One of the reasons why I love Scrabble3D application is that because single Scrabble games with large tile set such as 500 tiles or 1000 tiles can be played in it, either with large 2D boards or with small 3D boards. Tamil is my native language and games with large diverse letter set can be simulated in it, with large number of letters being sparsely used. Also that number of tiles in rack can be varied inbetween 2 and 20. Even having 20 tiles in the rack does not slow down the performance of computer opponent players by that much for large 2D boards and small 3D boards. Small 2D boards would not represent all of large letter set and large 3D boards would slow down the performance of computer opponent players significantly enough. Do you think that other people would not be interested in playing games with 500 tiles or 1000 tiles, then?
Or do you think that other people would not be wise enough to refer to here for frequency of each character in the Tamil dictionary? Or do you think that other people would not be wise enough to modify the letter set by themselves according on to their own tastes before starting on a game to play of Scrabble with 2D or 3D and board size according on to their own wishes?
Zitat von Scotty im Beitrag #16Okay, thanks. How about my concerns regarding the large number of letters?
This is a question for which it would be good enough to have your joint opinion with me. I have provided the dictionary. You can also generate a letter set for 100 tiles, or may be at least for 200 tiles, if it is at least OK for you, by yourself, based (depending) upon the frequency of the different characters that are being used in the dictionary.
How about 95 predefined tiles, 3 random tiles and 2 blank tiles? Or would it be better to have 90 predefined tiles, 8 random tiles and 2 blank tiles? If I were to implement classic Scrabble, or at least Superscrabble for Tamil, if it is at least OK for you, then for best results I would always generate random tiles for it respecting the frequency of the different characters that are being used in the dictionary.
Random tiles respecting the frequency of the different characters that are being used in the dictionary can also be generated by the players by making use of a computer program prior to the start of a game to be played. And that letter set (and that character distribution) would be different for each of the games to be played.
Remember Yahoo! Literati, randomly generated tile set at end of each player's turn and not at the start of the played game? So that if a tile is swapped then the same tile is not guaranteed to come back again at all returning later onwards within the game to be played. Points of tiles in that varies from 1 points to 5 point only, 0 points for blank tile and that supports English Scrabble only. 1 point tiles: ACDEGILMNORSTU, 2 point tiles: BHP, 3 point tiles: FKY, 4 point tiles: VW, 5 point tiles: JQXZ. Letter set (character distribution), dictionary, number of blank tiles, number of random tiles, number of tiles that are being used within rack, everything can be customized within Scrabble3D application prior to the start of a game to be played.
Large number of characters would round down to 0 in Tamil tile set if it were implemented enough only for a smaller tile set. Contrary to your own opinion, in my own humble opinion, for a smaller tile set, it would be much better off to divide down from the precise requirement of each character that is represented (got) in a larger letter set for any played game of Scrabble, rather than (instead of) multiplying up the requirements for a larger tile set from a smaller tile set, which would always be rounded as a multiple of the multiplier used from the original (base) number of the tile set that is being used, rather than (instead of) representing (getting) the precise requirement of each character in a larger letter set for any played game of Scrabble, then.
Zitat von Scotty im Beitrag #13And maybe I missed the point but what's the difference between Tamil and TamilZH? Which one should be on SF.net?
Please note that it is Tamizh and not Tamilzh, kindly enough. The voiced retroflex approximant is replaced by making use of the voiced alveolar lateral approximant and not appended (concatenated or juxtaposed next) to it at all, then.
Most people know the classic Scrabble game with ~100 letters. Scrabble3D has a predefined board for this (and option to switch to Superscrabble or 3D). You need to dig deep into the configuration to adjust everything for 1k letters. Starting from that it should be easy to multiply the number so a larger field is possible, not vice versa. I would use the distribution with zero count for rare letters and have a good number of randoms (that are evenly distributed not prioritizing the letters with zero count). But in the end it's up to you, maybe I'm wrong in respect to your language.
I am thinking of a new design. Please give me 2 or 3 days of time to implement it. I suggest that you do not rush to upload the Tamil dictionary to sourceforge to make it available for everyone. I will post over hereby once again whenever that I am ready enough, ever.
I would like to know about: (1) If Tamil dictionary is uploaded to sourceforge, then will I have access to make corrections to words in the dictionary if needed into it and change letter distribution and points for tiles, as I try out more demos with different number of tiles in the rack to come up with a furthermore flexible solution later onwards?
Even if you think so of right now, is it possible for you to add quintuple, sextuple, septuple letter score, word score, letter score malus to the board design of the Scrabble3D application source code and make a new release? It would be good enough for 2D larger enough board sizes or 3D larger enough board sizes. I do not insist you to do the same, just I am asking if it would be possible for you to do that so easily enough.
I would like to know about: (2) If it is possible to add up desired number of random tiles and blank (joker) tiles as needed within in the predefined letter set in the dictionary directly to load them automatically whenever the Tamil dictionary that is being loaded by the players ever, rather than loading up 90% to 95% of predefined letter set from the dictionary in the configuration settings and manually adjusting the desired number of random tiles and blank (joker) tiles as needed by the players manually over thereby.
Also that please look up into my own desired density of premium squares in 15×15×15 Scrabble 3D board from Tamil1000.ssg from Game.zip zip file that is being attached in my own post number 21 in this thread, kindly enough. I do not recommend sparsely populated density of premium squares that comes along within Scrabble3D application inbuilt package for into 15×15×15 and then 21×21×21 Scrabble 3D board at all. My own 21×21×21 Scrabble 3D board can also be built up within my own desired density of premium squares by making use of computer program into it up, too, later onwards, that is being, whenever I would be rather than free enough, instead of whenever that ever.
With latin script alphabets along X-axis and with numbers along Y-axis, the board co-ordinates are in forward order. But with the Greek alphabets, why are the board co-ordinates along Z-axis in reverse order?
By the way, if new dictionary, or new features (for example free multiplier, larger than 48×48 board, support for rectangular and then cuboid boards, board design for 27×27 board, 33×33 board, 39×39 board, 45×45 board) are added to the Scrabble3D application, then will it get updated for everyone having Scrabble3D application whenever that it checks up for updates, up ever?
Just out of my own curiosity only, the maximum allowed board size is 48 because it is the sum of count of uppercase and then lowercase Greek letters along Z-axis? Much to say actually that maximum size of allowed tiles in rack is 20 because of the fact that it is being a round number. Exceeding that would cause too many permutation of letter set in rack for a playable game that would slow down computer opponent player's performance, always offline, never online and that it is not being suitable for human opponent players online or offline as well, except that it is well suited for letter sets and language dictionaries with diverse character set such as Tamil. In my own made version of Tamil in 2014 with Scrabble board (Superscrabble board) that is being printed on chart paper printed on both of front side and then back side of cardboard sheet that is being cut with scissors, we used 10 tiles in rack with front side and back side choice of selection that is being allowed to be used. Whenever that using 20 tiles in rack, that is being equivalent enough and then good enough substitution of used 10 tiles in rack with front side and back side choice of selection that is being allowed to be used, only ever, actually enough.
Zitat von Scotty im Beitrag #20You can update the dictionary, but please not every day. And yes, free multiplier is possible in theory. Would be an interesting addition.
I think that I would add up a few words in the dictionary, the ones with bracketed (ளு), (ணு), (லு), (னு), (நு), (வு), (ழு), (ரு) and (வரு) and provide a finalized version of my own Tamil dictionary in 2 or 3 days. I will post about it once again around over hereby by then. I would also change the letter set to a good enough start up of 100 tiles which people can freely enough change them manually as they would desire from the configuration settings letter set. I think that you could publish that as an initial release.
I would also remove a few words which are meaningless anagrams of other words present in the dictionary and I noted about them that are being present in the Tamil Lexicon that is being published by the University of Madras whenever that I tried out ever of a Tamil Scrabble 15×15×15 Scrabble 3D board played game with 1000 tiles in it, that is being attached within into my own previous post number 21 in this thread, upon upto. This would be my own initial release and it would be good enough to have other collaborators whoever that are being joining with me as well to provide the different inflected forms of all of types of nouns and verbs that are being present within into the dictionary, then. This would be good enough to limit the strength of computer opponent players only, always offline and never online, then human opponent players can always make use of validation by poll if word is not being present within into the dictionary. That is being, invalid words present in the dictionary that are being to be removed by me right now, within into the next fewer enough days.
I think that you can subscribe to this thread if you want to, and then come online to reply to my own posted messages within into this thread, whenever that you get an e-mail alert ever, upon upto.
Please take Tamil.dic from Tamil.zip for publication of Tamil dictionary to sourceforge as initial release. I have made all of types of corrections to it to the best of my own knowledge as of right now, and I will not be willing to modify it once again at least for a while. Tamil.dic is for words with definitions. Tamizh.dic is for words without definitions. Both of those types of files should be successfully read by making use of Scrabble3D application, without dropping off any of the types of the words that are being present within those dictionaries. I added 534 new words in the dictionary, the ones in the Tamil Lexicon that is published by University of Madras containing bracketed letters of types of (ளு), (ணு), (லு), (னு), (நு), (வு), (ழு), (ரு) and (வரு) in the word, at all. 9 of these types of words were already present in the previous version of my own Tamil dictionary that was being attached in my own post number 8 in this thread and new definitions have been concatenated along with the old ones, at all. Old number of words = 115914 and new number of words = 116448. 2 of the words in my own previous version of my own Tamil dictionary that was being attached in my own post number 8 in this thread have been corrected, that is being as follows: ஒற்றுக்கேடடல் -> ஒற்றுக்கேட்டல். தவுக்கார்பணுதல் -> தவுக்கார்பண்ணுதல். 2 of the words that were previously dropped by me were added up back again because I believe that reasonable meaning is given for them in Tamil Lexicon that is published by University of Madras, that is being as follows: க்கு. க்குதல்.
I have left number of tile count at 1000 in my own Tamil dictionary file, to better give an idea to players of the frequency of each Tamil letter that is present in the Tamil dictionary. Players can adjust them freely enough at their own wish for smaller or larger letter set than that. 1000 tiles is the most ideal letter set for Tamil dictionary to represent each Tamil letter and their frequency that is present in the Tamil dictionary properly to players in the letter set, but that I just changed one piece of சீ to ா to better represent the frequency of the different Tamil letters that are present in the Tamil dictionary. It is something that is best fit for 45×45 square board or 15×15×15 cube board, because of the diverse number of Tamil letters that is made use of in the Tamil dictionary and numerous rarely used Tamil letters over thereby. My version should be something that is readable joining vowels ி, ீ, ு, ூ and ் with their consonants and separating vowels ா, ெ, ே, ை, ொ, ோ and ௌ from their consonants and presenting a version that is at least suitable for play with decent number of distribution of each Tamil letter that is present in the Tamil dictionary. This version separates ஔ into ஒ and ள as two different tiles. Alternate designs such as joining of all of types of vowels ா, ி, ீ, ு, ூ, ெ, ே, ை, ொ, ோ, ௌ and ் with their consonants would give a fewer enough tile set and that is furthermore flexible enough to play, and that would not be well readable by any of the players, isn't it? This version joins ஔ from ஒ and ள as one single tile. In both of the above cases, all of types of vowels of அ, ஆ, இ, ஈ, உ, ஊ, எ, ஏ, ஐ, ஒ, ஓ, ஔ and ஃ remain separate and do not joining with any of types of consonants.
Code.zip contains 2 new additional files: revise.py python script that I used to revise my own previous version of Tamil dictionary that was attached in my own post number 8 in this thread to my own new version of Tamil dictionary that is being attached in my own post number 23 in this thread, which is this. Input file = revise.dic for reading old words, tamil.txt for adding new words and output file = tamil.dic. revert.py python script is used to convert all of types of symbolic characters and accented non-ASCII characters that are used for substitutes for lowercase Latin Script ASCII letters back again into lowercase Latin Script ASCII letters. Input file = tamil.dic and output file = convert.dic. The input folders used in python scripts are folders that I used in my computer system. For other people making use of that same python script, they need to change that based and depending upon their own configuration settings. Output folder is not mentioned in python script. It is given as argument in command prompt while running python script, > for writing newly enough to some of type of file and >> for appending to it.
Flawed.zip that is being attached in my own post number 24 in this thread, which is the next post contains Tamil dictionary with all of types of symbolic characters and accented non-ASCII characters being replaced with lowercase Latin Script ASCII letters. As the name says, it is flawed and does not work out properly at all. Scrabble3D application confuses some of, if not all of, type of uppercase and lowercase Latin Script ASCII letters and they could be used interchangeably, either by making use of human players or computer players, either online or offline. Although that it is definitely worth noting that any of type of computer player is only available offline and never online at all, then. In Flawed.zip, convert.dic contains words with meanings, convert_lexicon.dic contains words without meanings. Both of those types of files should be successfully read by making use of Scrabble3D application, without dropping off any of the types of the words that are being present within those dictionaries. Being a shorter enough file, convert_lexicon.dic can be easily and quickly opened by making use of me, making use of Windows 10 operating system for any of types of reference sakes and purposes at all, then. Both of types of files can be used for Tamil Scrabble helper by making use of find.py python script in Code.zip, by parsing through all of types of words that are being present in the Tamil dictionary. tamizh.py python script in Code.zip converts Tamil dictionary containing all of types of words that are present in it with meanings to one containing all of types of words that are present in it without meanings. Input file = tamil.dic and output file = tamizh.dic as that is being present in Tamil.zip, but that it is also adjustable as well to input file = convert.dic and output file = convert_lexicon.dic as that is being present in Flawed.zip too.
Further that I am not sure if whether it would be possible to attach two files or more with combined size of greater than 12 MB furthermore, at least if not for one single file size, it is that, it is being.
By the way, actually that maths dictionary that was being authored by making use of Jouni Tolonen <kaissa@lycos.com> contains quite a number of invalid mathematical equations such as those like 18 - 9 = 7, 0 ^ 0 = 1, ∞ > 15, ∞ = ∞ + 1, etc. and so on. Would you agree to it if some other author such as those like me collaborate with any of type of other people's own work, whether that it also applies to other of types of collaborators contributing along with my own prepared enough Tamil dictionary? It is that, it is being prepared enough by making use of me. In my own humble opinion, that maths dictionary can never be fully implemented as number of mathematical equations that can be present in it of less than a specific enough size can be very large enough and unlimited if that restriction is being removed as well, also, too. Best of type of thing to implement is that, it is being parsing of Context Free Grammar by making use of Lexical Analysis, Syntax Analysis and Semantic Analysis at all, then. So, that would also apply to be best of type of implementation for computer play as opponent, may be that is, it is being by making use of a trie data structure only, enough, actually.
, (whatever that corresponds to the lowercase Latin Script ASCII letter of x ever) is not made use of in the predefined letter set of 100 tiles or 1000 tiles that is being defined in my own prepared Tamil dictionary, it corresponds to the Sanskrit letter of ஷ் and has been replaced by making use of ý. ? (whatever that corresponds to the lowercase Latin Script ASCII letter of z ever) is not made use of in the predefined letter set of 100 tiles or 1000 tiles that is being defined in my own prepared Tamil dictionary, it corresponds to the Sanskrit letter of க்ஷ் and has been replaced by making use of ÿ.
Sanskrit letters of ஸ, ஜ, ஷ, ஹ, க்ஷ, ஶ்ரீ and not ஶ are present in the dictionary at all, but that they are not being defined in the predefined letter set of 100 tiles or 1000 tiles that is being defined in my own prepared Tamil dictionary, then. ஶ is undefined in Valluvan true type Tamil font, but it cannot be used for replacement by other symbols and accented non-ASCII characters that are defined in the Valluvan true type Tamil font.
Words starting with க்ஷ also start with க. And words starting with ஶ்ரீ also start with ஶ.
Whenever loading a dictionary with a predefined letter set, Scrabble3D application always prompts for 'Current letter set does not match. Updated from dictionary info?' even if the letter set in the configuration settings that was being last updated and stored whenever that the Scrabble3D application that had been exited ever, matches with the letter set that is being present in the predefined letter set of the loaded dictionary.
@Scotty: Would you mind making one last correction to the Tamil dictionary? Please change the Letters, Values, Counts as follows (each of these 3 should be in one single line):
I initially prepared a letter set, but if count of 0 is assigned to very rarely used letters and Sanskrit letters, people would be able to use them for jokers and randoms. (I do not recommend people to use randoms for very rarely used letters as they are very hard to push them on the board). People could add the 0 count letters to the letter set by themselves if they want to, by referring those letters in this thread, but if they are predefined in the dictionary itself, they could use them without any hurdles, as they could simply modify the counts and values as per their own wishes, without the need to add or delete any extra row of letters from the letter set. The letters ஙி, ஙீ, ஙூ are not used in the Tamil dictionary. The letters ங, ஙு, ஞி, ஞீ, ஞு, ஞூ, ழீ, ழூ have been assigned 0 count because they are very rarely used in the Tamil dictionary. The letters ஸ், ஸ, ஸி, ஸீ, ஸு, ஸூ, ஜ், ஜ, ஜி, ஜீ, ஜு, ஜூ, ஷ், ஷ, ஷி, ஷீ, ஷு, ஷூ, ஹ், ஹ, ஹி, ஹீ, ஹு, ஹூ, க்ஷ், க்ஷ, க்ஷி, க்ஷீ, க்ஷு, க்ஷூ, ஶ்ரீ have been assigned 0 count because they are Sanskrit letters and are not part of the regular Tamil letter set.
For people interested in playing Classic Tamil Scrabble with 100 tiles, the following python script will be useful for generating 98 tiles based upon the frequency of the letters in the Tamil dictionary (attached in tile.py in Code.zip and that python script had been added up to the zip file newly): This will generate different tile distribution as needed for different games to be played and the output of this program can be copied and pasted into letter set data of Scrabble3D application. This would work out as long as the words in the dictionary are not protected by their own authors only: Best is to output results to a text file (python tile.py > tile.txt for writing output to new file and python tile.py >> tile.txt for appending output to a new file): tile.py
from random import random from math import floor file = open("C:\\Program Files (x86)\\Scrabble3D\\tamil.dic", "r") MAX_TILES = 98 MAX_COUNT = 400 # MAX_LENGTH = 25 # categories = [] d = {} t = {} total = 0 while True: header = file.readline() if header == '[Words]\n': break for line in file: word = (line[:-1], line[:line.find('=')]) ['=' in line] # if len(word) > MAX_LENGTH: # continue # if ';' in word and int(line[line.rfind(';')+1:-1]) not in categories: # continue j = 0 while j < len(word): letter = word[j] if letter == '\xc2' or letter == '\xc3': letter = word[j:j+2] j += 1 j += 1 total += 1 if letter in d: d[letter] += 1 else: d[letter] = 1 for s in range(MAX_TILES): r = floor(random() * total) + 1 count = 0 for (key, value) in d.items(): count += value if count >= r: if key in t: t[key] += 1 else: t[key] = 1 break for (k, v) in t.items(): f = (d[k] * float(MAX_COUNT)) / total l = 0 if f >= 4.5: l = 1 elif f >= 2.5: l = 2 elif f >= 2: l = 3 elif f >= 1.5: l = 4 elif f >= 1: l = 5 elif f >= 0.7: l = 6 elif f >= 0.5: l = 7 elif f >= 0.25: l = 8 elif f >= 0.15: l = 9 else: l = 10 print k, v, l print '--' file.close()
The following script is designed for Tamil dictionary, but the dictionary file can be modified for the other languages also by changing tamil.dic to another dictionary file name. Change MAX_TILES to the number of random tiles to be generated respecting the frequency of the different letters that are present in the dictionary. Change MAX_LENGTH to the maximum number of letters in words to be parsed in the dictionary (uncomment MAX_LENGTH and categories if it is needed to filter parsing dictionary words by these options also). Add comma separated integer list of values to the categories to parse only words that are present in the dictionary that belong to these categories. MAX_COUNT = 400, but this can also be changed. This is for generating values for letters respecting the frequency of the different letters that are present in the dictionary. In my written python script, my written computer program generates points for tiles as follows, but that they can also be changed for Tamil language and other languages: Frequency of letter in dictionary ≥ 4.5/400: 1 point. Frequency of letter in dictionary between 2.5/400 and 4.5/400: 2 points. Frequency of letter in dictionary between 2/400 and 2.5/400: 3 points. Frequency of letter in dictionary between 1.5/400 and 2/400: 4 points. Frequency of letter in dictionary between 1/400 and 1.5/400: 5 points. Frequency of letter in dictionary between 0.7/400 and 1/400: 6 points. Frequency of letter in dictionary between 0.5/400 and 0.7/400: 7 points. Frequency of letter in dictionary between 0.25/400 and 0.5/400: 8 points. Frequency of letter in dictionary between 0.15/400 and 0.25/400: 9 points. Frequency of letter in dictionary < 0.15/400: 10 points.
Some queries regarding the Tamil dictionary: Valluvan font (available from Google or in Font.zip that is attached by me in post 1 in this thread) is needed to play Tamil Scrabble. Is it possible to inform about this to people whoever use the Tamil dictionary for play? I do not prefer to use the Unicode font (which does not require any additional font to download) because the letters ா, ெ, ே, ை do not appear separately without the dotted circle letter. I am not joining vowels ா, ெ, ே, ை, ொ, ோ, ௌ from their consonants so as to have fewer number of letters in Tamil Scrabble letter set and provide more longer words in the Tamil Scrabble board and not to make two or more tile clusters to hit each other with too many shorter words played. I am not separating vowels ி, ீ, ு, ூ and ் from their consonants so as to make words more readable in Tamil Scrabble board.
What-If-Only variant might be well suited for Tamil Scrabble because of very diverse letter set and to provide with more flexibility of words in Tamil Scrabble board. What-If-Only variant might be well suited for Scrabble with mathematical equations as it requires a good balance between numbers, operators and =.
[Bug] In the letter set, if multiplier at the top of the page is changed, then the total letter count does not get updated if the number of jokers is assigned to be 0. The number of tiles in the rack and the number of randoms used in the game are independent from the letter set and so changing the multiplier at the top of the page does not change them. However that the number of randoms assigned at the bottom of the page should affect the total letter count, which does happen properly.
I tried out a Tamil Scrabble game that was played with computer day before yesterday in 15×15 Scrabble board with 90 predefined tiles as mentioned in my previous post, 8 random tiles and 2 jokers. [Bug] (This is the first game that I am trying out random tiles). It appears that random tiles are not properly displayed when they are placed on the board. The random tiles that are used are வ் in J5, வ் in M12, ை in N12, ம in D9, றி in F14, த in K13, ங் in H12, த் that was left over in computer's rack at the end of the played game. Only one த was generated as random but 4 த are shown as random in the board. Only one ை is generated as random but 3 ை are shown as random in the board. ம in D9 which is a random is not displayed as a random in the board. The game was played with 15 tiles in the rack, and the rarely used letters can only be generated by randoms. Very rarely used letters and Sanskrit letters were not even defined with 0 count in letter set in that played game and so they cannot be even part of jokers and randoms. In my humble opinion, it would be best to use the font that is used for displaying the tiles also with letter that should be chosen for blanks, played moves in Messages and Game Course, alert dialog box for validation of move by poll by other human players if a word played is not present in the dictionary. In Scrabble3D application, for Tamil Scrabble, Scrabble with mathematical equations and Scrabble with other languages, a word is not necessarily invalid if it is not present in the dictionary, other human players can validate by poll and computer can accept it if all other human players concur with it. But if an invalid word that is present in the dictionary is played in the Scrabble board, other people cannot object to it no matter whether they like it or not. Quick addition and deletion of words from the dictionary can however be done from File -> Word Search from the menu bar and that feature is available both offline and online. These changes will reflect only for a specific player and not for other players whoever are using the dictionary, isn't it?
For answering my questions, please take your own time. No rush at all. You have answered all of my major questions. These are small questions about my afterthoughts that I can come to know at my own pace. [Question] The random seed that is generated at start of new game and one that is used in saved games determine the random sequence of tiles that appear in the rack that is generated for the played games, isn't it? Is generation of random tiles (if number of randoms is assigned to be 1 or more) also dependent upon random seed or independent of it? What algorithm is used to generate them? If it is dependent on the random seed, then is it generated before the random sequence of tiles generation that appear in the rack for the played games or after it?
[Question] This can also answered by myself by experimenting with trials. In configuration settings in Scrabble3D application, it is possible to set threshold to exchanging tiles in rack to whenever there are as low as 0 tiles that are left remaining in the bag. What happens when players try out to exchange more number of tiles than that are left remaining in the bag? Under WESPA rules, there should be at least 7 tiles left in bag to exchange tiles to restrict players from exchanging Q in English Scrabble whenever there are less than 7 tiles that are left remaining in the bag. Also the tiles to be exchanged are kept aside and new tiles are drawn from the bag. But what about that in the Scrabble3D application? While players try out to draw new tiles from the bag for exchanging of tiles, then the old tiles are put for replacement back in the bag once again or not?
It appears that at least one human player should play whenever going to Game -> New Game menu bar. All of them being computer players is not allowed in this selection option. Demo can be used to test the flexibility of letter set, number of tiles in rack, number of jokers and number of randoms used in a game to be played for a specific enough language or mathematical dictionary.
This is a demo Tamil Scrabble game played with 4 computer players in 45×45 Scrabble board, with 20 tiles in rack (maximum allowed), with 1000 predefined tiles as mentioned in my previous post, 0 jokers and 100 randoms (maximum allowed) including very rarely used letters and Sanskrit letters that are present in the Tamil dictionary by assigning 0 count to them in the letter set of the configuration settings in the Scrabble3D application, as follows: By the way, maximum number of jokers that are allowed in any of type of game that can be played in Scrabble3D application is also 100. [Feature Suggestion] When a very rarely used letter in the dictionary (less than 10 occurrences in the dictionary) lands on computer player's rack, can the computer player be programmed to push such letters in the Scrabble board more efficiently, instead of getting stagnated in the computer player's rack at the end of the game that is played? Instead of going always for high scoring moves, computer players can be cleverly programmed to accumulate all of the letters that are needed for any word more efficiently to push the very rarely used letter in the dictionary (less than 10 occurrences in the dictionary) that land on computer player's rack and then to play that word in the Scrabble board. Also that human players will always certainly cleverly find out a way to focus upon accumulate all of the letters that are needed for any word more efficiently and then to push away that very rarely used letter in the dictionary (less than 10 occurrences in the dictionary) that lands on human player's rack and then play remainder of the game also with the power of the tiles in the rack with that letter replaced by another frequently used letter and not playing many more moves with the very rarely used letter in the rack and playing with one less letter with the remainder of the letters that are in the rack. Exchanging that letter is not a good idea, as it is a high scoring tile, and at the end of the game that is played, one of the players need to play it if they do not want to stagnate in the player's rack at the end of the game that is played and it would be much more harder to play that letter as the game progresses onwards. Such letters would always be assigned 0 count in the predefined letter set in the dictionary by me, and they would only be used if their letter count is changed to non-zero by any of types of players prior to start of new game that is played and they can be at least be also be used by making use of jokers and randoms. (I do not recommend people to use randoms for very rarely used letters as they are very hard to push them on the board).
Spot the best move in this board position. These are good puzzles to give to people who are interested in Scrabble games that are played in Tamil language just by making use of Tamil dictionary. (My previously sent images and posted messages to you contain information and details about all of these types of solutions, I believe in them so).
[Bug] When importing a screenshot of a board position in Scrabble3D application, then why are the letters that were made use of for the joker tiles also displayed in the joker tiles instead of just displaying them as joker tiles without any of types of letters that are written on them at all?
There are 2 different ways in which that desired letter set that is required for using and playing with any of type of dictionary that is available in Scrabble3D application can be uploaded into its own desired installation package, as follows, isn't it? One is by predefining that desired letter set in the dictionary itself, and the other is by uploading into its own desired place and loading with tournament rules in under that of rules -> tournament rules menu bar of configuration settings, after all, isn't it?|addpics|ubb-23-31b6.png-invaddpicsinvv,ubb-24-0971.png-invaddpicsinvv,ubb-25-3fea.png-invaddpicsinvv|/addpics|
Zitat von RamanPlease change the Letters, Values, Counts as follows...
Done, version 1.0.1 is online now.
Q1: All random() functions are synchronized with the seed. If you pick a random locally it needs to be the same across network players. Q2: IIRC, it's implemented according the rules and exchange is blocked if number of tiles in the bag is less than the number on the rack. F: Sounds like a bad idea. Actually there are really good algorithm that go far beyond my coding skills. Use Google scholar and search for "genetic algorithm scrabble" B: It's a feature not a bug. Jokers are in Scrabble3D not just blanks, they are assigned to a character and this is shown.
Zitat von Scotty im Beitrag #27Done, version 1.0.1 is online now.
Tamil dictionary right now looks fine to the best of my knowledge. Thank you. Zipping the dictionary compresses the file by 75%. If you want, I can give a smaller dictionary file, with meanings of words included only upto the first semicolon. Simply writing a computer program to print meanings of words upto first semicolon would not work out properly. It needs revising as if there had been multiple meanings for the same word especially when they had been separated by a space, then the primary meaning of the word could not be necessary the first meaning of the word before the first semicolon in the comma separated meanings of words, which were appended alphabetically according to the list of the Tamil Scrabble words by my python script as a computer program. In Scrabble3D application, meanings of words placed on Scrabble board are only shown upto first semicolon when mouse hovers upon them. Anyway that displayed meanings of Scrabble words are secondary, the validity of the word is primary for human play or computer play, online or offline.
During the past week, I see that majority of the Tamil dictionary downloads were from Tunisia and France. What has Tamil language got to do there? Tamil is a South Indian language, although spoken by minority groups in Sri Lanka, Malaysia, Singapore, etc. So, if the downloads were from there, at least then it would have made sense.
It would be good if English dictionary is updated to latest TWL and CSW2019 versions and English phonetic dictionary had option to turn on and turn off using diphthongs and option to choose whether a diphthong can be combined from two monophthongs. CSW2019 words without definitions: https://drive.google.com/file/d/1oGDf1wj...IWMh5uJs8s/view. CSW2019 words with definitions: https://drive.google.com/file/d/1XIFdZuk...CrrgJbLBxV/view. I am from Chennai, I am right now working as a software engineer in Bangalore. I participated in Mu Sigma International Scrabble tournament in Bangalore on January 11 - 12, 2020 in D division (also known as Newbie Division) and Bangalore club tournaments on February 2, 2020 and March 8, 2020. And also various online tournaments held in isc.ro by Karnataka State Scrabble Association and Scrabble Association of India during the period of curfews and lockdowns because of COVID-19. http://indianscrabble.com/games/mu-sigma-2020/ CSW2019 dictionary is used in all such Indian tournaments and in most of the places around the world, including Thailand, Australia, New Zealand, etc. If Scrabble3D application updates the English dictionary to the CSW2019 version, then everyone can make use of it without any hurdles at all. I have local copies of these dictionaries and people can find them in Dictionaries.zip zip file that has been attached in post 13 in 'Unfixed bugs in Scrabble3D application' thread in 'Bugreports & Suggestions' forum under 'English' group. sowpods.dic for Collins Scrabble Words and ipa_english.dic for English phonetic Scrabble word list with option to turn on and turn off using diphthongs and option to choose whether a diphthong can be combined from two monophthongs. Anyone can create for themselves CSW2019 dictionary file from CSW2019 word list with or without meanings by writing a computer program. With my sowpods.dic file, people can also choose whether or not to include all of the words which are specific enough only to each of CSW2012, CSW2015 and CSW2019 word lists.
Zitat von Scotty im Beitrag #27B: It's a feature not a bug. Jokers are in Scrabble3D not just blanks, they are assigned to a character and this is shown.
[Bug] Actually that I missed about mentioning 'tiles in the rack' in my previous post. If anyone wants to import screenshot of any board position to give it as a 'spot the best move in this board position' puzzle to another person, then why are blank tiles in the rack shown with the letter that had been assigned to it after the move had been made in the game, instead of showing it as a blank tile by itself? I am talking about the பூ tile in the rack in the last image that has been attached in the previous post of mine in this thread.
In Twitter, I tweeted about Tamil dictionary addition to Scrabble3D application yesterday to @sorkalam, @UntangleIndia, etc. to find out minded community of people interested in Tamil Scrabble. Other people are also welcome to contribute to fill gaps of different types of inflected verbs in Tamil dictionary and contribute to design of letter set, etc.
Zitat von Scotty im Beitrag #12Some minor issues: I'm not happy with 1k tiles. It's easier to multiply the number (there is a control on top of the list) than shrinking. Maybe you have numerous rarely used characters that need a frequency of zero, which would multiply to n in that case. But even then, isn't it easier to adjust these zeros to ones? Perhaps I'm wrong and you cannot play classic Scrabble at all and need super large boards.
You are not happy with 1000 tiles, but the letter set included along with the Tamil dictionary is only to give an idea of the precise frequency of each letter that is present in the Tamil dictionary to the various players whoever would use it. Players can modify the letter set at their own wish before the start of a game starting from there. Players can also decline the update of the letter set every time when the Tamil dictionary is loaded by clicking 'No' to the alert dialog box of 'The currently active letter set does not match with the information stored in the dictionary. Updated now?' if any player's own desired letter set had been previously stored in the Scrabble3D application. By the way, it should have been 'Update now?' and not 'Updated now', isn't it? And if players want to update the letter set any time with the information that is stored in the Tamil dictionary, then they can always easily load the Tamil dictionary and click 'Yes' to that above mentioned alert dialog box.
Starting from smaller number of tiles, using multiplier at top of letter set page to get larger number of tiles would give round numbers only, in multiples of that multiplier used only. In my humble opinion, I would actually prefer it to manually set the precise number of letters needed for larger number of tiles depending upon the frequency of the different letters in the dictionary and to reflect the same on the Scrabble board in any game, which can be always easily listed out by using a computer program, and then actually shrinking from there. For English Scrabble if that would be played with CSW2019 dictionary, starting from 100 tiles, that would give for 1000 tiles, the following letter distribution: A: 85, B: 23, C: 36, D: 41, E: 117, F: 17, G: 30, H: 25, I: 82, J: 6, K: 15, L: 54, M: 29, N: 58, O: 62, P: 29, Q: 4, R: 67, S: 67, T: 57, U: 39, V: 12, W: 13, X: 7, Y: 18, Z: 7. instead of: A: 90, B: 20, C: 20, D: 40, E: 120, F: 20, G: 30, H: 20, I: 90, J: 10, K: 10, L: 40, M: 20, N: 60, O: 80, P: 20, Q: 10, R: 60, S: 40, T: 60, U: 40, V: 20, W: 20, X: 10, Y: 20, Z: 10.
Based upon the languages that I am familiar with, in Scrabble3D application, right now that I can play English Scrabble, Tamil Scrabble, Maths Scrabble and English phonetic Scrabble. I also have helpers for most of the other languages that come along with the Scrabble3D application either in some website or in some mobile app. I have also written a python script as a computer program to do just that from the various dictionary files that come along with Scrabble3D application if the dictionary file has not been protected by their authors by making use of some key, and that reflects the exact version of the dictionary file for the Scrabble helper that is also made use of in the Scrabble3D application.
For Scrabble helpers that are implemented by me, for starting up with, people can take a look at help.py in Help.zip zip file that has been attached in post 39 of mine in 'Unfixed bugs in Scrabble3D application' thread in 'Bugreports & Suggestions' forum under 'English' group for English phonetic Scrabble. And find.py in Code.zip zip file that has been attached in post 25 of mine in this thread for Tamil Scrabble.
In this forum, it is not possible to view the revision history of any post in this forum, and starting to revert to or modify any post starting from that commit point, even for moderators of a particular forum and for supermods and administrators of all the forums, unlike that of many of the types of the other forums for which this type of feature would be available.
Times of India article about Sorkalam, on Monday 27 July 2020. https://timesofindia.indiatimes.com/city...ow/77187987.cms The Hindu article about Sorkalam, on Saturday 1 August 2020. https://www.thehindu.com/news/national/t...cle32247463.ece My own humble opinion just simply about Sorkalam played game, for people whoever would be interested to read this forum post of mine, if they would find this thread of mine in this forum through Google Search Engine about Tamil Scrabble:
Most people would psychologically prefer playing with wooden tiles for Scrabble, instead of board and tiles printed on chart paper, pasted with gum on front side and back side of cardboard and cut with scissors even if they can be made for themselves cheap with their own board layout, letter distribution and making extra tiles for ensuring different tile distribution in each game if some small amount of tiles picked up at the beginning of the game to play from a big pool. Wooden rack, board and tiles for any language, or any letter set including maths equations, English phonetic, Tamil phonetic, perfect squares, triangular numbers, Roman numerals, etc. can be ordered from Ahaa Furniture Store in Bangalore, India, https://www.ahaa.in/ which is owned by Abhijit Shilotri, a member of Karnataka State Scrabble Association (Bangalore Scrabble Club). I am not sure how much it will cost for set of 100 tiles. Shop owner has to be enquired about that. Also that most people would psychologically prefer playing with computer programs such as those like Scrabble3D application if the user interface is good enough, isn't it? With computer (that is actually always available only offline and never online) or other human players (that is actually always available either offline or online). In Scrabble3D application, I have created the version of Tamil dictionary that can be best played with it, which is more readable and given that number of tiles can be rack can be set to any number ≤ 20. It can be also increased ≥ 21 easily by editing the source code and compiling Scrabble3D application from that, but crash occurred during computer's turn from my own modified and compiled version of Scrabble3D application. Here are some alternate designs to Tamil Scrabble that would be more better to play but not supported by Scrabble3D application:
(a) In my 2014 make of Tamil Scrabble with board and tiles printed on chart paper, pasted with gum on front side and back side of cardboard and cut with scissors, front side or back side can be chosen for each tile. This is a player's choice just like there would be choice in question paper especially in Tamil school exams. So in this way, a difficult letter need not be compulsorily played and it can be simply pushed with an easy letter on the back of it, and given that difficult letters are always only paired with easy letters on the back of it. First we tried with 10 tiles in rack and it did not form words so easily, so in later games we decided to pick up 50 tiles per person at the start of a game for the entire game randomly from the big pool. And push away tiles wisely as fewer and fewer number of tiles remain with each player. All these games were played always only with board layout of 21×21 Super Scrabble board, that I printed on chart paper to play it for cheap when I did not want to buy original English Super Scrabble set from abroad for so much money when I could also make my own desired letter distribution for it by myself, with different points for different tiles even for the same letter. This is an interesting setting currently only possible to play with other human players online and Scrabble3D application hangs during computer's turn offline under current setting as it does if digraphs are included in letter set without replace function and also in that case only the first letter is shown in the tiles.
(b) This is the way how Tamil Scrabble would be good to be played, as designed by Sorkalam game designers R. Santhanakrishnan (kris_rs on Twitter) and his wife Preetha Raghavan: Tamil has 12 vowels, 18 consonants (plus 5 other Sanskrit consonants that are not part of core Tamil), 216 combos of consonants and vowels (plus 60 other from Sanskrit consonants that are not part of core Tamil) and 1 special character ஃ which is rarely used. For handling 216 letters which are combos of consonants and vowels, they keep separate wooden tiles for consonants, and transparent glass tiles for vowels which can be placed on top of them (not for 12 pure vowels in played games of Sorkalam at all, which are always printed and made only in wooden tiles). 2 separate racks are maintained for wooden tiles (which would be 10 in number) and glass tiles (which would be 5 in number) for each player in any played game of Sorkalam. It is played in 17×17 board with 90 wooden tiles, 55 glass tiles and 20 blank tiles. Some of the letters are not used, for example, ழௌ ளௌ றௌ னௌ, some of the letters are seldom used, for example, ங வரிசை except ங், so is ஞ வரிசை except ஞ் ஞ ஞா. Tell Tamil words with ழீ ழூ? தமிழீழம் குழூஉ? With vowels at the middle of the words which would mostly and usually occur only at the start of the words? Anyway that in my version of Tamil dictionary in Scrabble3D application, I have separated vowel combos for ா ெ ே ை ௗ ஔ by creating separate tiles for them, and to provide more flexibility and space by forming longer words and to reduce too many number of Tamil letters in 2D or 3D Scrabble board which would be rarely used.
My created version of English dictionary for Scrabble3D application or Tamil dictionary that is taken from Tamil Lexicon published by University of Madras after 3 days of processing work by me by making use of python script as a computer program would always include the meanings of each word in dictionary that would be displayed in Scrabble3D application whenever mouse hovers over any played word in Scrabble board in any played game, currently ongoing or one that has been saved and loaded. So for perfect squares and perfect cubes to tell the number what number's square or cube is that of, or for Roman numerals what number is that of, or for triangular numbers upto what number n from 1 if continuously summed is that number obtained of: instead of showing that as a binomial coefficient ⁿ⁺¹C₂ is shown instead of that.
In my humble opinion, many aspects in design of Sorkalam game can be improved before trying in tournaments which might be several years later. Currently that their only aim is to bring them in on e-commerce platforms and there is not even PC app or mobile app for it right now: (1) There are no transparent tiles in Sorkalam for vowels u (ு) & oo (ூ) to combine with other consonants. Or they are not kept separate. I didn't see them in any photos of game. Their reply: How u(ு) & oo(ூ) sounding உயிர்மெய் are to be used in Sorkalam can be seen in the playing instructions. They are written on blank tiles instead of wood+seethru tile combo. This is bcos உயிர்மெய் like கு,கூ/து,தூ/பு,பூ need diff symbols to add on க/த/ப Vs just புள்ளி (க்/த்/ப்) or கொக்கி(கி/தி/பி). My humble opinion: ு & ூ can be used with these symbols on transparent tiles to give them same weightage as other combos.
(2) Also Sanskrit letters ஸ, ஜ, ஷ, ஹ, க்ஷ, ஶ்ரீ & ஶ aren't properly handled that appear rarely in Tamil loanwords. Their reply: Wrto letters from Sanskrit - they aren't core Tamil so ignored. Of course, for Sorkalam gamers inclined to use such letters, blank tiles (of which there are 20) provide the flexibility for such needs. One blank tile per round per player is allowed.
(3) I am not happy with 20 blank tiles. That is too much. Letter distribution of Scrabble set in all other western languages make use of exactly and only 2 blank tiles per 100 tiles of any played game. All Indian languages (Tamil, Telugu, Kannada, Malayalam, Hindi, Odia, Bengali, Gujarathi, Punjabi), Sinhala and Tibetian make use of Abugida writing system. 20 blank tiles means that wild plays with bingo words of greater length running from one premium word score to another premium word score would be possible during some given player's turn that might or might not be unfair enough to one or more other opponent players in same played game. That is based out of their own reply of theirs: One blank tile per round per player is allowed. Instead more tiles in the rack can be used, in my humble opinion, whether it is of either of the 2 rack sets that are maintained in the played game, that for wooden tiles or that for glass tiles. These types of things can be customized in Scrabble3D application: board size, board layout, number of tiles in rack ≤ 20, tile distribution, points for each letter and dictionary words and I have done the same type of things in all of types of my own played games of mine in the past 3½ months in Scrabble3D application, since whenever that I had started making use of Scrabble3D application ever to try out and play games of all of types of variants of Scrabble played games.
Classic Scrabble is popular to most players in whatever language Scrabble is played while some other players like to play longer games with Super Scrabble also with more balance of tiles and number of turns to prove their vocabulary and caliber. Beyond that people would not be interested to play with larger boards and larger letter sets for them to become popular in any way. They take too much time to play to be patient and special boards have to be printed for each size and each board layout if not in computer and letters in tiles in computer are too small and computer opponent players take too much of time to play their turn if playing in computer such as those like Scrabble3D application depending upon size of dictionary, number of distinct letters in language alphabet and number of branches in trie data structure if it is to be built right from scratch for adding every subsequent letter of a formed word in any given specific language for playing in every board square along both 2 axes for 2D boards and for playing in every board cube along all 3 axes for 3D boards. For 3D boards, 15×15×15 board with 1000 tiles is the minimum size of cube Scrabble board and letter set that would be good enough to play, not anything less than that for size of cube Scrabble board and letter set. For 2D boards, instead of playing 1 game of 45×45 board with 1000 tiles, 10 games of 15×15 board with 100 tiles can be played in tournament games to give precise and balanced score to all players of played games at end of 10 played games to prove all player's vocabulary and caliber. Also that all 10 games of 15×15 board with 100 tiles would use base letter set and not precise and balanced scaling to 1000 tiles to represent precise frequency of each letter in that corresponding language dictionary. For Tamil Scrabble however, some letters would be dropped off every game if 10 games of 15×15 board with 1000 tiles are used, although that Scrabble3D application can make use of random tiles, that are totally random and should be minimally made use of and not reflecting precise frequency of each letter in the dictionary. In my opinion, reflection of precise frequency of each letter in the dictionary can be made manually by adding a specific letter in Scrabble3D application letter set in configuration settings multiple times but that would mean that Scrabble3D application would hang during computer's turn (that is actually always available only offline and never online). It might be only suitable for play with other human players (that is actually always available either offline or online).
(4) My points for some tiles are higher to reward more points for playing difficult tiles, especially for 12 pure vowels which mostly occur only at the beginning of the words, but in some special cases, such as those like as they appear in poetry, they can also appear in the middle of some words which are actually also of special interest to all of types of Tamil Scrabble playing gamers.
(5) All 12 pure vowels and 1 special Tamil letter ஃ which is rarely used are made use of exactly and only once in same 17×17 board without respecting the actual frequency of each of the 12 pure vowels in the Tamil dictionary which would amount to more than one tile for some of the Tamil letters and no tiles for some of the other Tamil letters. And exactly and only one tile it would amount to for the remainder of the Tamil letters which do not fall into both of types of the 2 categories that I have already mentioned above. In my humble opinion, it would be more fun enough to have different tile set each game in not necessarily 15×15 board each game, but of not necessarily different board size each game as it can be played each game in Scrabble3D application with customization each game in Scrabble3D application of board size, board layout, number of tiles in rack ≤ 20, tile distribution, points for each letter and dictionary words.
It is a good enough thing that in Scrabble3D application, if a language dictionary can be exclusively and exhaustively produced, then if a word is not present in dictionary then it is not necessarily invalid. If playing against computer opponent players (that is actually always available only offline and never online), it asks only for human players for validation if it can be accepted or not and it will be accepted only if that human player agrees to it. Whenever playing against other human players (that is actually always available either offline or online), then it asks for all other human players by poll for validation if it can be accepted or not and it will be accepted only if all other human players agree to it. No need to add or remove valid or invalid words respectively from dictionary and to close and reload played game with the new dictionary settings. By the way, these changes would always only be personal and would not reflect to other human players making use of Scrabble3D application all over the world and that would reflect automatically when over specific interval of time or manually when checking for updates if the changes have been updated in sourceforge website page. You would only do updates to dictionaries in sourceforge website page if they are requested by their own authors, and not other people, although you do not have to depend upon the original authors of the dictionary to make with their own updates to the dictionary in the sourceforge website page. This does not apply for computer opponent players (that is actually always available only offline and never online) playing invalid words while other human players (that is actually always available either offline or online) will honestly never do it, and so it is only worthy enough and good enough to play against other human players than to play against computer opponent players. |addpics|ubb-4d-968a.jpg,ubb-4e-6d9f.jpg,ubb-4f-bfa8.jpg,ubb-t-a7f4.png-invaddpicsinvv,ubb-s-136a.png-invaddpicsinvv,ubb-v-a973.jpg-invaddpicsinvv,ubb-w-8b84.jpg-invaddpicsinvv,ubb-u-b32b.jpg-invaddpicsinvv,ubb-x-40bc.png-invaddpicsinvv,ubb-y-35aa.png-invaddpicsinvv,ubb-1x-e6e1.png-invaddpicsinvv,ubb-1y-3d73.png-invaddpicsinvv|/addpics|
Sie sind vermutlich noch nicht im Forum angemeldet - Klicken Sie hier um sich kostenlos anzumelden
Download (Sourceforge.net)
Manual (Wiki)
Fanship (Facebook)
News (Twitter)
Offline
