rich-c: verify rich-c: refresh
moved to room Meeting Place
changed username to Pamela rich-c: hi daughter Pamela: Hi Dad Pamela: you're the only one here - where is everyone? rich-c: you're the second arrival tonight Pamela: did someone else come and go? rich-c: I was wondering the same thing - it is Wednesday, isn't it? Pamela: last time I checked rich-c: no, I came first, you're second Pamela: I'm actually down at Lindsays rich-c: really? well, I'm sure her computer works well too rich-c: what are you doing down there? Socializing? Pamela: well enough to do the job Pamela: yes, Lin's in Ottawa and Barbara is here cat sitting Pamela: we've been chatting for an hour rich-c: is she on DSL or dialup Pamela: I suddenly realized what time it was Pamela: Lin? She's on DSL rich-c: well, Barbara can look over your shoulder or chat or whatever too rich-c: guessed as much, your responses are coming up fast Pamela: no she's busy watching TV Pamela: I think that's only because it's just the two of u Pamela: s rich-c: btw Cynthis phoned this morning while we were out, left a message saying all her tests came out PK Pamela: what the heck is PK? rich-c: OK Pamela: farn dingers rich-c: smartass daughters... Pamela: nothing smartass about it - I was serious! rich-c: is Barbara cat sitting or kitty sitting? rich-c: OK, sorry if I gave you a moment Pamela: meaning fosters? No, Lin doesn't have any fosterlings at the moment - it's just Smudge rich-c: guess that is cat-sitting then Pamela: well, Smudge doesn't do much sitting : ) rich-c: did you hear that some medical researchers have a clue about cat allergies? Pamela: what kind of clue? rich-c: seems they figure there may be a control in about ten years rich-c: seems mice are allergic to catrs and they've found the protein that caus3es teh problem Pamela: how did they find out the mice are allergic????? rich-c: now if they can just find a cure for wayward typing fingers Pamela: I only wish Pamela: me, I'm a great believer in the backspace key rich-c: interesting question; I have no answer Pamela: well it certainly begs the question rich-c: it surely does, though I must say it makes sense Pamela: anyway DAd, i just came in to tell you I'm going to be late Pamela: I still haven't had dinner or done anything productive rich-c: like, you're going back upstairs now to have dinner? Pamela: I should, but I don't want to leave you here by yourself rich-c: oh, I'll play freecell and wait for someone else to show Pamela: i'll be back before too long rich-c: you go do your thing and come back asap Pamela: was there anything from Rich in e-mail? rich-c: ok see you shortly rich-c: not so far Pamela: okay, maybe he'll show before too long. I'm for the kitchen for now. See you in a bit. rich-c: see you then Pamela: little kerpoof : )
Pamela left chat session
moved to room Meeting Place
changed username to Ron Ron: hi rich-c: hi Ron, glad you're early Ron: scary eh rich-c: Pam is the only one who's been on and she had to go away for a bit Ron: aha rich-c: so hows by you?
moved to room Meeting Place
changed username to Daniel B Daniel B: wow! rich-c: bonsoir, Daniel Ron: still packing Daniel B: smaller number of members each week rich-c: sold it yet or any good offers? Ron: hi daniel Daniel B: hi Rich! hi Ron! Ron: not yet Daniel B: Is it normal at this period of the year? rich-c: Pam will be back, don't know whre Dr. D. or Harvie or BobS are rich-c: no, usually the core gang turn up quite reliably - I imagine Guy will be along later rich-c: I gather james is still wandering about Canada so we won't see him tonight Ron: m rich-c: you an the Mac or Dosbox tonight, Ron? Ron: mac Ron: laptop rich-c: Mac laptops are an expensive luxury rich-c: I'm hoping to gt a new laptop soon Ron: but i have a sub in my lap so you guys do not have my complete attention rich-c: can get a very nice new one for $1300 but it has more bells and whistles than I want or need rich-c: what flavour sub? Ron: subway club rich-c: right - Subway are a very wide-spread franchise Ron: yes - tentacles even here in comox Daniel B: note : I'm searching info about fast fourier transform. I forgot how i applied the formulas in my programs and I notice that the inverse transform is not well done. rich-c: I don't east between meals and would not get that for a regular meal so I'm unfamiliar iwth it Ron: this is supper rich-c: that's one of those things Dr. D. or Dale would know about rich-c: yes, when you're batching it convenience counts Ron: you mean there is an application for fourrier series in gaming Ron: ?? Ron: There- the laptop is back on my lap Ron: I'm expecting compoany in about 1/2 hour to 45 min, so might have to leave earlly Ron: Must say I'm growing tired of BC Liberal as on TV at suppertimne... paid for with my money Ron: Anybody alive? Daniel B: weather here : night, still have snow on the ground, received more snow this monday, raining during 3 or 4 days, small problem with water accumulation, and that's all. Ron: messy Ron: I remember it well from my days in Ottawa Ron: This time of year, you can expect just about anything Daniel B: I'm working on my coleco web site before continuing my other projects. I installed a trial version of somekind of html editor tool with a limited period of 30 days. Daniel B: I'm not updating my web site online, I want to make it works on my computer first. Ron: Well yes, that's the best thing to do. What edeitor do you use? Daniel B: I downloaded a tool to verify my html syntax and ... my old web site had more errors than i expected. Ron: you write html from scratch then? Daniel B: Yes, because I'm a programmer, I didn't use a wysiwyg tool that generate code for me. I prefer keeping control of it. Daniel B: cse html validator lite is free Ron: right. And that way you get only the code that you want Daniel B: quite useful when we know the hot-keys Ron: I can do a very basic site from scratch, but I need help with tables and such Daniel B: excatly, only the code i really want Ron: It's not all that difficult really, it's just that I don't do it often enough Daniel B: because of the limited geocities free service, I have no much bandwidth to keep all the information, zip files and text without doing a huge cleanup. Ron: do they charge for use of space or bandwith? Daniel B: I also downloaded compression softwares winrar, 7zip, PowerArchiver and many more to be able to produce standard zip files but with the best compression as possible. Daniel B: no, they don't charge for the extra bandwidth, they simply block the access to the site. Ron: I see. And how many meg are you allowed for your website? Daniel B: 5 meg Ron: gonna have to leave Daniel.... I have company just arrived Ron: g'nite Daniel B: I'm actually using two accounts Daniel B: otherwise it's impossible to put all the programming tools and extra things like games and emulators in 5 meg Daniel B: good night!
moved to room Meeting Place
changed username to Jillian Jillian: Hello! Daniel B: hello Daniel B: Ron had company just arrived so he is not there. Jillian: How are you today? Daniel B: I'm fine Jillian: have others been here tonight? Daniel B: problem with water... raining during 3 or 4 days Daniel B: Pam will probably be back later Daniel B: but the others... Daniel B: i don't know Jillian: I haven't been on in a long time, but I have made progress with Adamcon, so thought I would jump on. Jillian: I finally found a hotel that will suit, so should have pricing and be accepting registrations within a week or so. Daniel B: ho yeah, the adamcon this year... where? Jillian: Have you had flooding because of the rain? Jillian: It's in Whitby Ontario. Jillian: Just east of Toronto. Daniel B: flooding? not a big one but enough to be mad. Jillian: I hope you would be able to join us. Jillian: two apartments ago, Dale and I had lots of problems with floods. Daniel B: If I can find someone who want to do the travel from quebec to toronto, why not... but it will be difficult to find someone. Jillian: we were in a basement and more than once found a foot of water at one end of the basement when we woke upl. Jillian: Hmm... I might be able to find someone who would be willing to come from Montreal, I'll keep my ears open. Jillian: You prefer not to drive alone? or don't drive? Daniel B: I have no driver liscence Daniel B: I'm not a driver, i'm a walker. or a user of bus Jillian: If you came down on the train, bus or a plane, we could arrange for transport from the station. Jillian: If that opens options. Daniel B: We will have to talk again about the possibilities Jillian: Yes. Daniel B: It could be a great experience for me to join the adamcon meeting this time.
moved to room Meeting Place
changed username to Guy B. Guy B.: Greetings!!! Jillian: Hi Guy Guy B.: Hi Jill, what's the latest frim Dale on the con? Jillian: I found a hotel that meets our needs, so I should have pricing within a week or so. Jillian: Once I get pricing set, I'll be ready to accept registrations. Jillian: The dates are the fourteenth of JUly to the seventeenth.
moved to room Meeting Place Jillian: Daniel, would you be interested in presenting? Jillian: I know that people really enjoyed "your" presentation last year. Daniel B: I'm not good enough to talk in english infront of you guys. I will probably prefer to speak in fren-glish or something like that. Jillian: Well I think that would be fine, it would give the Ontarians among the group to practice translation. Jillian: What do you think Guy?
(Jillian stops and smells the flowers.) Guy B.: All right. Sounds good. I will put the dates on my vacation schedule. Hi Daniel Jillian: I'll send out an email when I have a registration web page ready. Guy B.: Problem with me. Foreign accents are hard for me to hear. Jillian: hmm that's true. I think that there are several people who could(and would) translate though. Daniel B: well, speaking slowly can be the solution ;-) Jillian: Pam speaks French well and I used to do phone support in French so we could muddle along. Jillian: jeffrey just collapsed into my lap. i think it's time for bed Guy B.: That will help, but I never took any foreign language when I was in school. Although my mom told me one time that she wanted to have my grandfather teach me Italian, but that didn't happen. Guy B.: Sounds like one little tike is tired today.
changed username to james Guy B.: Hi James james: hi Daniel B: hi james james: sorry for the tardiness Jillian: i took french until grade twelve, but actually learned most of my french while working at the local burger king. Jillian: hi james. james: hi, i'm back home in japan now Daniel B: tardiness? i didn't know it was a good english word... tard -> late :-) Jillian: somewhat more formal than most use.
Jillian changed username to Dale Dale: Hi james: anyway, back. was a long trip home Daniel B: hi Dale Daniel B: long time no see Dale: Howzitgoing? Dale: Yes. Dale: You weren't on the last two time I've been on. Dale: I can't figure out what the answer is to your RLE compression question is. Dale: I think that you should organize the characters in columns, and generate the name table algorithmically though. Daniel B: well, i think i didn't talked too well of my problem because it's not about the rle compression itself Daniel B: it's how to "arrange" data to help the rle compression Dale: So the pattern table would use character 0 in row 0,1,2,3 column 0 and character 1 in row 4,5,6,7 columm 0. Daniel B: I have a pdf file (in french) with pictures to show you my problem Dale: and character 8 in row 0,1,2,3 column 1. Dale: Did you email it already? james: too bad neither pam nor rich are here today Daniel B: not yet, because of the adamcon mailing list limited email size. Daniel B: but i can email the pdf to you directly Dale: Okay. Dale: How are oyu doing James? Dale: When is that trip to Canada scheduled for? Did I miss it? james: i'm back home, dale :) james: got back to japan two nights ago Dale: That's what I get for not paying attention. Oh well. Daniel B: I sent the pdf file to you dale Dale: Still have to go visit your neck of the woods. Dale: Maybe not this year. james: it's quite a trip james: from my place here to my parents' place in ottawa, it's close to 28 hours door to door Daniel B: Dale: keep me informed if you find a solution about my optimisation problem. :-) Dale: I've been to Austrailia, which was 37 hours (including a 6 hour layover) james: yeah, australia is quite long. if i ever go there, it'll be from here and not from canada Daniel B: Rich-C send me this email : Daniel B: I seem to have fallen offline and can't get back on - james: lol Dale: Daniel, are you planning on coming to AdamCon this year? Daniel B: hope this will get through. If I can't get back on Adam primary I'll try secondary. james: maybe that's where the others are Daniel B: I would like to try to go at the adamcon this year. but i'm not a driver, no car, and no money. Dale: Hmmm... Dale: I'll have to think about that one. Try and keep the dates free at least. Dale: I recommend Via Rail as a good way to get here from there. Dale: You'd want to take the train to Oshawa. Daniel B: the only money I got came from a TPS check I received today. not enough to use train i think. Dale: What do you think of my idea of encoding the columns of the Mode 3 pattern table? Dale: Certainly the left two pairs of blocks would encode well in your example. Dale: Is it the pattern table, or the colour table that you're encoding? Daniel B: in mode 3, the pattern is also the color Daniel B: because each byte in the pattern are 4 pixels Dale: 4 bits per block, 2 side-by-side blocks per byte. Daniel B: err. 2 pixels Daniel B: yes Dale: 4x4 pixels, right?
moved to room Meeting Place
changed username to rich-c Daniel B: yeah big pixels rich-c: hey! the gang's all here! Daniel B: equ 4x4 hi-res pixels rich-c: sorry, something 3 jumps downstream from my ISP crashed - dumped me earlier Dale: big pixels I've been calling blicks. Dale: blocks Daniel B: no problem rich! we were talking on a solution to get me in the adamcon this year. rich-c: that would be really super, Daniel Daniel B: In my pdf file, i think I used the word block to design 2-bytes. Daniel B: so a block of 2x2 big pixels rich-c: what we need to know now though is whre and how much Daniel B: jillian was there and wanted to talk about it rich-c: good show - your presence would be VERY welcome Dale: The budgeted amount for the 'Con is $300 US for a regular full delegate. rich-c: OK - got teh hotel booked yet? Daniel B: mmm.. I don'T have this money Dale: Jill has one in mind, but I'm not sure she's left adeposit yet. rich-c: how many meals will be included? Dale: Breakfast is provided by the hotel with sausage, eggs and bacon. Dale: There is lunch planned. rich-c: that should look after that, though for me it's heavy Dale: For the banquet we have a favorite resteraunt picked out. Dale: I'm not sure yet about the other suppers yet. rich-c: my normal breakfast is one glass orange juice, two slices toast, two cups coffee Dale: Depends on the value of the CDN dollar, I think. Guy B.: Hi Rich rich-c: well, our dollar is moving up nicely right now - really more depends on when teh U.S. dollar crashes Dale: I imagine they also have juice, toast and coffee. Dale: Not a breakfast for me. rich-c: hi Guy, finally got back on Guy B.: Dale, you will let us know when we can register. Dale: It will be soon. Ihave been swamped at work, so I don't have the bank account set up yet, but I'm not far from that. rich-c: james, you still around? how was teh trip? rich-c: for other nights, does Whitby have a Mardarin yet? Daniel B: well, if i don't eat anything i suppose i can go to adamcon this year with your help. Dale: There is one just slightly in Oshawa. About 10-15 min drive maybe. rich-c: bring a big bag of sandwiches, Daniel ; - ) Dale: There is a closer chinese buffet called Buffet World about 5 min away. rich-c: hey, if we can get teh gang there, they'll love it Guy B.: Ok. sounds good. Well, folks got to run. I'll see you all next week. rich-c: Dr. D. still hasn't got over his first experience with them rich-c: see you Guy, take care Dale: Daniel, I might be able to get you a ride with my brother Neil who is in Ottawa. Guy B.: Poof
Guy B. left chat session rich-c: Mandarin I know and trust - and my truck will hold five, eight if I put teh rear seat back in Dale: Maybe meet in Montreal or something? rich-c: can you make it down to Montreal, Daniel? Daniel B: i don't know. actually notbut maybe if i can convince someone. rich-c: well you do have a couple of months to find a lift down from St. Foy Dale: I'll keep thinking about it then. Daniel B: we will need to talk about it again and again. rich-c: well, the chat is every Wednesday night, so if Dale is on you should have lots of opportunity Dale: So each name table entry represents one block? rich-c: you two have some techie stuff going? Daniel B: one block of 2 bytes yes Dale: And each pattern table entry (8 bytes long) and hold 4 blocks, right? Daniel B: and the rle compression is done one bytes Daniel B: on bytes Daniel B: 4 blocks right rich-c: it's my time to sign out - good evening, gentlemen Daniel B: so there are 6 columns of 4-blocks in my example Dale: And it just depends what name table line you're on to decide which of the 4 blocks are displayed. Daniel B: that can be compressed with rle rich-c: color me gone
rich-c left chat session Daniel B: the name table will be compressed too Daniel B: but done based on the pattern table and the picture we want Dale: Rich, sorry, we are thrashing through one of Daniel's compression problems. Dale: You're saving the name table? Daniel B: yes Dale: Wouldn't it be better to generate the name table algorithmically? Dale: And used the same one for every picture? Daniel B: to avoid all problems, yes. but i want comething very optimized Daniel B: and I need to do this way because of my goal of doing a colorfull and nice border for my 3d game porject. Daniel B: the algorithm way is already done in my screen mode 3 editor sofware Dale: I would argue that a generated name table + RLE compressed pattern table would be smaller than a compressed pattern + compressed name. Daniel B: but coding 192 patterns it's huge Daniel B: based on my 3d game project, i need to produce a smaller pattern first, and optimized for the rle compression. Dale: I have used a litteral run, repeated run, and sequentally increasing run type algoritm in my RLE compression before. Dale: I think you'll loose the repeated pattern by scrambling the name table. Dale: Baseed on your example image anyway. Daniel B: and if you understand my example, all we need to do is to align blocks of 2-bytes in a way that the rle compression on bytes should be maximum Dale: Is it representative? Daniel B: the name table for my 3d project will necessary be modified to be able to see the 3d frames inside of it. Daniel B: the border i want to draw doesn't look exactly like the example. Daniel B: but it's close too Dale: I recommend that you organize the pattern entries vertically then. Dale: So all of the pattern entries from 0 to 5 will be shown in column 0 of the name table. Dale: IF that isn't pleasing, then sort all of the 8 byte pattern table entries, eliminate duplicates, and generate an RLE compressed name table. Dale: See if it is smaller or bigger. I think it will be bigger. Dale: Does it ring true to you? Daniel B: i already programmed the blocks collector to avoid duplicated patterns. Daniel B: now i need to re-organize them to align the bytes to be rle compressed. Dale: Well, that is most of the entropy gone. That'll be difficult to compress as a result. Daniel B: and this is my problem Dale: It is usually more efficient to RLE compress pattern table entries that don't have that preprocessing step done to them already. Dale: I think that that step is a mistake. Dale: you're probably missing out on an extra 5% of compression because of your method. Daniel B: for my project, it's a necessary step because the 3d frames to be showed in the screen need a huge part in the pattern table. Dale: Plus you are stuck with quoting a name table. Daniel B: and it need to be an array of consecutive bytes Dale: I don't understand. Mode 3 is "APA" (all points addressable) -- any color in any big pixel. Daniel B: ok, refer to my pdf... Daniel B: see all the char #3 in the name table? Dale: Either there's something you're not telling me, or the is an important fact that I don't understand. Dale: 77,11,77,33,33,33,33,33? Daniel B: the 3s characters in the name table designed the black zone in the picture Dale: Sure. Daniel B: so i change them by character numbers (ex. 30-80) based on the algorithm way you talked about Daniel B: to be able to show the 3d frames Dale: Isn;t it better to leave the name table fixed, and change the pattern table in this mode? Dale: Less overhead, easier calculations? Dale: Just twice the VDP bandwidth? Daniel B: nah, because it will need to consider which pattern can be modified for the 3d and this it's fast enough to reach the frame rate i want Daniel B: having consecutive patterns to be the 3d frame , it's better Daniel B: and the frames i'm drawing are half-frames Daniel B: sometimes, the left part is a exact mirror of the right part. Dale: Maybe I'm just tired, but I still think you're asking the wrong question. Daniel B: did you remember the 3d cave html file i sent to you in the past months? Dale: Yes. Dale: Are you rendering it realtime? Daniel B: maybe you don't have the version 2 with the half-frame specification Dale: So it is all prerendered? Daniel B: all the half-frames are encoded 2bits per pixel to use only 4 colors Dale: Version 2? I don't remember. Dale: Ah yes. Dale: I remember. Daniel B: and to avoid coding the mirrored part, and also a specific code to read the data in rom in a reversed way to show the second half frame, the names on screen are also reversed in their order Dale: In that case it is starting to sound sensible to have an algorithmically generated pattern table and store only the name table. Daniel B: it's hard to explain Daniel B: the rom will have plenty of (part of) pattern for the 3d half-frames. Dale: Why not do that? In 4 colours, with 2 bits per pixel, every possible 2x2 block is representable in 256 pattern table entries. Daniel B: the 3d frames will be encoded 2bits per huge pixel on screen and i use also a color table for the coloring part. Daniel B: so, a block of 4 big pixels will be encoded in one byte... but for the frames, not for the border. Daniel B: they are 2 seperate things Dale: Because they use different colour palettes? Daniel B: exactly, the border can use all the 15 colors, not the frames to be showed. Daniel B: i didn't draw yet the border i want because i wanted to fix my optimisation problem first Daniel B: anyway, thank you for your help! and good night! Dale: Should I work on an example of what I was thinking of? Dale: Or did I give you any ideas? Dale: anyway good night. Daniel B: well, i wanted a way to to the pattern optimisation and i think i will simply search for block of 2 similar bytes and align them whatever if they are 4,3,2 or only 1 in the pattern table. Daniel B: it's not the best, but it will works Dale: That;s why I suggested the sort. Daniel B: a sort on what? Dale: A sort will align long runs fairly well. Daniel B: it's not easy to sort the pattern and then have the similar bytes aligned Dale: The 2 or 8 byte pattern table entries. Dale: The longest run you can get is 3 bytes, right? Daniel B: 3 bytes can be compressed yes by saying "3 times this byte" Dale: Or maybe 4 bytes (2nd byte of first entry, 2 identical bytes of middle entry, 1st byte of later entry) Dale: There will be no runs of more than 4 identical bytes. Daniel B: based on my example, there are 5 or 6 columns of 8 bytes Dale: That entropy was eliminated in your first pass thorough. Daniel B: 8 identical bytes Daniel B: and the sort is done on 2-bytes, not on bytes. Dale: 33,33 then 33,33 then 33,33 then 33,33? Daniel B: yes, but it's rle compressed into 8 times 3 Daniel B: 8 times 33 Dale: 8 times 33? Daniel B: sorry Dale: I hadn't factored that into the theoretical maximum. Daniel B: But how we can acheive this sort that works on 2-bytes to align bytes for the rle compression? Dale: So you could have a pattern ending with: 77,33 then one with 8x33 then 33,77 for example Daniel B: impossible Dale: So the theoretical maximum run length is actually 10. Dale: What's impossible? Dale: It is the city name alphabest game. Daniel B: because of the no duplication, the 8x33 code said there is 33,33 then 33,33 then 33,33 then 33,33 so no other 33 should be coded in the pattern Daniel B: err... no, you are right Dale: What about a block that is 77,33? Won't it have a seperate entry? Daniel B: i mixed bytes and 2-bytes Daniel B: arg... this problem makes me crazy Dale: Okay, then we are back to sorting each of the 4 2-byte entries. Daniel B: ok, if you have 77,33 then one with 8x33 then 33,77 ... the rle compression should be 77, 10x33 then 77 again Dale: Then use the alphabet city game to order them. Dale: ie.: try and find a pattern that starts with the same byte that the current pattern ends with. Dale: Toronto Ottawa Ashburn North York Knowlton, ... Daniel B: that makes a lot of possible paths Dale: Sorting the two byte entries should maximize the run length in a single 8 byte pattern. james: sorry guys, i was lurking and following your conversation a bit, but i have to go Daniel B: bye james james: i'll try to be on more next week james: good to talk to you boht james: *both Dale: Then alphabet game will seek to maximize the intrapattern run length.
james left chat session Dale: Bye James. Daniel B: i will have to find information about the alphabet game you are talking about... i forget this theory after the final exam. Dale: There's a theory like that? Dale: It is a game we play in the car to keep from falling asleep while driving. Daniel B: i never heard about it, but i remember a project where i tried to find the optimized path between nodes... but i think it doesn't help Dale: C&O. Dale: Yes, that is the right field of expertise. Dale: You're goal is to grow a smallest compression organization. Dale: That's a least path problem for sure. Daniel B: oh well, my nightmare beggin Dale: The trick is to use Dynamically programming, etc so that you don't get the busy beaver function. Daniel B: how can i program this thing again based on my problem. shit Daniel B: i can't remember a thing about what i have done for the path problem Daniel B: i think i never programed it, but simply apply the theory in an exam. Daniel B: it was about internet Dale: What I recommend is that you worry about sorting the 4 sections of blocks, and then take the patterns and find the max run, then leave the rest in any order. Daniel B: iso model, etc.. Dale: That is at least computable. Dale: Computable is good. Dale: Non-computable is the busy beaver function. Daniel B: looks easy ... but it is not Daniel B: i think i will need algorith or pseudo-code Daniel B: to start Dale: I'll think about it this weekend. Daniel B: thanks Dale: I should go. Daniel B: me too Daniel B: good night! Dale: Good night.
Dale left chat session Daniel B: take care *poof*
Daniel B left chat session
moved to room Meeting Place
changed username to <undefined>
<undefined> left chat session