[QUOTE=Hexxeh;34413758]All these Facepunch scraping apps are cool, but there's still the problem that you're downloading way more data than needed.
Would a third-party server that scraped for you and which Facepunch server trusted as a third party (passing the real IP address through) not be the best solution?[/QUOTE]
In theory yes, but that would mean garry would have to support your app and the users would have to trust your server.
[B]Edit:[/B]
And higher cost.
[QUOTE=Robber;34413768]In theory yes, but that would mean garry would have to support your app and the users would have to trust your server.
[B]Edit:[/B]
And higher cost.[/QUOTE]
If someone, such as myself, is happy to provide it as a service to application developers publishing free apps at no cost to the developer, and I can get garry to agree to it, what's the problem? I'm sure garry would like good quality apps for Facepunch as much as the rest of us. Indeed he's used the old FPAPI one before I believe.
The advantage to a centralised scraping API that's available for public use is that garry only has to work with a single third-party.
I should mention that I have /plenty/ of capacity to provide this service. If it helped, I could also distribute servers to allow for edge-nodes closer to users to provide a faster response, but given the nature of the application, this would provide only minimal performance improvement.
Added some acid :3
[URL=http://imageshack.us/photo/my-images/401/acidf.png/][IMG]http://img401.imageshack.us/img401/8097/acidf.png[/IMG][/URL]
[QUOTE=Sebixxx;34413892]Added some acid :3
[URL=http://imageshack.us/photo/my-images/401/acidf.png/][IMG]http://img401.imageshack.us/img401/8097/acidf.png[/IMG][/URL][/QUOTE]
Needs moar Vignetting :P
Why did Garry remove the old API anyway?
Was it just Compwhizi's 'relaxed' attitude towards security, or something else?
I didn't want to maintain and secure it. And I don't want to be locked to vbulletin. So you're best off just scraping for now.
Okay, set myself a challenge to get scraping for the forum list, thread list and pages of a thread working by the end of the day. Started an hour ago, I've finished the forum list and I'm about half way through the thread list.
Output is in JSON currently, but it'd be easy to make it support XML too.
[video=youtube;SGSxnCrVA94]http://www.youtube.com/watch?v=SGSxnCrVA94&feature=youtu.be[/video]
did some flocking such that they won't clump as much as before
and implemented PBAE skills
[QUOTE=garry;34414427]I didn't want to maintain and secure it. And I don't want to be locked to vbulletin. So you're best off just scraping for now.[/QUOTE]
Would you be interested in outputting a more scrapable version? I've not got any experience in vb themes, but surely you could make one that only outputs pertinent data.
[QUOTE=Lexic;34414500]Would you be interested in outputting a more scrapable version? I've not got any experience in vb themes, but surely you could make one that only outputs pertinent data.[/QUOTE]
Scraping isn't the problem, the FP layout is easily scrapable. The problem is accepting request via a third party proxy.
Edit: Finished scraping the thread list.
I pulled an all nighter and now have my very own (possibly)reliable UDP network layer. I don't yet know how stable it is or if I might be leaking tons of memory but from my limited on the spot testing it seems stable.
The reliability is very simple and straight forward right now... I just record all packets(including a copy of the sent data and a unique packet id) that need acknowledgement before sending them. Then I wait a full RTT + RTT/2 before sending them again if the acknowledgement hasn't yet been received. The wait is for network anomalies that might be present.
Once an acknowledgement has been received for that packet, I run through the list and delete the matching packet.
I should really read some network books so I can get a better grasp of what the right way of doing this would be. I'm pleased with what I've done so far, though. Hopefully I will get some reusable code out of this.
[QUOTE=false prophet;34414897]I pulled an all nighter and now have my very own (possibly)reliable UDP network layer. I don't yet know how stable it is or if I might be leaking tons of memory but from my limited on the spot testing it seems stable.
The reliability is very simple and straight forward right now... I just record all packets(including a copy of the sent data and a unique packet id) that need acknowledgement before sending them. Then I wait a full RTT + RTT/2 before sending them again if the acknowledgement hasn't yet been received. The wait is for network anomalies that might be present.
Once an acknowledgement has been received for that packet, I run through the list and delete the matching packet.
I should really read some network books so I can get a better grasp of what the right way of doing this would be. I'm pleased with what I've done so far, though. Hopefully I will get some reusable code out of this.[/QUOTE]
Does this let you say that some packets absolutely must reach the other end while others are allowed to be dropped?
[QUOTE=swift and shift;34414966]Does this let you say that some packets absolutely must reach the other end while others are allowed to be dropped?[/QUOTE]
yes
[release]Python 2.7.2 (default, Jan 26 2012, 02:32:59)
[GCC 3.4.4 (msys special)] on msys
Type "help", "copyright", "credits" or "license" for more information.
>>>
[/release]
[QUOTE=voodooattack;34414994][release]Python 2.7.2 (default, Jan 26 2012, 02:32:59)
[GCC 3.4.4 (msys special)] on msys
Type "help", "copyright", "credits" or "license" for more information.
>>>
[/release][/QUOTE]
Why would you post this?
[QUOTE=voodooattack;34414994][release]Python 2.7.2 (default, Jan 26 2012, 02:32:59)
[GCC 3.4.4 (msys special)] on msys
Type "help", "copyright", "credits" or "license" for more information.
>>>
[/release][/QUOTE]
credits
[editline]27th January 2012[/editline]
it's not working
Try help
[QUOTE=voodooattack;34414994][release]Python 2.7.2 (default, Jan 26 2012, 02:32:59)
[GCC 3.4.4 (msys special)] on msys
Type "help", "copyright", "credits" or "license" for more information.
>>>
[/release][/QUOTE]
oh man no way me too
[release]Python 3.2.2 (default, Nov 21 2011, 16:51:01)
[GCC 4.6.2] on linux2
Type "help", "copyright", "credits" or "license" for more information.
>>>
[/release]
from __future__ import braces
[editline]27th January 2012[/editline]
that's p much my entire knowledge of python
[QUOTE=BlkDucky;34415466]from __future__ import braces
[editline]27th January 2012[/editline]
that's p much my entire knowledge of python[/QUOTE]
Try "import antigravity" (no seriously, try it)
For those of you working on the Steam web presence API, here's a useful idea: use it to facilitate chat logging.
(So people leave me alone about updating my old chat logger. fml)
[QUOTE=Dlaor-guy;34415313]Why would you post this?[/QUOTE]
I just built python for msys. Which was a feat comparable to building the latest perl. Considering that I had to edit nearly every file and had to build tcl/tk to get it to work.
I'm currently in the process of packaging it with mgwport, I'll diff the original sources against mine, write a script to build it from scratch then wrap it up and post it to the mingw/msys mailing list in a day or two, I think some of the folks there are going to be ecstatic about this.
In case you're wondering, this is quite important as the current python builds for windows can't handle msys paths and tend to break builds for pretty much anything dependent on a python script.
Also, the python devs are somewhat smug and refuse to code/accept patches to support 'minor platforms', they barely accepted the ones for cygwin support.. figures..
[editline]27th January 2012[/editline]
Also, emphasis: [B][I][U][GCC 3.4.4 (msys special)] on msys[/U][/I][/B]
[editline]27th January 2012[/editline]
[QUOTE=Chandler;34415510]Try "import antigravity" (no seriously, try it)[/QUOTE]
[code]import webbrowser
webbrowser.open("http://xkcd.com/353/")[/code]
That's actually in the 'Lib' folder.. wow.
Finished scraping for forum view, thread view and post views. Adding posting and rating now.
Realised I had the normals backwards for 3D explicit plots in x and z. Whoops :v:
[url=http://postimage.org/image/lym8mttzb/][img]http://i42.tinypic.com/11b3toi.png[/img][/url]
Garrisoning units works now for civilian buildings, as well as collision detection
[url=http://postimage.org/image/fz32wwykn/][img]http://i43.tinypic.com/xanxqq.png[/img][/url]
Notice the 3D jet fighters. The infantry actually spawned inside the trees.
[url=http://postimage.org/image/8x55apuyv/][img]http://i40.tinypic.com/352ji1l.png[/img][/url]
There's not a lot really out there...
So far the tiles are loaded from a bitmap file, displayed there on the working minimap. Next thing I'm doing is adding spawn scripts to fill the map, then maybe some fog of war or lighting, garrison-able ships, etc...
Destructible terrain will be a must, eventually i'll get down to dialog and objectives
I made an applet for a friend to upload and download files from a MySql database.
[img]http://dl.dropbox.com/u/99717/applet.png[/img]
The download bar to the right works only for downloading though :v:
GranPC is a collaborator for the [url=https://github.com/andrewmcwatters/steamweb-lua-sdk]steamweb-lua-sdk[/url] now, so it looks like we'll see the same turn out time we did for the FPSDK.
I'll be starting work on this today, and if there are any individuals who would like to join our team, yell at us and we'll discuss it.
Well, fuck, I should've seen this coming, but it turns out you can't use the API without owning the app.
The thing is, for us who do have beta access, even once we're done with reversing everything, nothing should change, so everything will be usable post-beta. But I'm telling you what you already know, so.
[img]http://i.imgur.com/SK0zf.png[/img]
Doing some shit with the WoW api.
Sorry, you need to Log In to post a reply to this thread.
