Commit message (Collapse) | Author | Age | Files | Lines | |
---|---|---|---|---|---|
* | docs: Make sphinx build process match the current HEADrelease-0.8 | Marcel Hellkamp | 2011-03-20 | 1 | -36/+103 |
| | |||||
* | fix: "Digits cannot be used in route parameter names" (#108) | Marcel Hellkamp | 2010-12-14 | 2 | -1/+9 |
| | | | | Thanks to Alexey Borzenkov | ||||
* | Version bumb to 0.8.5 (security fix)0.8.5 | Marcel Hellkamp | 2010-11-11 | 1 | -1/+1 |
| | |||||
* | Security fix: cookie_decode() vulnerability to statistical timing attacks. | Marcel Hellkamp | 2010-11-11 | 1 | -5/+11 |
| | | | | | | | | | | | | | | | This vulnerability reduces the amount of time required to guess a valid signature for a forged cookie. Only applications that use the secure-cookie feature are affected. Detail: Prior to this patch, the transmitted signature and the calculated (valid) signature were compared using the python '==' operator. The runtime of this operation depends on the length of a common string prefix. The time difference is barely measurable for a single request, but it exists. An attacker could send a large amounts of request with different signatures, measure the time for each request and statistically determine which signature is most likely to have a valid prefix. On a fast network, this can drastically reduce the time required to guess a valid signature for a forged cookie. | ||||
* | Version bumb to 0.8.4 (small bugfix)0.8.4 | Marcel Hellkamp | 2010-10-09 | 2 | -1/+5 |
| | |||||
* | made DictMixins usable with the @view decorator | Brandon Gilmore | 2010-10-09 | 1 | -1/+1 |
| | |||||
* | Version bumb to 0.8.3 (bugfix release)0.8.3 | Marcel Hellkamp | 2010-08-26 | 1 | -1/+1 |
| | |||||
* | docs: Changelog | Marcel Hellkamp | 2010-08-26 | 1 | -0/+5 |
| | |||||
* | fix: (issue #90) Reloading server dies on slow hardware. | Marcel Hellkamp | 2010-08-11 | 1 | -1/+1 |
| | | | | | | | | | | | | Python seems to have some serious issues with the GIL when IO bound and CPU bound threads are mixed on a multi-core CPU [1]. This may prevent the observer thread to wake up and touch the lock-file it time. I observed this even on otherwise idling hardware. This patch does not fix the bug completely, but it should give the observer enough time and work around this issue in most cases. Dying observer processes are relatively rare. An increased timeout should not hurt that much. [1] http://dabeaz.blogspot.com/2010/01/python-gil-visualized.html | ||||
* | Version bumb to 0.8.2 (bugfix release)0.8.2 | Marcel Hellkamp | 2010-07-29 | 1 | -1/+1 |
| | |||||
* | Added release documentation for 0.8.2 | Marcel Hellkamp | 2010-07-29 | 1 | -0/+9 |
| | |||||
* | fix: This solves issue #87 (FileCheckerThread seems to fail on eggs) | Marcel Hellkamp | 2010-07-29 | 1 | -1/+1 |
| | |||||
* | Fixed spelling of "recipes", other typos in docs. | David Buxton | 2010-07-21 | 2 | -10/+10 |
| | | | | Moved "apidoc/recieps.rst" to "apidoc/recipes.rst". | ||||
* | fix: mount() was mis-labeled as wrapping Bottle.get_url(). | Marcel Hellkamp | 2010-07-19 | 1 | -1/+1 |
| | | | | Thanks to Ian | ||||
* | fix: (Issue #83) Bottle.get_url() does not return correct path when ↵ | Marcel Hellkamp | 2010-07-19 | 2 | -4/+12 |
| | | | | SCRIPT_NAME is set. | ||||
* | fix: Added backward compatibility and deprecation warnings to some of the ↵ | Marcel Hellkamp | 2010-07-19 | 2 | -4/+25 |
| | | | | | | API changes. docs: Added undocumented API changes to changelog | ||||
* | Fixed some documentation spelling, punctuation and typos. | David Buxton | 2010-07-19 | 7 | -121/+121 |
| | |||||
* | Updated MANIFEST.in | Marcel Hellkamp | 2010-07-15 | 1 | -2/+3 |
| | |||||
* | Version bumb to 0.8.1 (new release)0.8.1 | Marcel Hellkamp | 2010-07-15 | 1 | -1/+1 |
| | |||||
* | Last documentation fixes. | Marcel Hellkamp | 2010-07-15 | 5 | -17/+127 |
| | |||||
* | Merge branch 'reloader-0.8' into release-0.8 | Marcel Hellkamp | 2010-07-15 | 1 | -56/+89 |
|\ | |||||
| * | Code cleanup. | Marcel Hellkamp | 2010-07-15 | 1 | -36/+29 |
| | | |||||
| * | Bugfix: If the reloading observer process got killed, the child process ↵ | Marcel Hellkamp | 2010-07-13 | 1 | -49/+89 |
|/ | | | | | | survived. Bugfix: Infinite recursion in reloading code. | ||||
* | Merge branch 'stplnolist' into release-0.8 | Marcel Hellkamp | 2010-06-30 | 2 | -59/+59 |
|\ | |||||
| * | fix: Issue #77 - Templates should return strings as documented, not list of ↵ | Marcel Hellkamp | 2010-06-30 | 2 | -59/+59 |
|/ | | | | | | strings. Thanks to "tristanz" | ||||
* | Documentation (lots of it) | Marcel Hellkamp | 2010-06-27 | 10 | -369/+377 |
| | |||||
* | Version bump to 0.8.rc10.8.rc1 | Marcel Hellkamp | 2010-06-26 | 1 | -1/+1 |
| | |||||
* | Merge remote branch 'origin/405router' | Marcel Hellkamp | 2010-06-26 | 3 | -87/+111 |
|\ | |||||
| * | Refactored the method-awareness from the Router-class into the Bottle-class. ↵ | Marcel Hellkamp | 2010-06-25 | 2 | -200/+120 |
| | | | | | | | | | | | | | | | | | | | | | | | | The router now contains dicts instead of callbacks and the Bottle.match_url() method picks the right entry from the dict. The router should be kept as simple as possible and the Bottle class is the right place to raise any HTTPError(405) exceptions. I also removed the backup-methods and made sure all visible APIs stay backwards compatible. | ||||
| * | implemented 404/405 with optimized routing | Alex Popescu | 2010-06-22 | 3 | -30/+134 |
| | | |||||
* | | Merge remote branch 'origin/stpl_bugfix-0.8' | Marcel Hellkamp | 2010-06-26 | 3 | -9/+72 |
|\ \ | |||||
| * | | fix: In stpl, there was a chance that text blocks were split into single ↵ | Marcel Hellkamp | 2010-06-25 | 1 | -1/+1 |
| | | | | | | | | | | | | characters and handled inefficiently. | ||||
| * | | bugfix: SimpleTemplate no supports some python multi-line statements | Marcel Hellkamp | 2010-06-25 | 3 | -7/+54 |
| | | | |||||
| * | | bugfix: SimpleTemplate now detects python comments correctly | Marcel Hellkamp | 2010-06-25 | 2 | -4/+20 |
|/ / | |||||
* | | fix: A small bug in the test toolbox which rendered the assertInBody() tests ↵ | Marcel Hellkamp | 2010-06-22 | 1 | -3/+3 |
|/ | | | | | | useless. Thanks to Alex Popescu. | ||||
* | Merge branch 'file_wrapper' | Marcel Hellkamp | 2010-06-19 | 1 | -3/+18 |
|\ | |||||
| * | fix: A fallback for 'wsgi.file_wrapper' MUST have a close() method that ↵ | Marcel Hellkamp | 2010-06-19 | 1 | -3/+18 |
|/ | | | | | | invokes the original file-like object's close() method. Thanks to Enrico Zini. | ||||
* | Merge branch 'file_wrapper' | Marcel Hellkamp | 2010-06-19 | 1 | -2/+2 |
|\ | |||||
| * | fix: Return wsgi.file_wrapper wrapped files directly to the wsgi server. | Marcel Hellkamp | 2010-06-19 | 1 | -2/+2 |
|/ | | | | Thanks to Alex Popescu, Enrico Zini and David Paleino. | ||||
* | Merge branch 'bequiet' | Marcel Hellkamp | 2010-06-18 | 1 | -6/+10 |
|\ | |||||
| * | Silence WSGIRefServer on run(quiet=True). | Marcel Hellkamp | 2010-06-18 | 1 | -2/+6 |
| | | |||||
| * | fix: Calling run(quiet=True) broke some of the server adapters. | Marcel Hellkamp | 2010-06-18 | 1 | -4/+4 |
|/ | |||||
* | Whitespace | Marcel Hellkamp | 2010-06-13 | 1 | -6/+6 |
| | |||||
* | new: Support for tuples as return value. | Marcel Hellkamp | 2010-06-13 | 2 | -1/+6 |
| | |||||
* | fix: 304 responses must not return entity headers (rfc2616 10.3.5). | Marcel Hellkamp | 2010-06-13 | 3 | -4/+30 |
| | | | | | | | fix: 204 responses should not return a content-type header (rfc2616 10.2.3*). * The RFC does not say this, but wsgiref.validate() does and it makes some sense. Thanks to Alex Popescu for this bug report. | ||||
* | Removed a redundant test for wsgi.file_wrapper in Bottle._cast() and added ↵ | Marcel Hellkamp | 2010-06-13 | 4 | -6/+23 |
| | | | | some tests for status codes that should result in an empty response body. | ||||
* | fix: Changed HTTP status code for send_file() calls that fail because of ↵ | Marcel Hellkamp | 2010-06-13 | 2 | -4/+4 |
| | | | | | | root-path restrictions or missing permissions from 401 (Unauthorized) to 403 (Forbidden) to comply the RFC. Thanks to Alex Popescu for the bug report. | ||||
* | Changed run and fixed bottle reloader, so child process exits, | Johannes Schönberger | 2010-06-03 | 1 | -3/+13 |
| | | | | if parent process got killed | ||||
* | bugfix: HeaderDict.get() was case sensitive. | Marcel Hellkamp | 2010-06-02 | 2 | -0/+11 |
| | |||||
* | fix: Be quiet if a CGI-like server adapter is used | Marcel Hellkamp | 2010-05-20 | 1 | -2/+6 |
| |