 HurtnTime United States. May 04 2003 02:18. Posts 288 | Profile |
jca, i'm working with someone on BW rep file API and file manager. we like what you did with BWchart and would like to know if you want to work with us to have everything in one app.
the rep manager app we have is very simple at the moment, we plan to add searching capabilities, superview features, etc.
let me know if you interested.
if you want to see what we have so far:
http://www.battlereports.com/users/HurtnTime/bwrepmgr.zip
It has little functionality at the moment. You can browse your folders and if you select a BW rep file then it will display information about it such as the map, players, races, etc.
Let me know if you have any trouble with it. It has not been tested very well. |
|
|
 Smorrie Netherlands. May 04 2003 04:00. Posts 2501 | Profile Blog |
| Very cool and nice to see ppl doing stuff like this oO;~~ |
| | It has a strong technique, but it lacks oo. |
|
|
 AK-Nemesis May 04 2003 05:13. Posts 2005 | Profile |
OMG, tha's such a great program..
GREAT JOB!!! |
| | Nemesis has left the building~ |
|
|
 ky[Z] China. May 04 2003 07:32. Posts 1677 | Profile | |
| | Terran is SOoOo over-powered~!! Especially in TvT~! |
|
|
 jca France. May 04 2003 09:55. Posts 273 | Profile |
HurtnTime, I just tried bwrepmgr and it looks pretty good so far (C++ with MFC?)
bwrepmgr can already display the replay header so I assume that you know how to decompress the whole thing and get the list of actions, am i right?
|
| |
|
 Muhweli Finland. May 04 2003 10:21. Posts 5292 | Profile |
HurtnTime, I LUV U, Contact me at msn: muhweli@hotmail.com, icq: 18642988 or e-mail: muhweli@paivola.net
I need to ask a couple of questions concerning the technical structure of the program and about the program source code. (Will it be public by any part btw?) |
| |
|
| jca France. May 04 2003 12:19. Posts 273 | Profile | |
| |
|
 Saab Sweden. May 04 2003 15:14. Posts 28 | Profile |
Hey jca, do you think you can add a feature to the program that allows one to adjust the y-scale?
because often in the action/minute there are peaks which cause the 'autofit' (?) to make scales up to 2000 act/min, making it being hard to interpretate the graph
|
|
|
| imRadu Romania. May 04 2003 15:52. Posts 1775 | Profile Blog |
| saab i doubt you had a peak of 2000 because that means that you performed 33.3 actions in that second |
| | Its really good to see that some people dont let education get in the way of their ignorance |
|
|
 Pol Poland. May 04 2003 15:55. Posts 1186 | Profile |
thx very gosu  |
|
|
| Saab Sweden. May 04 2003 16:07. Posts 28 | Profile |
2000 is very high indeed, i cant find the replay where it was anymore, but i can find ones where it reaches up to 700
numbers can cheat in various ways |
|
|
| jca France. May 04 2003 16:55. Posts 273 | Profile |
| Indeed, in some reps the players seem to repeat some actions (probably from keyboard autorepeat) many times in a row, which results in a very high peak on the action/minute curve. I think I will set a maximum for that curve, maybe 500? |
| |
|
| HurtnTime United States. May 04 2003 17:28. Posts 288 | Profile |
My ICQ is: 1514534
MSN: jorgeRlopez@hotmail.com
jca, I did it in C++ and MFC using Visual Studio .NET. However, I expect it to work just fine under Visual Studio 6.0. In fact, I might start compilining it in VC 6.0 so I dont have to include all those DLLs with the application.
About the rep file itself. I received code that uncompresses it all and the header section is figured out pretty good (important info anyways). The map section is documented at another website since its just like a map file. The third section is all the user actions and I got some info from the developer of SuperView (William WC3), Fractaled, and Brefan so should be able to decode it in short order the next time I have some time available.
By the way, I read in some other threads some people concerned about Blizzard going after anyone who reads rep files. That is just non-sense. Rep files are user data. Any Word processor worth anything can import MS Word .DOC files and it is not a published format. User data is fair game. Another example is the reverse engineering of AutoCAD drawings by OpenDWG Alliance (www.opendwg.org). |
|
|
| Saab Sweden. May 04 2003 18:04. Posts 28 | Profile |
500 seems good yes...
by the way, isnt it possible to show current supply stock by simply counting everything that goes in and out? (as mentioned at your site)
but then as an option in the program |
|
|
| jca France. May 04 2003 18:15. Posts 273 | Profile |
| There is already a curve for supply. But it's not the "current supply". It's the "total supply" since beginning of game. This is because from the .rep we can tell when units are built, but not when they are killed. So we don't know when supply goes down. |
| |
|
 OrangeTerran United States. May 04 2003 18:35. Posts 963 | Profile | |
| |
|
| Saab Sweden. May 04 2003 19:37. Posts 28 | Profile |
| oh... also, often at the end of the game (or lategame) you see the gascurve shows during some intervals a straight line, asif nothing is being mined, do you know why? |
|
|
| Saab Sweden. May 04 2003 20:18. Posts 28 | Profile |
| oh... also, often at the end of the game (or lategame) you see the gascurve shows during some intervals a straight line, asif nothing is being mined, do you know why? |
|
|
| Saab Sweden. May 04 2003 20:27. Posts 28 | Profile |
sorry for the double post, guess i pressed reload before back..
anyhow, would it be possible to make a time derivative of the supply stock to have a production?
btw, i dont want to pressure you or anything, but if you need idea's, i have some |
|
|
 Bill307  Canada. May 04 2003 21:33. Posts 8583 | Profile Blog |
jca, there is a problem with your BWChart program. I'm not sure how exactly it's calculating how many units a player built during the course of the game, but it definitely seems to be doing it wrong. For instance, I made a short rep where I built 7 probes and started my 8th before leaving, then viewed it and was told that I had made 12 probes and that my opponent had made 13 drones =/.
I don't see how you could correctly calculate something like that, anyway, since the rep records all actions whether they work or they fail. |
|
|
| | 1 2 3 4 5 6 7 8 9 Next All |
