The reason I don’t show all of them is because it’s extremely wasteful. It increases load times and also encourages people to do stuff like banking up hundreds of lessons, which messes up the word lists for other books (since lessons added from one book are filtered out of other books).
Anyway, it’s like a necessary evil to encourage people to use the website properly.
I may add an option for the alchemizer at some point. I guess anyone who’s actually paying for the server can go do whatever they want with the server
Ahh, I hadn’t considered that. I suppose there could be a lesson cap, but I don’t quite understand what you mean by the messing with wordlist thing. So like if they added words from one book to lessons, didn’t do the lesson, and then switched to another book? And with the load times, could there just be an option to load all words, defaulted to no?
Also maybe only allow people to trash after the xth item to prevent people from getting huge lesson queues like you were talking about?
Ive got like 2700 words left in oreshura, but with all the stuff I have in houhou, anki, and words that are misparsed names or simply grammar points I know…I dont know if that number is actually closer to 2600 or 2000. So far I’ve trashed around 1/3 of the words, but I feel like ill trash less and less as time goes on. If it was a patreon perk to be able to figure out the actual number, I’d say its just annoying enough to have me throwing money at you, though :3
Pretty much that. Then they’d have to start deleting lessons themselves or the word lists for other books are gonna start looking inaccurate. And I get enough crap from being about lists being “inaccurate” as is. If they start mass adding lessons and stuff gets out of whack then it damages the credibility of my website, even though it’s not my fault people are messing up their account.
Anyway, it’s just a huge PIA to do all that when people could just… click the refresh button instead.
Ive only actually really started using floflo as of like a week ago, but I think ill end up supporting it regardless of if that feature gets added. Was not aware about anki and houhou importing, which does sorta help actually (although a majority seems to be misparses).
Oof, I just went through and got most of the ones for the next part…but ill try and remember. The most recent was was missing 当たってない and parsing the kanji alone and I think the hiragana was put into its own word. イイ was parsed as iraq rather than 良い. It didnt properly get 信じがたい although it may have actually had the right definition and I just didnt see it. It parsed ぜんっぜん as just ぜん which is pretty understandable but still. It didnt properly get すみずみ or ナマイキ iirc and parsed them as something different. It parsed 何よその as 何, よそ, and の. Thats all I can really remember off the top of my head. If it would help, I can list them as I come across them, but I assumed there wasnt too much you could do. In general, it seems to be issues with casual stuff and slang mainly.
I think this should be parsed correctly as 信じる＋がたい (an auxiliary verb) but I trashed it awhile ago so I dunno off the top of my head. 信じがたい has its own dictionary entry but I don’t think it’s technically a real word.
Anyway, thanks for giving me such a detailed report. It seems like a lot of the misparsing woes are caused by misparsed katakana, often caused by substandard spelling (ex: ナマイキ). If I could find a way to just filter out all katakana words (as a foreign learner, 90% of them are useless anyway) then I think I could improve the site’s accuracy a lot.
It had the body build definition, but maybe it also had the aux verb definition and I didnt see it. おいおい had the wrong definition (correct was 追い追い) and I just had to select to see all guesses for the right one. Maybe I didnt do that for がたい.
But yeah, I mean…the people complaining complaining about inaccuracy are in for a rude awakening. There are plenty of words I have already come across that arent even in any J-A dictionary and thats always gonna mess up parsing. たまるか and モテまくり already gave me a run for my money lol.
As someone very passionate about this method of learning, I feel like its really only important that the easier texts are parsed well. The people reading those likely will have issues with actually parsing the text themselves and need floflo to be accurate since they’ll rely on it more. I remember how hard it was to parse text when I started and I hated that there were no spaces because I thought that made it impossible…but people tryna read konosuba or sao should be able to follow along and not need to fully rely on a parser to figure out where words stop and start lol. So if theres really no easy fix for weird slang and katakana words found in harder texts, I’d say heck that stoof for now. Although it is sorta your site and you can deal with it however you please :p.
EDIT: がたい did also have the aux definition and I guess I just missed it.
So for some reason my browser has been having trouble saving cookies. I haven’t changed any of the website’s code related to that in like a month, so I’m not really sure if it’s just an issue for me or if something spontaneously broke.
So like, here’s some examples of things that should save:
Guess mode for vocab tables
Beta SRS options (assuming you’re not doing a lesson)
Bookmarked books (all books page)
The last book you filtered for (lessons page)
Should not be repeatedly showing you these warnings:
Asking if you want to use the beta SRS for lessons
That alert that pops up when you hit the refresh vocab list button
Have any of these cookie-related things stopped saving for you recently?
It sometimes asks me whether I want to use the beta SRS, even though I clicked on ‘beta’, but I don’t think that’s what you’re asking about… It definitely remembers my bookmarked books, and I haven’t noticed any issues with the lesson filter or guess mode, but I might just not be remembering what I left it on.
It’s asking whether you wanna default to the beta SRS when you queue up a lesson.
Anyway, it seems to be kinda random. Like my code isn’t throwing an error but sometimes the cookies straight up don’t save. There’s a limit to how many cookies you can have per domain, so maybe I hit that or something…
Okay, my best guess is that I either hit the browser’s capacity for either A) # of cookies per site or B) total length of cookies. I’ll start by rolling all the cookies on the beta SRS into one and then see what happens from there.
That’s bad. Tell me if it happens again because that basically means there was a question where it thought you’d already done both the reading and the meaning (which means it shouldn’t have been picked to begin with).