Difference between revisions of "BattleTechWiki talk:Project Planets"

Line 2: Line 2:
 
*Okay, Leadman, I've uploaded Skuggi's member userbar suggestion onto the Project page. What do you think? --[[User:Revanche|Revanche (admin)]] 18:35, 22 November 2006 (CST)
 
*Okay, Leadman, I've uploaded Skuggi's member userbar suggestion onto the Project page. What do you think? --[[User:Revanche|Revanche (admin)]] 18:35, 22 November 2006 (CST)
  
 
+
==PHP Generated Star Systems?==
== PHP Generated Star Systems? ==
 
 
 
 
I've noticed that hundreds of planets are not listed with coordinates or surrounding systems much from the periphery and clan space. Though I'm sure I can find them through ISCS and Source Books, and probably has already been suggested and implemented in someway, should a PHP image output complete with an image-map linked to respective planet wiki pages be created for each blank wiki page? At least not server-generated images but generated images and uploaded for those wiki-lost systems. I can contribute to the programming/generation process and then upload outputs for every system not listed.
 
I've noticed that hundreds of planets are not listed with coordinates or surrounding systems much from the periphery and clan space. Though I'm sure I can find them through ISCS and Source Books, and probably has already been suggested and implemented in someway, should a PHP image output complete with an image-map linked to respective planet wiki pages be created for each blank wiki page? At least not server-generated images but generated images and uploaded for those wiki-lost systems. I can contribute to the programming/generation process and then upload outputs for every system not listed.
  
Also, the 60-lightyear zoom can be confusing as people (including me :p) may not know where in relative to the entire Inner Sphere the region may be. Some current X, Y coords are still problematic by the way. --[[User:Dracoix|Dracoix]] 04:48, 27 July 2009 (UTC)
+
Also, the 60-light year zoom can be confusing as people (including me :p) may not know where in relative to the entire Inner Sphere the region may be. Some current X, Y coords are still problematic by the way. --[[User:Dracoix|Dracoix]] 04:48, 27 July 2009 (UTC)
  
 
:Dracoix, I know there is indeed issues with the planetary DB and interest in both correcting and improving how they are handled, but I am so not the guy ready to handle this. I'm going to ask [[User:Nicjansma|Nicjansma]] (the site's founder) and [[User:Frabby|Frabby]] (probably our most dedicated planetologist) , as well as two of our code-crunching admins to weigh in over here. Please check back, because I think you may be on to something. --[[User:Revanche|Revanche]] <sup>([[User_talk:Revanche|talk]]|[[Special:Contributions/Revanche|contribs]])</sup> 11:31, 27 July 2009 (UTC)
 
:Dracoix, I know there is indeed issues with the planetary DB and interest in both correcting and improving how they are handled, but I am so not the guy ready to handle this. I'm going to ask [[User:Nicjansma|Nicjansma]] (the site's founder) and [[User:Frabby|Frabby]] (probably our most dedicated planetologist) , as well as two of our code-crunching admins to weigh in over here. Please check back, because I think you may be on to something. --[[User:Revanche|Revanche]] <sup>([[User_talk:Revanche|talk]]|[[Special:Contributions/Revanche|contribs]])</sup> 11:31, 27 July 2009 (UTC)
Line 70: Line 68:
  
 
:::::::::But that's the point Nic: It's ''not'' essentially static. It should be, but it plainly isn't. The data will require corrections and additions again and again, which is why I think we should implement something that can easily be modified with updated data. [[User:Frabby|Frabby]] 09:29, 29 July 2009 (UTC)
 
:::::::::But that's the point Nic: It's ''not'' essentially static. It should be, but it plainly isn't. The data will require corrections and additions again and again, which is why I think we should implement something that can easily be modified with updated data. [[User:Frabby|Frabby]] 09:29, 29 July 2009 (UTC)
 +
::::::::::If we're going to go this route, why not just go all the way? Have Sarna simply switch to auto-generating nearly all its content on the fly. There are relatively few things that can't be auto-generated; all of it "flavour". Anything with statistics behind it can be accommodated through automation. Note that this would require partially forking Sarna's MediaWiki installation. --[[User:Xoid|Xoid]] 05:17, 3 August 2009 (UTC)
  
== PHP Generator Part 2 (Sandboxing) ==
+
==PHP Generator Part 2 (Sandboxing)==
 
 
 
(New section on regards to sandboxing drafts.)
 
(New section on regards to sandboxing drafts.)
 
So we know that the systems are in disarray, and the original data is still useful, and some of the planets may need to be corrected and re-regenerated down the line. I propose an off-wiki sandbox on another site. That way the wiki will not be affected until the draft template and generated images are finalized. First things first, we need an image template basis. I can work with almost anything image related in PHP (job related story for another day.)  
 
So we know that the systems are in disarray, and the original data is still useful, and some of the planets may need to be corrected and re-regenerated down the line. I propose an off-wiki sandbox on another site. That way the wiki will not be affected until the draft template and generated images are finalized. First things first, we need an image template basis. I can work with almost anything image related in PHP (job related story for another day.)  

Revision as of 01:17, 3 August 2009

Member Userbar

  • Okay, Leadman, I've uploaded Skuggi's member userbar suggestion onto the Project page. What do you think? --Revanche (admin) 18:35, 22 November 2006 (CST)

PHP Generated Star Systems?

I've noticed that hundreds of planets are not listed with coordinates or surrounding systems much from the periphery and clan space. Though I'm sure I can find them through ISCS and Source Books, and probably has already been suggested and implemented in someway, should a PHP image output complete with an image-map linked to respective planet wiki pages be created for each blank wiki page? At least not server-generated images but generated images and uploaded for those wiki-lost systems. I can contribute to the programming/generation process and then upload outputs for every system not listed.

Also, the 60-light year zoom can be confusing as people (including me :p) may not know where in relative to the entire Inner Sphere the region may be. Some current X, Y coords are still problematic by the way. --Dracoix 04:48, 27 July 2009 (UTC)

Dracoix, I know there is indeed issues with the planetary DB and interest in both correcting and improving how they are handled, but I am so not the guy ready to handle this. I'm going to ask Nicjansma (the site's founder) and Frabby (probably our most dedicated planetologist) , as well as two of our code-crunching admins to weigh in over here. Please check back, because I think you may be on to something. --Revanche (talk|contribs) 11:31, 27 July 2009 (UTC)
See User_talk:Nicjansma#Other_extension_suggestions :)
The datafile on which this wiki's system data is based was an older version of the ISCS atlas and contains numerous factual errors, omissions and typos. Plus, the bot used failed to transcribe the XY coordinates correctly. Bits and pieces of this discussion are unfortunately spread across many talk pages on BTW, but essentially we need to redo the entire underlying database (that includes re-checking all XY coords where we have them, and provide approximations for where not - and then there's my pet project about unmapped worlds where we don't have as much as a clue to where they are). Once this is done, we need to agree on a proper format to present systems because the current system mixes up system names and planet names. And finally some code wizard would be required to create clickable maps from that database, with the option to alter the maps as new info becomes available. I can help collecting the data but I am totally helpless when it comes to coding. Frabby 10:56, 28 July 2009 (UTC)
I actually specialize in image manipulation and algorithms (watermarks, pixel functions, etc) what shall be thy bidding? >:) Except going through every source book and re-entering systems and planets into a new 99.9% error free database would be a pain. I'll get cracking on a star map tool with labeled image output (as well as CSV output) offline. It can then probably be ported into PHP. However, any image function in PHP requires GD2, server CPU usage, and bandwidth to send the generated image to the user. And yes, I'm just a guy that came out of no where, but I think I can help you guys out on this issue.
Also how accurate were the original 2003 ISCS files? --Dracoix 23:41, 28 July 2009 (UTC)
They were mostly accurate. I'd say 95% or so. I corrected a few, but not much work was needed. That said, there were probably errors that I did not catch. On a related note, might I suggest moving this discussion to BattleTechWiki talk: Project Planets? --Scaletail 01:15, 29 July 2009 (UTC)
Suggestion sensible; move made. --Revanche (talk|contribs) 02:16, 29 July 2009 (UTC)
Hey all. I believe I can offer a structure to handle all the data elements using Google Maps' new GMapType class. Unfortunately, I'm swamped with work right now and it will be a few days before I can demonstrate a prototype. If you can wait a little bit I can whip together something more concrete for planning purposes.
Draconix (and Nic), if we can get a data host and a sufficient server, we will need three db tables structured something like this (assuming it's MySQL, of course): Smiley.gif
CREATE TABLE system (
  system_id INT auto_increment,
  system_name VARCHAR(64) default 'name missing',
  system_x SMALLINT unsigned default 0,
  system_y SMALLINT unsigned default 0,
  pixel_x SMALLINT signed default 0,
  pixel_y SMALLINT signed default 0,
  control_id INT default 1,
  PRIMARY KEY (system_id),
  FOREIGN KEY (control_id) REFERENCES control(control_id)
) ENGINE=InnoDB

CREATE TABLE control (
  control_id INT auto_increment,
  date_3057 TINYINT default 1,
  date_3062 TINYINT default 1,
  PRIMARY KEY (control_id),
  FOREIGN_KEY (date_3057) REFERENCES sovereign(sovereign_id),
  FOREIGN_KEY (date_3062) REFERENCES sovereign(sovereign_id)
) ENGINE=InnoDB

CREATE TABLE sovereign (
  sovereign_id TINYINT auto_increment,
  sovereign_name VARCHAR(32) default 'name missing',
  PRIMARY KEY (sovereign_id)
) ENGINE=InnoDB
so we can run queries like this:
SELECT system_name, coords_x, coords_y, pixel_x, pixel_y, sovereign_name 
FROM system JOIN control USING (control_id) JOIN sovereign USING (sovereign_id) 
WHERE system_id=XX AND date_3062=YY LIMIT 1;
Another table might be needed for alternative names, coordinates, etc. The date columns would require a good deal of foresight, unless we approached them like date ranges. Hmmm. The more I think about it, the more I like that idea, but it would require more complex queries. Oh well, a subject for another time.
I hope this puts a bit more structure to the conversation. Thanks. --Ebakunin (talk|contribs) 01:22, 29 July 2009 (UTC)
I suggest not using googlemaps, good idea to keep it stupid simple? I can probably handle the database on my host probably (though it's a dedicated shared server account.) It's a firewalled SQL setup though, can't make external queries, but a PHP "GET" proxy can always bypass this. By the way, anyone have a completed CVS or other text listing all the systems that's up to date? --Dracoix 03:11, 29 July 2009 (UTC)
Inner Sphere Cartography Society Archive - fantastic resource, if somewhat outdated. Øystein Tvedten (the BT geographer) helped with the data. --Ebakunin (talk|contribs) 03:27, 29 July 2009 (UTC)
I already have those files, I was referring to the corrected ones. Funny you should mention about Oyst, I actually asked him about terrain for planets (to 3D model) a year back or so. Long story for another day. --Dracoix 03:35, 29 July 2009 (UTC)
Original ISATLAS of ISCS I knew it existed somewhere. Can I use this collected data to build a new database? Or will there be a copyright issue? They seem to have the concept nailed down...Though it's lost to its own site map links. --Dracoix 03:53, 29 July 2009 (UTC)
The original 'Planets' pages were created from the ISCS data, specifically I think this file. Is there an updated version? It looks like ISCS's website has been updated recently to include a lot more info. I used the ISCS data file, in conjunction with a components database, and built a C# app to pre-build the 'wiki-code' for all of the planets. This code is located here. Ideally, I would prefer to have all pages and images pre-built instead of re-querying the database for data that we know is essentially static. Nicjansma 06:41, 29 July 2009 (UTC)
But that's the point Nic: It's not essentially static. It should be, but it plainly isn't. The data will require corrections and additions again and again, which is why I think we should implement something that can easily be modified with updated data. Frabby 09:29, 29 July 2009 (UTC)
If we're going to go this route, why not just go all the way? Have Sarna simply switch to auto-generating nearly all its content on the fly. There are relatively few things that can't be auto-generated; all of it "flavour". Anything with statistics behind it can be accommodated through automation. Note that this would require partially forking Sarna's MediaWiki installation. --Xoid 05:17, 3 August 2009 (UTC)

PHP Generator Part 2 (Sandboxing)

(New section on regards to sandboxing drafts.) So we know that the systems are in disarray, and the original data is still useful, and some of the planets may need to be corrected and re-regenerated down the line. I propose an off-wiki sandbox on another site. That way the wiki will not be affected until the draft template and generated images are finalized. First things first, we need an image template basis. I can work with almost anything image related in PHP (job related story for another day.)

I'll start coding the outputs at http://mindoftheq.net/btw/planetgen.php?&p=PLANETNAME&y=YEAR where PLANETNAME would be the case-insensitive planet/system name and YEAR would be the non-specific control year. I like to keep things idiot-proof and dynamically friendly :). I'll even provide the SQL builds for download.

In all the output would be like a completed wiki page complete with a generated image with image-map links. Input would be greatly appreciated regarding the image template. --Dracoix 16:05, 30 July 2009 (UTC)