish there was a proper ripper for Melon that isn't an unstable canvas ripper extension. Melon app used to just give you PDFs sometimes, many years ago, but not anymore.
>>272918488I must have forgotten to include the updated package.json in the zip, sorry. 4 AM coding is a bad idea. I'll update the archive with the same link.With oxipng, according to their github issues the latest version that supports Windows 7 is this one, so you can replace the exe with it:https://github.com/shssoichiro/oxipng/releases/tag/v9.0.0
>>272918893Guess it was a 7 issue, now it works.Here's a suggestion: copy tags from the book description to metadata file as they often have useful information like when it was published and what magazine it was serialized in. And tiny cover.webp file is kinda useless, not sure why it rips it now.
>>272919068I thought that the covers might be different from the first page sometimes, so I was ripping them just in case. If they all end up being the same, we can remove them. They're lightning fast to download and indeed tiny, so it shouldn't affect ripping speed.For the tags, that has been on my mind, but the way the script works is that it doesn't load the book description page, it goes directly to the viewer page based on the book ID once it's retrieved. I was thinking to write a separate script for tags later, though maybe it would make sense to already start saving maybe even the whole page, dealing with extracting tags later or separately... but in any case, can't do much until after work.
>>272904124use hakuneko
>>272911715That's applying it with lower and lower and lower quality settings.Applying the same settings over and over again won't change anything.Applying slightly different but all relatively high quality settings will degrade the quality very slowly, you'd need to do it 100 times to see a major degradation
Found another issue where the script fucks up - some of the volumes in some of the series have metadata without the volume name, so the script tries to save them into the same folder, as it relies strictly on the metadata and thinks these are standalone volumes. Series ID 223971 is one example. I will fix that somehow...
>>272919425Updated the script to account for this in a dirty hack manner of just using the ID as the folder name, if this issue occurs when downloading a series. Kinda fucked if downloading by book IDs though, but should still be able to save to a different folder if series id is not equal to book id, yet volume name is missing.https://drive.google.com/file/d/1D6dGBcukWcatRaEvYk3MDXrPMxJRfkmI/view?usp=drive_linkShould have probably included IDs into folder names to begin with, but alas.
How's the progress on the backing up front?
>>272918679Try reading https://www.melonbooks.co.jp/special/b/0/service/e-books-guide/en/index.html
>>272919791/g/ is yet to deliver the logotype.
>>272919948I don't get it
>>272919981Were you not around when Mangadex went down?
>>272899764Amen.
>>272919791Most of the hopes are for the Archive Team efforts, but at least two anons here are trying to save non-trivial amount of the comics.My personal progress is 210/6674 series, so still a long road ahead and I'll run out of time and disk space eventually.
>>272920331how much space is that so far?
>>272919948Kek>>272919791154 series, 555 volumes. Just under 50GB.Going slower than I expected, but I'm backing them up on a server that stays on 24/7 so I'm confident I can get a good amount done.
Looks like 2.5TB is sufficient? I'm assuming this is the png variant. Also not including the paywall.
>>272920216Oh I was every much around for that, and I remember /g/'s general incompetence for anything (like how Pantsu failed). I just don't know what the logotype thing was meant to be a reference to.
>>272921190Everyone volunteered to make the logo for /g/'s manga reader website, no one volunteered to make the website itself.
How many series are paywalled?
>>272920216the meme is much older than that
>>272921236860 at least, the entire R18 catalogue is paywalled, but anon's script can get them despite that.Unsure if there's any standard manga that's paywalled.
>>272921210Oh yeah, I remember that now.
>>272921079By my very crude estimations, that's the general ballpark. I have around 3.5TB of free space (plus I can free up a bit more) so I might not even run out of space, but still likely to run out of time. There's only so much the website is capable of providing to one machine, besides the general load on the service.Hopefully anon with home servers can rip things faster, but I only have my single laptop that I also need for work.
>>272921210Was this related to some sort of dex panic? I think I remember that.
>>272922424Either that, or when the old batoto died and dex was not yet a thing.
The archive team is indeed beginning to start archival, they are also de-scrambling images like we do, and brute-forcing book IDs one by one to get everything possible. So the result should be largely identical.
>>272899211
>>272923686
>>27292094861 GB (size on disk listed as 72 GB though)
>>272923516Do they host it anywhere, or just keep it archived?
