Jump to content

Nandemonai

Members
  • Posts

    840
  • Joined

  • Last visited

  • Days Won

    3

Reputation Activity

  1. Like
    Nandemonai got a reaction from captainroyy in Shin Koihime Musou Translation Project   
    Okay, it's the first of the month (for a sufficiently large value of 'first of the month').  Time for an update!
    About 4400 lines were transledited in a month.  I decided to change things up a bit and focus on character routes for awhile.  Makes a nice change of pace from working on the main story, and I'm quite a bit farther in the main story than in the side routes; the main story is about 63% done, where the side routes are about 35% done.
    Shoku
        Main                             63% (6495/10321)
        Battle                       100% (299/299)
        Aisha, Touka                 37% (1546/4137) (+16%)
        Ei, Yue                      36% (695/1954) (+23%)
        Enya                         1% (13/2777)
        Hinari, Shuri                22% (645/2907)
        Iishe, Pairen, Reiha, Toshi  36% (1226/3377) (+36%)
        Kikyou, Shion                1% (34/2277)
        Mii                          27% (396/1448)
        Nenene, Ren                  16% (535/3330)
        Rinrin                       43% (1245/2912) (+24%)
        Sei                          69% (1885/2734) (+16%)
        Sui, Tanpopo                 77% (2610/3409) (+27%)
        Shoku TL:97% (+3%) (39236/41882)  Edit:42% (+10%) (17624/41882) Untranslated H-scene routes: (Hinari, Shuri), and Sei & Rinrin have partial h-scene translations.  (Rinrin got left off the list before)
    Gi:     4% (1390/38852)     Go:    5% (1360/25983)
    Total:    19% (+4%) (20378/108888)
  2. Haha
    Nandemonai reacted to Sayaka in Kodoku ni Kiku Yuri/Lonely Yuri randomly gets a Steam release   
    Y-you mean backlogs are clearable?
    All hail @VirginSmasher of the once empty backlog. Please teach me your ways, sensei.
     
  3. Thanks
    Nandemonai reacted to Ranzo in Craziest Synopsis You Read for a Nukige   
    What have you done?!?
  4. Confused
    Nandemonai reacted to Dreamysyu in Craziest Synopsis You Read for a Nukige   
    This already sounds like the best VN ever.
    ~~~~~
    Looked through some other messed up tags and found this:
     
  5. Like
    Nandemonai got a reaction from Strike105X in Shin Koihime Musou Translation Project   
    Okay, it's the first of the month (for a sufficiently large value of 'first of the month').  Time for an update!
    About 4400 lines were transledited in a month.  I decided to change things up a bit and focus on character routes for awhile.  Makes a nice change of pace from working on the main story, and I'm quite a bit farther in the main story than in the side routes; the main story is about 63% done, where the side routes are about 35% done.
    Shoku
        Main                             63% (6495/10321)
        Battle                       100% (299/299)
        Aisha, Touka                 37% (1546/4137) (+16%)
        Ei, Yue                      36% (695/1954) (+23%)
        Enya                         1% (13/2777)
        Hinari, Shuri                22% (645/2907)
        Iishe, Pairen, Reiha, Toshi  36% (1226/3377) (+36%)
        Kikyou, Shion                1% (34/2277)
        Mii                          27% (396/1448)
        Nenene, Ren                  16% (535/3330)
        Rinrin                       43% (1245/2912) (+24%)
        Sei                          69% (1885/2734) (+16%)
        Sui, Tanpopo                 77% (2610/3409) (+27%)
        Shoku TL:97% (+3%) (39236/41882)  Edit:42% (+10%) (17624/41882) Untranslated H-scene routes: (Hinari, Shuri), and Sei & Rinrin have partial h-scene translations.  (Rinrin got left off the list before)
    Gi:     4% (1390/38852)     Go:    5% (1360/25983)
    Total:    19% (+4%) (20378/108888)
  6. Like
    Nandemonai got a reaction from Strike105X in Shin Koihime Musou Translation Project   
    Okay, update time.  The interval for this one was a little longer than the previous few.  I decided to move these to the first of the month and see if I can't get on a regular schedule.
    Even though this interval was longer, there was vacation in there and also I had to spend some time re-editing the existing work.  I decided to change to present tense, not past tense.  Because lines like "Victory was mine!" and "I was dead" (instead of "I'm dead") are just silly.  This involved going back and re-edited the beginning parts of the game to be in present tense.  So the numbers didn't move quite as much as I'd have liked.
    Shoku
        Main                             63% (6495/10321) (+8%)
        Battle                       100% (299/299)
        Aisha, Touka                 21% (878/4137) (+5%)
        Ei, Yue                      13% (257/1954) (+7%)
        Enya                         1% (13/2777)
        Hinari, Shuri                22% (645/2907)
        Iishe, Pairen, Reiha, Toshi  0% (0/3377)
        Kikyou, Shion                1% (34/2277)
        Mii                          27% (396/1448)
        Nenene, Ren                  16% (535/3330) (+3%)
        Rinrin                       19% (558/2912) (+4%)
        Sei                          53% (1436/2734) (+10%)
        Sui, Tanpopo                 50% (1704/3409) (+49%)
        Shoku TL:94% (+4%) (39236/41882)  Edit:32% (+6%) (13250/41882) Untranslated H-scene routes: (Hinari, Shuri), (Ren, Nenene), and Sei
    Gi:     4% (1390/38852)     Go:    5% (1360/25983)
    Total:    15% (+3%) (16004/108888)
  7. Haha
    Nandemonai reacted to Tyr in How to Install and Run Untranslated Visual Novel Text Hooking Tools   
    The last three questions in this thread, all asked by an account registered solely for the purpose of asking these questions:
    [Lamputa]:                        ITH Updater doesn't work!
    [Tyr]:                                Don't use the ITH Updater. Download the newest ITH release instead.
    [Razur]:                            ITH Updater doesn't work!
    [sanatlig]:                         Don't use the ITH Updater. Download the newest ITH release instead.
    [The man with no name]:  ITH Updater doesn't work!
    I'm out.

  8. Like
    Nandemonai reacted to arosia in Our New first Visual Novel   
    A lot of developers don't really see the value of getting involved in a community. A lot of them don't do much in the way of marketing, even.
    It takes a few games--if you last that long--before you realize that it's a privilege to be able to create and have people give you their attention.
    People respond well to mutual respect and grace.
    Poor thing really did mess up here. Not only did they get a lot of negative attention, but they also missed out on one of the great VN communities available to them.
  9. Like
    Nandemonai reacted to novurdim in Our New first Visual Novel   
    Erm, so, like every forum ever created? If true, 15% is an amazingly big number.
  10. Haha
    Nandemonai reacted to Ranzo in Our New first Visual Novel   
  11. Thanks
    Nandemonai reacted to tahu157 in Our New first Visual Novel   
    Do me a favor. Edit your post, press ctrl+A, hit this button: , and then save.
  12. Like
    Nandemonai got a reaction from DarkZedge in Shin Koihime Musou Translation Project   
    Okay, it's the first of the month (for a sufficiently large value of 'first of the month').  Time for an update!
    About 4400 lines were transledited in a month.  I decided to change things up a bit and focus on character routes for awhile.  Makes a nice change of pace from working on the main story, and I'm quite a bit farther in the main story than in the side routes; the main story is about 63% done, where the side routes are about 35% done.
    Shoku
        Main                             63% (6495/10321)
        Battle                       100% (299/299)
        Aisha, Touka                 37% (1546/4137) (+16%)
        Ei, Yue                      36% (695/1954) (+23%)
        Enya                         1% (13/2777)
        Hinari, Shuri                22% (645/2907)
        Iishe, Pairen, Reiha, Toshi  36% (1226/3377) (+36%)
        Kikyou, Shion                1% (34/2277)
        Mii                          27% (396/1448)
        Nenene, Ren                  16% (535/3330)
        Rinrin                       43% (1245/2912) (+24%)
        Sei                          69% (1885/2734) (+16%)
        Sui, Tanpopo                 77% (2610/3409) (+27%)
        Shoku TL:97% (+3%) (39236/41882)  Edit:42% (+10%) (17624/41882) Untranslated H-scene routes: (Hinari, Shuri), and Sei & Rinrin have partial h-scene translations.  (Rinrin got left off the list before)
    Gi:     4% (1390/38852)     Go:    5% (1360/25983)
    Total:    19% (+4%) (20378/108888)
  13. Like
    Nandemonai got a reaction from havoc in Shin Koihime Musou Translation Project   
    Okay, it's the first of the month (for a sufficiently large value of 'first of the month').  Time for an update!
    About 4400 lines were transledited in a month.  I decided to change things up a bit and focus on character routes for awhile.  Makes a nice change of pace from working on the main story, and I'm quite a bit farther in the main story than in the side routes; the main story is about 63% done, where the side routes are about 35% done.
    Shoku
        Main                             63% (6495/10321)
        Battle                       100% (299/299)
        Aisha, Touka                 37% (1546/4137) (+16%)
        Ei, Yue                      36% (695/1954) (+23%)
        Enya                         1% (13/2777)
        Hinari, Shuri                22% (645/2907)
        Iishe, Pairen, Reiha, Toshi  36% (1226/3377) (+36%)
        Kikyou, Shion                1% (34/2277)
        Mii                          27% (396/1448)
        Nenene, Ren                  16% (535/3330)
        Rinrin                       43% (1245/2912) (+24%)
        Sei                          69% (1885/2734) (+16%)
        Sui, Tanpopo                 77% (2610/3409) (+27%)
        Shoku TL:97% (+3%) (39236/41882)  Edit:42% (+10%) (17624/41882) Untranslated H-scene routes: (Hinari, Shuri), and Sei & Rinrin have partial h-scene translations.  (Rinrin got left off the list before)
    Gi:     4% (1390/38852)     Go:    5% (1360/25983)
    Total:    19% (+4%) (20378/108888)
  14. Like
    Nandemonai got a reaction from Santiash123 in Shin Koihime Musou Translation Project   
    Okay, it's the first of the month (for a sufficiently large value of 'first of the month').  Time for an update!
    About 4400 lines were transledited in a month.  I decided to change things up a bit and focus on character routes for awhile.  Makes a nice change of pace from working on the main story, and I'm quite a bit farther in the main story than in the side routes; the main story is about 63% done, where the side routes are about 35% done.
    Shoku
        Main                             63% (6495/10321)
        Battle                       100% (299/299)
        Aisha, Touka                 37% (1546/4137) (+16%)
        Ei, Yue                      36% (695/1954) (+23%)
        Enya                         1% (13/2777)
        Hinari, Shuri                22% (645/2907)
        Iishe, Pairen, Reiha, Toshi  36% (1226/3377) (+36%)
        Kikyou, Shion                1% (34/2277)
        Mii                          27% (396/1448)
        Nenene, Ren                  16% (535/3330)
        Rinrin                       43% (1245/2912) (+24%)
        Sei                          69% (1885/2734) (+16%)
        Sui, Tanpopo                 77% (2610/3409) (+27%)
        Shoku TL:97% (+3%) (39236/41882)  Edit:42% (+10%) (17624/41882) Untranslated H-scene routes: (Hinari, Shuri), and Sei & Rinrin have partial h-scene translations.  (Rinrin got left off the list before)
    Gi:     4% (1390/38852)     Go:    5% (1360/25983)
    Total:    19% (+4%) (20378/108888)
  15. Like
    Nandemonai got a reaction from Plk_Lesiak in Clockwork Leyline and Hoshimemo: What actually happened!   
    SakuraGame sold a game that was so poorly handled, that they had simply run the whole script through Google Translate.  Script commands and all.  The resulting script bore only a passing resemblance to something the game's engine could actually execute, so naturally it would crash early on.
    I don't remember which game that was, or if it was ever fixed.  But that's pretty powerful evidence that they do not, in fact, edit their work at all.
  16. Like
    Nandemonai reacted to Plk_Lesiak in Clockwork Leyline and Hoshimemo: What actually happened!   
    I haven't seen any proofs of SG translations being edited in any meaningful ways.
  17. Like
    Nandemonai got a reaction from Toranth in Clockwork Leyline and Hoshimemo: What actually happened!   
    I work in software myself.  Their explanation makes perfect sense to me.  They fell victim to one of the classic blunders: They made a risky change (upgrading a version in their script compiler) that they thought was a mostly-harmless change.  Because it was right before the games were supposed to release, they said 'we just need to do a quick check and make sure nothing broke' when what they really needed to do was another full runthrough of the game.  Why would they do this?  Because their tests are still manual.  Someone has to manually play the game and check for untranslated lines.  This is another of the classic blunders.  Something like checking for untranslated lines, or checking to make sure that there are no 'orphaned' parts of the script (lines that exist in the data but the engine never actually displays), can and should be automated.  Anytime a build is generated, those checks should be run automatically.  This is a bit more expensive to develop up front (because you have to write the tools to run these tests, and the infrastructure to do it automatically) but it saves a lot of QA effort, and avoids humiliating incidents like these.
    So they made a risky change they didn't realize was risky, and it blew up in their faces.  Okay, fine.  There's nothing out of the ordinary about this.  It happens.  And they can't revert the change, because it would cause even less acceptable issues.  Also nothing unusual.  Then they discovered fixing it requires rewriting the game engine, which is riddled with inline assembler (ugh!), and is generally a pain in the ass.  Again, there's nothing unusual about this.
    There's two things about this that get me.
    First, Sekai's QA processes are clearly insufficient.  The game should never have gotten out of the door like this.  They need better tools and processes so they don't do so much manually.  Each project needs to build a button that fires off a test to find out whether there are any untranslated lines.  And they need to push that button every time they make any change.
    Second, this is Sekai Project being Sekai Project again:  You must have known this weeks ago ago, guys.  This happened right before Christmas, they might not have even started looking at it until early January, then it takes a while to really understand the problem.  Then it takes a while to realize 'no, this isn't a quick fix'.  By this time it's probably around mid-January.  So why was it only announced this week?  Look.  Everybody makes mistakes.  Bugs happen.  But there's no excuse for not announcing this as soon as they knew what the root cause was.  This is like them announcing they'd talk about Maitetsu 18+ at a con, then not doing it, and then leaving it hanging for months.
  18. Like
    Nandemonai got a reaction from Asonn in Clockwork Leyline and Hoshimemo: What actually happened!   
    I work in software myself.  Their explanation makes perfect sense to me.  They fell victim to one of the classic blunders: They made a risky change (upgrading a version in their script compiler) that they thought was a mostly-harmless change.  Because it was right before the games were supposed to release, they said 'we just need to do a quick check and make sure nothing broke' when what they really needed to do was another full runthrough of the game.  Why would they do this?  Because their tests are still manual.  Someone has to manually play the game and check for untranslated lines.  This is another of the classic blunders.  Something like checking for untranslated lines, or checking to make sure that there are no 'orphaned' parts of the script (lines that exist in the data but the engine never actually displays), can and should be automated.  Anytime a build is generated, those checks should be run automatically.  This is a bit more expensive to develop up front (because you have to write the tools to run these tests, and the infrastructure to do it automatically) but it saves a lot of QA effort, and avoids humiliating incidents like these.
    So they made a risky change they didn't realize was risky, and it blew up in their faces.  Okay, fine.  There's nothing out of the ordinary about this.  It happens.  And they can't revert the change, because it would cause even less acceptable issues.  Also nothing unusual.  Then they discovered fixing it requires rewriting the game engine, which is riddled with inline assembler (ugh!), and is generally a pain in the ass.  Again, there's nothing unusual about this.
    There's two things about this that get me.
    First, Sekai's QA processes are clearly insufficient.  The game should never have gotten out of the door like this.  They need better tools and processes so they don't do so much manually.  Each project needs to build a button that fires off a test to find out whether there are any untranslated lines.  And they need to push that button every time they make any change.
    Second, this is Sekai Project being Sekai Project again:  You must have known this weeks ago ago, guys.  This happened right before Christmas, they might not have even started looking at it until early January, then it takes a while to really understand the problem.  Then it takes a while to realize 'no, this isn't a quick fix'.  By this time it's probably around mid-January.  So why was it only announced this week?  Look.  Everybody makes mistakes.  Bugs happen.  But there's no excuse for not announcing this as soon as they knew what the root cause was.  This is like them announcing they'd talk about Maitetsu 18+ at a con, then not doing it, and then leaving it hanging for months.
  19. Thanks
    Nandemonai got a reaction from ArgentstR in Clockwork Leyline and Hoshimemo: What actually happened!   
    I work in software myself.  Their explanation makes perfect sense to me.  They fell victim to one of the classic blunders: They made a risky change (upgrading a version in their script compiler) that they thought was a mostly-harmless change.  Because it was right before the games were supposed to release, they said 'we just need to do a quick check and make sure nothing broke' when what they really needed to do was another full runthrough of the game.  Why would they do this?  Because their tests are still manual.  Someone has to manually play the game and check for untranslated lines.  This is another of the classic blunders.  Something like checking for untranslated lines, or checking to make sure that there are no 'orphaned' parts of the script (lines that exist in the data but the engine never actually displays), can and should be automated.  Anytime a build is generated, those checks should be run automatically.  This is a bit more expensive to develop up front (because you have to write the tools to run these tests, and the infrastructure to do it automatically) but it saves a lot of QA effort, and avoids humiliating incidents like these.
    So they made a risky change they didn't realize was risky, and it blew up in their faces.  Okay, fine.  There's nothing out of the ordinary about this.  It happens.  And they can't revert the change, because it would cause even less acceptable issues.  Also nothing unusual.  Then they discovered fixing it requires rewriting the game engine, which is riddled with inline assembler (ugh!), and is generally a pain in the ass.  Again, there's nothing unusual about this.
    There's two things about this that get me.
    First, Sekai's QA processes are clearly insufficient.  The game should never have gotten out of the door like this.  They need better tools and processes so they don't do so much manually.  Each project needs to build a button that fires off a test to find out whether there are any untranslated lines.  And they need to push that button every time they make any change.
    Second, this is Sekai Project being Sekai Project again:  You must have known this weeks ago ago, guys.  This happened right before Christmas, they might not have even started looking at it until early January, then it takes a while to really understand the problem.  Then it takes a while to realize 'no, this isn't a quick fix'.  By this time it's probably around mid-January.  So why was it only announced this week?  Look.  Everybody makes mistakes.  Bugs happen.  But there's no excuse for not announcing this as soon as they knew what the root cause was.  This is like them announcing they'd talk about Maitetsu 18+ at a con, then not doing it, and then leaving it hanging for months.
  20. Thanks
    Nandemonai got a reaction from Dreamysyu in Clockwork Leyline and Hoshimemo: What actually happened!   
    I work in software myself.  Their explanation makes perfect sense to me.  They fell victim to one of the classic blunders: They made a risky change (upgrading a version in their script compiler) that they thought was a mostly-harmless change.  Because it was right before the games were supposed to release, they said 'we just need to do a quick check and make sure nothing broke' when what they really needed to do was another full runthrough of the game.  Why would they do this?  Because their tests are still manual.  Someone has to manually play the game and check for untranslated lines.  This is another of the classic blunders.  Something like checking for untranslated lines, or checking to make sure that there are no 'orphaned' parts of the script (lines that exist in the data but the engine never actually displays), can and should be automated.  Anytime a build is generated, those checks should be run automatically.  This is a bit more expensive to develop up front (because you have to write the tools to run these tests, and the infrastructure to do it automatically) but it saves a lot of QA effort, and avoids humiliating incidents like these.
    So they made a risky change they didn't realize was risky, and it blew up in their faces.  Okay, fine.  There's nothing out of the ordinary about this.  It happens.  And they can't revert the change, because it would cause even less acceptable issues.  Also nothing unusual.  Then they discovered fixing it requires rewriting the game engine, which is riddled with inline assembler (ugh!), and is generally a pain in the ass.  Again, there's nothing unusual about this.
    There's two things about this that get me.
    First, Sekai's QA processes are clearly insufficient.  The game should never have gotten out of the door like this.  They need better tools and processes so they don't do so much manually.  Each project needs to build a button that fires off a test to find out whether there are any untranslated lines.  And they need to push that button every time they make any change.
    Second, this is Sekai Project being Sekai Project again:  You must have known this weeks ago ago, guys.  This happened right before Christmas, they might not have even started looking at it until early January, then it takes a while to really understand the problem.  Then it takes a while to realize 'no, this isn't a quick fix'.  By this time it's probably around mid-January.  So why was it only announced this week?  Look.  Everybody makes mistakes.  Bugs happen.  But there's no excuse for not announcing this as soon as they knew what the root cause was.  This is like them announcing they'd talk about Maitetsu 18+ at a con, then not doing it, and then leaving it hanging for months.
  21. Like
    Nandemonai got a reaction from skrewball717 in Shin Koihime Musou Translation Project   
    Okay, it's the first of the month (for a sufficiently large value of 'first of the month').  Time for an update!
    About 4400 lines were transledited in a month.  I decided to change things up a bit and focus on character routes for awhile.  Makes a nice change of pace from working on the main story, and I'm quite a bit farther in the main story than in the side routes; the main story is about 63% done, where the side routes are about 35% done.
    Shoku
        Main                             63% (6495/10321)
        Battle                       100% (299/299)
        Aisha, Touka                 37% (1546/4137) (+16%)
        Ei, Yue                      36% (695/1954) (+23%)
        Enya                         1% (13/2777)
        Hinari, Shuri                22% (645/2907)
        Iishe, Pairen, Reiha, Toshi  36% (1226/3377) (+36%)
        Kikyou, Shion                1% (34/2277)
        Mii                          27% (396/1448)
        Nenene, Ren                  16% (535/3330)
        Rinrin                       43% (1245/2912) (+24%)
        Sei                          69% (1885/2734) (+16%)
        Sui, Tanpopo                 77% (2610/3409) (+27%)
        Shoku TL:97% (+3%) (39236/41882)  Edit:42% (+10%) (17624/41882) Untranslated H-scene routes: (Hinari, Shuri), and Sei & Rinrin have partial h-scene translations.  (Rinrin got left off the list before)
    Gi:     4% (1390/38852)     Go:    5% (1360/25983)
    Total:    19% (+4%) (20378/108888)
  22. Thanks
    Nandemonai reacted to Clephas in Opinions on the current visual novel fanbase/community?   
    More or less this... by that choice of words, he essentially stated that, whether conscious or subconscious, that was his prejudice not only against the games but against the lifestyle.  If he'd simply said 'I can't get into yuri or yaoi' or something similar, it would have served the same purpose without causing so much trouble.  To be straight about it, I can't get into yaoi... but I don't mind people next door living the lifestyle.  By that choice of words, he essentially indirectly condemned any member of the community who does pursue an LGBT lifestyle. 
    Even if he was just referring to the VNs, he went beyond stating a preference by that choice of words... he was indirectly attacking those who do like them.  While I frequently complain about charage/moege, I do manage to enjoy many of them and I don't condemn people for playing them... and I don't condemn people for enjoying LGBT-influenced games. 
  23. Like
    Nandemonai reacted to Zenophilious in Opinions on the current visual novel fanbase/community?   
    Yeah, it really was, considering how easy it is to interpret that statement as "gay and transgender people are degenerates".  There is literally nothing else in that sentence that refers to only disliking gay/transgender characters in VNs, and it seems to link being gay or transgender with being a degenerate.  It also doesn't list any other qualities these "degenerate" characters have that makes them so reprehensible to him, other than being gay and transgender.  I'd say it's pretty hard to say he doesn't come off sounding like a bigot.
    Then there's the choice of the word "degeneracy", which implies that there's something inherently wrong with their lifestyle itself, and echoes the ideology and speech of homophobes that find the mere existance of LGBT people offensive.
    So, yeah, that was a pretty unfortunate word choice, if that's what it actually was.
  24. Like
    Nandemonai reacted to arosia in Opinions on the current visual novel fanbase/community?   
    I think it's pretty clear what you were trying to say.
    Moderation is what separates 4chan from Reddit from Fuwa.
  25. Like
    Nandemonai reacted to Darklord Rooke in Opinions on the current visual novel fanbase/community?   
    That and the word "theme" which was then clarified later on to reference to homosexual characters. Classifying homosexuality as a degenerate theme is most definitely an unfortunate choice of words
×
×
  • Create New...