@Alon #2 is not a concern, but I don't see why this is forced, perhaps just a simple boolean to tell it if we need it would be sufficient.
@Floh As I mentioned above, a HEAD request would be much more efficient than GET for this kind of operation. On Friday, October 16, 2015 at 3:39:31 AM UTC-4, Floh wrote: > > As for point 3: there's isn't really a useful way in HTTP to check if a > file exists, apart from firing a request and checking the returned HTTP > status (the emscripten_async_wget function would call the onerror callback > in that case). > > What I do usually is to create a table-of-content (TOC) file in an > offline-build process, this contains all existing files on the web servers, > and sometimes additional information like file size and MD5 hash. The TOC > file is downloaded first and stored in a map/dictionary in RAM. When the > actual asset files are loaded, an exists-check is done first through the > TOC, thus if a file doesn't exist no server round-trip is necessary. You > can also check for the existence of directories, and iterate over their > contents, since all this information is in the TOC-file (and HTTP doesn't > provide this functionality). > > Cheers, > -Floh. > > Am Donnerstag, 15. Oktober 2015 19:41:19 UTC+2 schrieb Robert Goulet: >> >> Hi all, >> >> I have a few questions about asynchronous file system API: >> >> 1. Can we use emscripten_async_wget or other async methods before we >> enter the emscripten main loop? >> 2. The documentation says "*In addition to fetching the URL from the >> network, the contents are prepared so that the data is usable in IMG_Load >> and so forth...*" Is this still true in *1.34.12*? I recall there was >> a change to disable content preparation by default; does this apply to >> these functions or are they still preparing content? >> 3. I don't see any API to verify if a file exist before we async wget >> it, would it be possible to add an API that only tell us if the file >> exist >> on the server? >> >> In our engine, the way it currently is we load resources asynchronously, >> but it must be completed before we start the update/render loop. Would that >> be an architecture issue? >> >> Thanks! >> > -- You received this message because you are subscribed to the Google Groups "emscripten-discuss" group. To unsubscribe from this group and stop receiving emails from it, send an email to emscripten-discuss+unsubscr...@googlegroups.com. For more options, visit https://groups.google.com/d/optout.
