- Appended to comment:
drh added on 2009-09-24 14:32:43:
The fossil repository should store all filenames as UTF8. I think the bug here is that the "fossil add" command is not converting the names from the microsoft encoding into UTF8 before storing them. - Change resolution to "Open"
- Change comment to "When I run "fossil.exe add dir" under Windows, fossil recursively adding all files. If there are files with non-ascii names (Russian in my case), they inserted into fossil repo using system default encoding ('windows-1251' in Windows with Russian locale), and fossil works with them w/o problems. But in the web-interface on the files page (/dir) these files shown with unreadable names. This is because of UTF-8 <verbatim>fprintf(g.httpOut, "Content-Type: %s; charset=utf-8\r\n", zContentType);</verbatim> in the HTTP reply header (cgi.c). Please <b>delete "; charset=utf-8" part</b> from Content-Type header because there no CP convertor in the fossil. This will allow to set any charset in the META tags in the page header template and use browser's internal CP convertor or CP autodetector. This will also solve a problem from the ticket [d23fb0b742ac2e989e36bfef98281f970ab34e8f], I think."
- Change foundin to "f969b6cdde"
- Change private_contact to "09bcfe9beb6bd1f9361d71b153c193630ed2ac65"
- Change severity to "Severe"
- Change status to "Open"
- Change title to "i18n for filenames"
- Change type to "Code_Defect"