Jump to content
  • Advertisement

gamma_strahler

Member
  • Content count

    42
  • Joined

  • Last visited

Community Reputation

122 Neutral

About gamma_strahler

  • Rank
    Member
  1. gamma_strahler

    mysql Question

    Hi, we currently evaluate some free databases for a project in our company. For mysql, i wondered wether it is possible to create schemata (like in sql server with the statement "create schema"). the documententation from the mysql homepage only says that the statement "CREATE SCHEMA" is a synonym for "CREATE DATABASE". thanks for any answers! greets, gamma_strahler
  2. gamma_strahler

    Terrain Lighting

    Thanks for your replies! I think i tend to use a normal map or light map since storing, preprocessing and recalculating the vertex normals needs to much processing power as well as RAM so that this is not the best choice. But is a single lightmap or normal map sufficient? My terrain has currently a size of 1024x1024 cells. Since i plan to use an even larger terrain (4096 and larger), i suppose it would look blocky if i overlay the terrain with a single map. What are your experiences? Thanks in advance, gamma_strahler
  3. gamma_strahler

    Terrain Lighting

    Hi, i have implemented a terrain engine which uses geomipmapping. Now i want to implement some lighting, but i have no idea of addressing the correct normals. Since i use different levels of detail (the geomipmaps), i have different sets of vertices and therefore different normals. So if i switch between LOD levels, the shading looks different and the user notices more "popping". So what is the correct solution? Which methods are used for terrain lighting? Can you point me to some good articles? I have searched the web but while there are tons of tutorials and articles for terrain rendering, links that focus on terrain lighting are somewhat rare. Thanks in advance, Gamma_strahler
  4. gamma_strahler

    Efficient database search?

    @moosedude The problem aren´t the SQL´s (we have checked that extensively). The bottleneck here is that there are too many connections (our customers have too many users that are simultaneously logged in and make a lot of queries, uodates and so on at high frequencies ). To make it more difficult, many of them work on terminal servers which lack of resources like RAM and CPU power. So you could thay in this case the sql´s are the problem, but i think that it is not possible to optimize these sql´s further. I can post an example later.
  5. gamma_strahler

    Efficient database search?

    Thanks for your tipps! I have evaluated lucene already, but it is not an option since it lacks support for umlauts.
  6. gamma_strahler

    Efficient database search?

    We have an index (the primary key) column defined, but i mean a real index-based search (like the one Windows XP does it)
  7. Hi, i have the order to improve the fuzzy search of our business application. Unfortunately our database connections (MSSQL) are very slow ( a single search query by name takes over 5 seconds for just 10,000 records. ) would it be worthwhile to implmenent an index-based search? or are there any other preferable ways that could improve the speed of our searches? thanks for any help gamma_strahler
  8. gamma_strahler

    Import of Html Tables

    Quote:Original post by erissian Quote:Original post by gamma_strahler Quote:Original post by d000hg Quote:Original post by RDragon1 You should have generated the data in xml, not html, and used a stylesheet to generate html for viewing in a browser. Then you would just have to process the real data.Where did he say he has control over the format of the input data? Writing software's about solving the problem, not thinking how we should have avoided the problem in the past (although learning these lessons for future projects is important). That´s the main problem. The data is from an external provider which has a collection of all annual accounts from our country, and the html pages are generated with various different tools. i have encountered as many as 40 variations in the structure which is a pain to handle. regular expressions would be not of much use in this case. im not sure, but could datamining be an appropropriate candidate to address this problem? Isn't that just renaming the problem? Anyways, is this a provider you're dealing with or are your scraping the data from a webpage? If you're dealing directly, you may be able to request the data in another format. Otherwise, if they are publishing this data, they are likely using another program to generate the tables. There may be several odd cases, but I'm sure there's some method to it. Can you give us an example? It would go beyond the scope of this thread to post examples here, but the situation is as follows: Our provider only publishes the collected data, and has also no control over the generated format. We are allowed to save individual html files to disk but are not allowed to scrape the data in any form of automated http/soap request. The data variations are a bit tricky, since: - There is no unique tag or attribute which helps identifying the correct tables. It is also not enough to ensure unuqieness by just investigating the position column names, since there may more tables which share equal position names. - The second problem is that different position names can map to the same position in our schema and vice versa. to adress the first, i could use synonyms, but successfuyll testing 100 samples is not meaningful enough to ensure that they also pass for all 600.000
  9. gamma_strahler

    Import of Html Tables

    Quote:Original post by d000hg Quote:Original post by RDragon1 You should have generated the data in xml, not html, and used a stylesheet to generate html for viewing in a browser. Then you would just have to process the real data.Where did he say he has control over the format of the input data? Writing software's about solving the problem, not thinking how we should have avoided the problem in the past (although learning these lessons for future projects is important). That´s the main problem. The data is from an external provider which has a collection of all annual accounts from our country, and the html pages are generated with various different tools. i have encountered as many as 40 variations in the structure which is a pain to handle. regular expressions would be not of much use in this case. im not sure, but could datamining be an appropropriate candidate to address this problem?
  10. Hi, i have the assignment to write an import function for our software that reads in financial annual accounts from a html table element. the problem is, that the generated html pages have structural variations which make it difficult to spot the relevant data and map it to the correct positions in out software product. i thought about textual recognition, but the variations are too big to make it reliable. fuzzy logic is also not an option here. i would be appreciated if you have any ideas or hints of how i could realize this feature. thanks in advance Gammastrahler
  11. Hi, in our company we want to redesign the search logic for our application. Part of our concept is the idea to find the database entries which contain ALL specified search terms in the database entries (ANDed search, something like google does it). so i searched the web for existing algorithms for this approach but didn´t find anything explaning this strategy. Can you give me some starting points and hints of how i could accomplish this? Thanks in advance! gamma_strahler
  12. Hi, i have already bought some good books, but since i´m more a weekend game programmer than a really hardcore one, my state of knowledge for newer graphics technologies lacks, particularly with regard to shader programming. there are so many books about shaders available, which all seem to be excellent (the ShaderX series, the GPU Gems series), making a decision almost impossible. What´s important for me, is that the level of difficulty for shaders is intermediate, and should not focus on shaders only, but also on using and implementing them with Direct3D / OpenGL. Also, since i used to program most of my time in OpenGL before, but want to focus more on Direct3D, i need a good book covering Direct3D. Last, but not least, i think i need some books about Game Programming, which covers topics like game engine architecture, scene graphcs, game objects, game data structures, etc... The books i have already read and found very useful are: - Mathematics for 3D Game Programming and Computer Graphics - Computer Graphics Principles & Practice - Game Programming Gems II - OpenGL SuperBible - 3D Math Primer - The CG Tutorial Do you have any further reccomendations for me? thanks in advance, Gamma_strahler -
  13. gamma_strahler

    Local objects in c++

    ups... sorry for the double-post
  14. gamma_strahler

    Local objects in c++

    Thanks for your answers, but there is one problem left which i don´t really understand. I have encountered that this only works as long as i don´t allocate dynamic memory in an object. For example, the following code crashes (which uses the free store for a data member): struct vector3 { float _x, _y, _z; char *_p; vector3(float x, float y, float z) : _x(x), _y(y), _z(z) { _p = new char [1024]; } ~vector3() { if (_p) delete [] _p; _p = 0; } }; using namespace std; vector<vector3> g_vertexArray; void do_another_thing() { g_vertexArray.push_back(vector3(0, 1, 2)); } void main() { do_another_thing(); for (vector<vector3>::iterator it = g_vertexArray.begin(); it != g_vertexArray.end(); ++it) { cout << it->_x << endl; // CRASH!!!! } }
  15. Hi, if i have some function like this: void my_add_vertex() { g_vertexArray.push_back(CVertex(1.0, 1.0, 1.0)); } then i create an instance of CVertex, which is only valid inside the my_add_vertex runtime, and it is passed as a reference to push_back. Is this dangerous? the constructor of CVertex only initializes x, y and z (and other data members). i see this practice in many tutorials, and from my state of knowledge, i would say that it is unsafe to pass a reference to a local variable, though i get no compiler warning. would be nice if you could shed some light on this. thanks Gamma_strahler
  • Advertisement
×

Important Information

By using GameDev.net, you agree to our community Guidelines, Terms of Use, and Privacy Policy.

We are the game development community.

Whether you are an indie, hobbyist, AAA developer, or just trying to learn, GameDev.net is the place for you to learn, share, and connect with the games industry. Learn more About Us or sign up!

Sign me up!