If the
line is indeed broken there, a hyphen must be displayed at the end
of the first line. If not, the character is not dispalyed at all.
In operations like searching and sorting, it must always be
ignored.
The requirements for the SOFT HYPHEN in HTML are unrelated to BMS. All the BMS players which support UTF-8 display “#TITLE Changelog” as “Change-log”.
Google inserts several U+200B ZERO WIDTH SPACEs to the translation result of webpages.
I do not know how the application which is not a web browser should treat these special characters.
It seems that DTX cannot express polymeters. Even if that is possible, although nobody is unlikely to use that. (Because unrelated to a game. )
By separation of presentation and content, software may become convenient. For example, chart-viewer which can subdivide measure which looks very long. (Note: This is not proposal. )
- 機械翻訳が...;_; 当該文章だけを英語でもう一度書きます: I think, if LR were well alive today and LR started supporting LRX then BMS would have been completely displaced by LRX. But most proposals do not have such an influential software supporting them, not to mention a lack of niche softwares like nanasigroove. (ななしのBMS拡張がどの程度受け入れられたのもななしがどの程度使われるソフトウェアだったからです。)
Very thank you for your supplementary explanation ! I think that I have understood at last.
I wanted BMS to be replaced by LRX positively. But “BeatMania Simula” is too catchy.
Indeed, I did not consider the time and effort which implements CSS. I am always stupid ...
When metadata is expressed only with <meta/> elements, reading and writing may become difficult about multiplex value.
For me, the following ways are more easier than IDREFS:
In sound-only mode, unquoted #COMMENT values are also displayed as quoted strings. Conversely, on “selecting screen”, quotation marks of #COMMENT are invisible. What an elaborate behavior this is!
I think that there may also be #SUBGENRE extension, seeing this sample. ... Indeed ?
It may be interesting if the skin and sound-effects file group for a BMS package can be referred to from a BMS side. I cannot explain properly, but it looks like “standardized #EXTCHR (or Materials)”. The extension which can assure that “shared resource can be referred to regardless of the structure of BMS application” may be required.
Unique identifier (ID type) may rather bring about trouble. It may be happier for us to be satisfied with “the state of being not valid XML but well-formed XML” (e.g. LRS, XBMS) ?
It is xml:id error if first-letter of a value is [0-9].
It is xml:id error if two or more same ID values appear in a document.
/* Reminder for me :: xml:id requires @namespace */
@namespace xml "http://www.w3.org/XML/1998/namespace";
#id {/* unmatch */}
[id] {/* all just ids match */}
[xml|id] {/* all xml:ids match */}
sound[xml|id="01"] {/* all <sound xml:id="01"> match */}
image[xml|id="01"] {/* all <image xml:id="01"> match */}
I learned a lot. I greatly appreciate your indication !
I agree with you. Perhaps GUI chart-editors are the most important.
It seems that the LRX editor was due to implement the function similar to #EXWAV. In addition, virtual waveform slicer / concater were already implemented. (Video: sm11018377, sm11067837) These are “the features for simplifying making chart data from music” literally.
Supposing GUI chart-editors can have the Melodyne DNA, a BMS scene will change dramatically.
About LRX: It may be misunderstanding. LRX had solved all the problems mentioned above. #SWBGA is virtual image grouping, and LRX editor already has audio concater.
I may not understand “flexibility of format”. BMS 2.0 seems to aim at M.U.G.E.N in music video games. LRX, Sonorous, TechnicalGroove, etc. seem to aim at “better BMS”. I think that BMS apps do not need to cover DTX extension. ... I do not know well yet about XBMS.
Sharp comment to LRX. However, that #xxx separators are not bound to actual measures (in musical notation) may bring about a new musical viewpoint.
For example, Polyrhythms, Polymetres, Polytempos, etc. can be visualized now more clearly than before, by objectized separators. Since this is impossible for the conventional BMS, it is commented as follows: 마디 선을 긋는 것은 유저에게 그저 압박을 주기 위해 사용되는 기능이므로
A feeling of LRX just like a low-level language has spoiled the musical abstractness of BMS. Instead, musical abstractness is delegated to a LRX editor. The notation of LRX will keep away man. (I know what most bugs of the conventional BMS were elaborated for by hand coding)
According to the rumor, LRX was able to translate BMS into XML at the time of a play, and was able to play it with nearly perfect reproducibility. This is just above-mentioned ideal polyglot software.
If this scenario had succeeded, LRX may have been set to the last “tolerant” BMS parser. Whenever I look at angry developers, I am reminded of LRX.
Long comments make me happy! Sorry for not replying sooner.
About specification being too huge, I am also worried; many of conventional BMS apps have been developed by not a group but the individual. But so, priority is most given to developer's desires.
I cannot understand subsequent sentences well, since the accuracy of translation probably is not good. (1) “BMS may disappear”: Even if LRX has Polyglot parser ? (2) “There is even no software like nanasigroove”: What does this nanasigroove symbolize ?
Since a developer with abundant experience thinks software portability as important and develops BMS apps, we are fortunate. Since I am almost ignorant about programming languages, I cannot judge the right and wrong of specification. However, Sonorous proposal is very interesting.
Probably, rankers who challenge #RANDOM BMS are enjoying the inequity of ranking, positively.
Therefore, it is doubtful whether collecting all the combination that may exist is required. However, it is a very interesting idea. If realizing is possible, someone is likely to implement.
(Although I am not interested in IR, I know that it is important. IR is required of de facto standards.)
JSON may be able to be used for replay data (called “ghost”), trial programs, etc.
Like OGP or Microdata, multiplex <meta/> elements seem to be best practicenow. However, I cannot imagine it well ... (<meta key="subartists" value="IDREFSIDREFS …" /> ?)
Mute note is interesting ! To be sure, it is appropriate to it to exist as one of the effects applied to a note as you tell.
Since I cannot imagine difficulty of implementing, I can be taking interest simply.
This resembles the awkward inline style as used in the field of (X)HTML. How to solve this could also imitate it of (X)HTML, if such localization is ill-mannered.
Values of type IDmust match the Name production. A name must not appear more than once in an XML document as a value of this type; i.e., ID values must uniquely identify the elements which bear them.
“[0-9]” must not appear as a character of the beginning of id attribute value.
(Therefore strictly speaking, LRS is not valid XML, either. )
If “document.getElementById("id")” is applied to the above-mentioned XBMS, probably the result which we expect will not be obtained.
XBMS seems to be unable to express “empty BGM lanes”. When the arrays of BGM objects are not held, it may be inconvenient for GUI chart-editor applications.
The programming language “Rust” (version 0.8) was released on .
“Angolmois Rust Edition” completed updating for Rust 0.8 conversion.
The document was also updated. The new section “What's a difference between the Rust Edition and the original Angolmois?” was added.
A new problem is pointed out in that section:
The C version allows "rewinding" the movie by repeatedly setting the same alphanumeric key to the BGA layer; the Rust Edition doesn't. I'm not sure that this feature is portable at all.
I don't know the document which made reference about this behavior. However, since #WAV is rewindable (FINAL ATTACK RIDE D-D-D-DECADE), I consider it more desirable (in my opinion) for #BMP to be also rewindable.
I don't know BMS which uses “rewinding of video” intentionally.
BMSE Tips
When BMSE opens a BMS file, a part of commands which BMSE does not support are removed.
However, these may be able to be protected by #IF.
To be sure, for example CSS3, it is already almost impossible to cover and describe vendor prefixes and old and new syntax about “Gradients” and “Flexible Box Layout Module”.
This proposal will become important from now on. [Wide display, Multi display, Retina display, High-Definition display, High DPI, Scalable Vector #BMP, etc. ]
#LANES ??? (I want to specify “either is fine, users' decide”)
Lane gap:
Style
Image
Code
Separate
| | | | | | |
| | |2| |4| |
|S|1| |3| |5|
#LANES ???
Half
| | | | |
| | 2 4 |
|S|1|3|5|
#LANES ???
Either
#LANES ???
#CANVASSIZE and #LANES interfere in appearance (called “theme” or “skin”). Therefore, first of all, the argument about “a desirable structure about appearance” may be necessity.
// translation may not be right
#TITLE Point, Line, Plane
#TITLEja点、線、面
#TITLEko점, 선, 면
#TITLEzh-CN点，直綫，平面
#TITLEzh-TW點，線，平面
#WAV01 vocal-en.wav
#WAV01ja vocal-ja.wav
#WAV01ko vocal-ko.wav
#WAV01zh-CN vocal-zh-Hans.wav
#WAV01zh-TW vocal-zh-Hant.wav
#BMP01 lyrics-en.mpg
#BMP01ja lyrics-ja.mpg
#BMP01ko lyrics-ko.mpg
#BMP01zh-CN lyrics-zh-Hans.mpg
#BMP01zh-TW lyrics-zh-Hant.mpg
#00101:01
#00104:01
When BMS application does not support “Simple Localization”, or when priority is given to neither Japanese nor Korean nor Simplified Chinese nor Traditional Chinese:
#TITLE Point, Line, Plane
#WAV01 vocal-en.wav
#BMP01 lyrics-en.mpg
#00101:01
#00104:01
When the language to which BMS application gives priority is set as Korean:
#TITLE 점, 선, 면
#WAV01 vocal-ko.wav
#BMP01 lyrics-ko.mpg
#00101:01
#00104:01
This feature prevents distribution of the hash in Internet Ranking, and makes it possible to offer BMS efficiently.
[Writer Requirements] Probably, I guess that when authors use commands with language tags, authors should write together the command without a language tag as a default value.
#TITLE quovadis
#TITLEla quovadis
#TITLE bolero
#TITLEfr boléro
beatmaniaIIDX has song called “one or eight”. This is Janglish which translated Japanese words meaning “sink or swim” literally (一か八か). Which are the suitable language tags to this?
#TITLE one or eight
#TITLEen one or eight
#TITLEja one or eight
#TITLEen-JP one or eight
This feature may be a busybody extremely. Although I can read only Japanese and can hear only Japanese, I do not want to hear “Ask to the wind” (바람에게 부탁해) in Japanese. I want the option which changes a language setup temporarily.
... In the Internet-Ranking, the mechanism in which “바람에게 부탁해” and “風にお願い (JP version)” can be treated as the same BMS may be desired. The situation which ranking separates even by the bug-fix of BMS is not desirable. But probably, solution of this problem will be difficult.
reply (2)
Explicit Key Sound of Long Note Endpointsはまだ構想中だから待ってこない方がいいです;_; ただロングノートの末に重なるTick Objectが同じような役割をするように設計され、ロングノートの末のkey soundが存在する場合、同じ位置にあるTick Objectと同じタイミングで再生されなければならないためにひょっとしたらこれで十分だっかもしれません。実装問題さえなければ当該sectionがTick Objectに吸収されるかもしれません。
tick-obj will be able to include lnend-keysoundcompletely.
However, lnend-keysound has a merit which does not require new channels (#xxx70-8Z).
lnend-keysound also has a merit which can be described briefly. But combo becomes ambiguous.
About this point, I agree with your opinion completely.
However, there are two problems.
The game balance in the recent BMS applications is moderately consistent.
The “COMBO” in ruvit is different from “TOTAL NOTES”. (“COMBO” does not influence #TOTAL. )
Practically, the more game balance becomes tight, the more environment is limited.
For example, BMS which is “#TOTAL 60” is assumed.
In nazobmplay which has an initial value of a groove-gauge 22%, this BMS is able to clear.
In MixWaver which has an initial value of a groove-gauge 9%, this BMS is not able to clear.
There are some implementations which can change the initial value or rate of change of a groove-gauge. A life bar (HARD gauge or EXPERT gauge) may be applicable instead of a groove-gauge. Should these be “standardized” ? I cannot judge it ...
Of course, the situation where the increase in a gauge is in inverse proportion to the number of LNs is not desirable. But my being able to propose is only setting a loose value to #TOTAL. I delegate all the game balance to each user. If the HARD option, the HAZARD option, etc. are evaluated on the Internet-Ranking, I think that it must be sufficient.
Hello! It's very very hot this summer ... Sorry for not replying sooner.
I installed SDL 2.0.0 into MinGW and checked updated Angolmois (C version).
I created angolmois.lnk and have changed “Property -> Target” as follows:
C:\WINDOWS\system32\cmd.exe /v:on /q /c for /l %n in (1,1,2) do if %n==2 (for %q in (!x!) do start "" /b /min "D:\bms\angolmois\angolmois.exe" any-angolmois-command-line-options %q) else set x=
If we drag and drop a BMS file to this shortcut icon, Angolmois will start.
In e5cea53a2cbd or later, supposing cmd.exe option “/min” is specified, an Angolmois execution window will not become active. Meta-informations (e.g. #TITLE) are not displayed, either. (Note: This is not a bug report. This is a memorandum record for me.)
I checked that the behavior about the “Overlap” of #BGA had been changed. (Now, “Overlap: Yes”)
[e5cea53a2cbd] Some images of newbgatest.bms are distorted just like a fish-eye lens.
