| |
| |
| |
|
Page: 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21
Comments:
<0> a client that wants C for their cgis? I think i like this guy.
<1> i don't want to write a template engine and compile it in, if i can find one readily available.
<2> i want to make money frmo home, w/o doing anything
<1> ornge: porn. it has to be fast.
<1> it's a traffic app
<0> feti_, awesome.
<3> haha
<3> feti_, nice. :-)
<1> PHP's not even an option at this point ;D
<0> i do all my webstuff in cgi and people pick on me like "why don't you use php? or ruby on rails? etc."
<1> plus, even though i love php, it just isn't suitable for this.
<1> heh yeah
<0> apple WebObjects is rather nice.
<1> i'm porting my php framework to this now.
<3> OrageTide, same here, I usually stick to C or even sh before going with PHP/Python or similar.
<0> it used to cost money, i think it's free now.
<1> it's going pretty well, with the exception of config parsing and templating issue
<0> sbahra, basically i already know C and am very good with it. why use something else?
<1> ease i think for some things
<1> for a complex app, php makes stuff fast. but it's like any argument. you sacrifice so much for it
<0> oh no. you can't do maps in C. well i have code laying in my toolbox to do that. so it's no big deal
<3> OrngeTide, because other languages can be more productive for specific situations, no matter how much you know C. :-P
<1> if you have other devs that don't do C, you run into an issue using C
<0> sbahra, it's not more productive for me to learn a new language just to do a handful of webpages though:)
<0> if i were doing them for a living that would be totally different.
<3> OrngeTide, indeed.
<0> i met a guy who did his wiki in gforth.
<3> I needed an excuse to learn Python though, next project will be completely Python (pretty big system too)
<3> OrngeTide, heh
<3> OrngeTide, a friend of mine is running his website with his own Haskell webserver. ;-)
<0> nice
<3> He has the code up somewhere if you want a look.
<0> i was fiddling around with doing some stuff in lua. i like lua since it integrates with C so easily.
<0> nah. i totally don't get haskell
<4> mornings
<3> OrngeTide, haven't worked with lua, BSDinstaller has been making use of it for a while now though.
<0> lua is pretty simple and has a pretty nice syntax to it. i use to be a tcl fan, but lua has completely replaced it for me. (since it's faster, and has types and is easier to extend)
<1> OrngeTide: do you intertwine your HTML into your C code directly or do you parse the HTML's and store them into memory and do your variable replacements in real-time?
<1> cause i need a tiny templating engine to do all that, that supports conditions/loops
<0> feti_, i interwine the HTML and use CSS to change the layout.
<1> hrm ;( darn
<1> that's an issue
<3> feti_, well, *usually* you can just have a "top" and "bottom" template, open and print, but that is slow.
<1> guess it's back to the drawing board for me.
<0> it's faster to use CSS and more flexible. you make the client do all the heavy lifting instead of having to load and parse a template everytime you load a page.
<1> right, i'm aware of all that.
<1> but i'm talking about dynamic pages and replacement of variables inside the HTML
<3> feti_, what I am doing for this project, since it's in Python, I don't want to deal with the overhead of system call switching (remember, a majority of data is cached), I will have a C daemon that will cache all appropriate templates/etc...
<0> i have a config system where i can insert a string into the header. this holds any extra stuff i want (like favicon, extra stylesheets, etC)
<3> feti_, yes, in that case, I just do a keyword search (strstr) and then on that point, output the replacement
<1> hrm
<0> i wrote a thing that replaced $foo$ with things. i had some hacks where you could repeat sections of html so you could generate tables and lists.
<0> but i never use it anymore, now that CSS is mature.
<1> well the thing is i want this to function more like PHP style, very basic though, but maybe compile the templates to C and anything inside <? or ?> would be C code, and anything outside would be HTML
<1> that's not asking for too much
<1> ;P
<0> i was thinking of doing <? lua code ?>
<0> you could write a parser that you feed it an html with <? c code ?> and it would output a C program
<1> so generate C from the templates in real-time (when change occurs) and compile it in raltime
<2> feti_: you'd want *interpreted* lang intermingled with html, not *compiled* language like C
<1> the only worry i have with compiling in real-time is if you're in a chroot, you obviously do't want GCC in tehre
<0> kind of like lexx/yacc/rpcgen and stuff that generate C from an input file
<1> so that'd be impossible
<5> hey OrngeTide :)
<1> weird: that'd work too even.
<0> feti_, you could put tcc in there and have it only compile .so files to load.
<0> Gambit-, hio
<1> tht's true
<2> or use cint
<5> OrngeTide, did you have any other comments re the schema? Do you think it'll support everything that might come up?
<1> cint?
<2> c/c+ intetrpreter
<0> Gambit-, it doesn't support everything that might come up.
<5> OrngeTide, Details details :)
<1> weird: how do you hook into the main C app with that so you can use stuff in it
<0> Gambit-, but i don't know what else might come up. so i don't have any recommendations
<1> like a custom function
<0> Gambit-, but i can promise you that you will change it at least once.
<5> OrngeTide, Well you've got a lot more mud experience then me, that's why I'm trying to get your perspective :)
<0> the reason being: you haven't written out your requirements.
<0> writing a schema without having hard set requirements is hard and you have to be willing to change the schema as your non-concrete requirements change
<0> Gambit-, i'm not sure i understood your schema entirely. but it covered a bunch of things i didn't think of. so that's good.
<5> OrageTide, s'true; really I don't know enough about the problem sphere to readily identify any gaps...
<5> OrngeTide, basically it lets columns be created using generic names that are consistent between multiple tables so when you request a property of a given item you can ask for the same property name -- i.e. the same field name -- for any given item.
<0> so this guy is actually going to pay you money to write a mud?
<5> that's basically it
<5> Not as such
<5> but the projects gotten my interest
<5> plus it gives me a good excuse to get out the mana-redux codebase and put some time and polish into it
<5> maybe turn it into something real.
<0> Gambit-, ehhe. fun
<0> i wish i could work on it too
<5> Yeah, basically :)
<0> btw. "THAC0" is obsolete.
<5> Well that's why I'm h***ling you for creative criticism :P
<5> that's ok, I never understood it anyways :P
<0> to hit armor cl*** 0
<5> Yeah, so you have to roll a 5 to hit armor cl*** 0, if you don't do you not do any damage at all?
<1> WeirdMish: this may be a stupid question, but can cint compile the C script into bytecode stored into memory so it's only compiled on startup?
<1> i've never used cint. don't bash me! ;D
<1> i'm reading through some pages on it now though, trying to see how it works exactly
<2> feti_: i think so
<0> Gambit-, right. you miss
<0> Gambit-, or you just bounce off their tough armor.
<2> "Castrated Calif. molester seeks freedom ". District Attorney Tony Rackauckas says Reilly remains a ***ual predator who should remain incarcerated.
<5> ahh eheh that ****s :)
<0> lets say that you are attacking a dragon with a sausage. you have a THAC0 of 16 and the dragon has an AC of 10. you'd need to roll above a 26 on a 20-sided die to hit the dragon. :)
<2> Can I attack Queen with a sausage ?
<5> well anyways, can you give me any examples of things the schema might not work? Think of all the things you've ever seen in a game.
<0> WeirdMish, yea. but a queen would have an AC of 90 in that case. unless she's a slut.
<5> s/in a game/in a mud/
<5> I'm mostly looking for examples to stretch my perspectives, ne?
<4> mud whoo
<0> Gambit-, well one thing i don't like is that you shouldn't be dynamically creating tables.
<0> i don't understand long_sword_item_table
<5> The dynamic table creation isn't in-game, it's by-admin
<0> yea. i don't think admins should be creating new tables.
<5> Right, so the "long sword" is a category, and all objects that belong to the category of "long swords" will have a record in that table containing those stats.
<0> why the extra level of indirection?
<0> just have table objects
<6> ya realy
<0> which has everything, items, rooms, whatever.
<5> because different categories have different attributes
<6> with a type field and have 'long sword' as a type
<0> then look up by a guid. if you use a key it will be fast since it's just a b+tree search
<0> Gambit-, i would use dynamic attributes.
<0> table attributes would have the attributes for every object, indexed by the object id
<5> sure a linear table search is ultra fast, that's not the point.. the category split lets objects belong to multiple categories and derive characteristics or scripts from each one.
<0> then you would look up all the attributes for an object in that table and get a list of results
<5> oh so you have an attributes table with: obj_id : attrib_id : atrib_value?
<0> Gambit-, i think you're trying to implement too high level of a concept ontop of SQL.
<0> yes.
<6> OrngeTide: i don't understand why you would do it any other way, isn't that the whole point of an rdbms in this kind of situation? :P
<0> and probably some versioning information.
<4> Gambit-: You're going to make a SQL backended MUD?
<0> all the attribute that are common to every object would just be in the object table. (like name, owner, creation date, etc)
<5> zid_, yeah,
<4> sounds cool
<5> zid_, yup
<4> I'm a mudder if you ever get it loginable
<4> **** I left my drink downstairs, brb
<5> OrngeTide, Yup... 'course it's faster to pull single records instead of a bunch of records.
<0> you could break it up into rooms, characters, items, users, npc_controllers, etc.
<5> break it up into seperate tables?
<4> I would
<5> so you have rooms_attribs?
<0> rather than just making everything an object. but then a world-wide id isn't really useful anymore, and you'd just use a type-based id like most muds (diku, rom, etc)
<5> but basically you're approaching back what I'm thinking right now.
<4> whoo ROM :P
<4> I used to tinker with that
Return to
#c
or
Go to some related
logs:
esx hds vmotion
#beginner
#gamedev
#computers
Gauntanamo bay prisons shut down
tropical rianforests
thenaughtyamericahouse.com pass
#nhl
#politics
#solaris
|
|