Skip to content
Commits on Nov 9, 2011
  1. @tilgovi

    Revert "use couch_db calls instead of couch_btree"

    We can't do this without other changes since couch_db:get_design_docs
    is hard coded to skip deleted documents, and that's incompatible with
    how _changes is supposed to behave.
    This reverts commit e41d226.
    tilgovi committed Nov 9, 2011
  2. @tilgovi

    Futon Cache-Control

    Specify that paths under /_utils be served with Cache-Control headers:
     - private: so permissions on dbs aren't violated by caching proxies
     - must-revalidate: so we don't serve a stale futon
    The revalidation is particularly useful for developers, but might also
    serve to prevent stale versions of documents being shown accidentally.
    tilgovi committed Oct 20, 2011
  3. @tilgovi

    remove version number from futon static resources

    No process was in place to update these query strings, which would
    typically serve to prevent excessive browser caching if they were
    made to change. No one wants to change these every release. Templating
    every single HTML file here feels silly. I'm just nuking this junk.
    Fix COUCHDB-860
    tilgovi committed Oct 20, 2011
  4. @tilgovi

    some refactoring in couch_httpd_db.erl

    Fix COUCHDB-1277. These changes are largely for code clarity and
    conciseness, but have a few other nice effects as well.
    * Responses to documents created/modified via form data POST to /db/doc
      or copied with COPY should now include a Location header.
    * ?batch=ok updates should include a Location header.
    * Form data POST to /db/doc now includes an ETag response header.
    * ?batch=ok is now supported for COPY and POST /db/doc updates.
    * ?new_edits=false is now supported for more update paths. This change
      is likely not generally useful, but listed here for completeness.
    tilgovi committed with tilgovi Sep 8, 2011
  5. @tilgovi

    use couch_db calls instead of couch_btree

    couch_changes can use the interface from couch_db to access documents,
    keeping the logic of the btree storage better encapsulated.
    tilgovi committed Nov 9, 2011
  6. @tilgovi

    couch_db:get_design_docs/1 -> #full_doc_info{}

    Rather than open every design doc at once with this call, just return
    the full doc info. It's the cheapest thing to do because it lets other
    code open the documents as needed.
    tilgovi committed Nov 9, 2011
  7. @tilgovi

    export couch_db:get_full_doc_info(s)

    It's useful to export these to keep the btree access well
    encapsulated and still provide fast access to the full_doc_info
    tilgovi committed Nov 9, 2011
  8. @tilgovi

    fix improper comparison on filtered changes

    Regression caught by the changes js test.
    tilgovi committed Nov 9, 2011
Commits on Nov 8, 2011
  1. @tilgovi
  2. @tilgovi
  3. @rnewson
Commits on Nov 7, 2011
  1. @KlausTrainer @fdmanana

    Add www/script/test/coffee.js to share

    Fixes COUCHDB-1336.
    KlausTrainer committed with fdmanana Nov 7, 2011
Commits on Nov 6, 2011
  1. @fdmanana

    Updated CHANGES & NEWS for 1.2.0

    Added mention to the new changes feed built-in filter `_view`.
    fdmanana committed Nov 6, 2011
Commits on Nov 5, 2011
  1. @fdmanana
  2. @fdmanana

    Fix badarg error in couch_os_process

    The value of the KillCmd is an iolist and not a
    binary, so the binary_to_list call failed.
    Issue introduced by COUCHDB-1186.
    fdmanana committed Nov 5, 2011
  3. @fdmanana
  4. @fdmanana

    Add view request duration to Futon

    Closes COUCHDB-509
    fdmanana committed Nov 5, 2011
Commits on Nov 4, 2011
  1. @fdmanana

    Set gen_server call timeouts to infinity

    When a server is too busy, the default call timeout of
    5 seconds may not be enough when getting an OS process
    to execute JavaScript code.
    Closes COUCHDB-1333.
    fdmanana committed Nov 4, 2011
Commits on Nov 3, 2011
  1. @fdmanana

    Fix active tasks Futon screen date display

    Date months were incorrect, JavaScript's Date class
    getMonth() method returns an integer in the range 0..11.
    fdmanana committed Nov 3, 2011
Commits on Nov 2, 2011
  1. @rnewson

    Add newline to end of utf8.h

    Missing newline causes build failures.
    rnewson committed Nov 2, 2011
  2. @rnewson

    Tolerate whitespace runs in .ini files

    Previously we tolerated exactly one character of white space on
    either side of the equals sign in .ini files. Now you can have lots.
    rnewson committed Nov 2, 2011
Commits on Nov 1, 2011
  1. @fdmanana

    Log error on failure writing to the URI file

    Report, and most of the work by Rogutės Sparnuotos.
    Thanks. Closes COUCHDB-1326.
    fdmanana committed Nov 1, 2011
  2. @fdmanana

    Fix handling of SSL option verify_ssl_certificates

    If this parameter was set to false in the .ini config, Couch
    would still complain with following error message:
    "Verify SSL certificate enabled but file containing PEM encoded
    CA certificates is missing"
    Closes COUCHDB-1325.
    fdmanana committed Nov 1, 2011
Commits on Oct 31, 2011
  1. @rnewson
  2. @janl

    Properly adhere to bytes and offset parameters for /_log

    Closes COUCHDB-887
    janl committed Oct 31, 2011
  3. @janl
  4. @janl

    First fully working version of the JavaScript CLI tests.

    This patch has multiple parts:
    1. Make remaining tests work with CouchHTTP.
    2. Add ./test/javascript/run to make check.
    3. Add handling of OPTIONS method to couchjs's http client.
    4. Add referer sending to couchjs's http client.
    To run *all* tests, run `make check`.
    Closes COUCHDB-1312
    janl committed Oct 31, 2011
Commits on Oct 30, 2011
  1. @janl

    Fix typos and whitespace.

    Closes COUCHDB-1324
    Patch by Rogutės Sparnuotos.
    janl committed Oct 30, 2011
  2. @janl

    Expose raw_path in req to show, list and externals.

    Closes COUCHDB-676
    Closes COUCHDB-341
    janl committed Oct 30, 2011
  3. @janl

    Futon: Make delete database dialog more expressive.

    Closes COUCHDB-208
    janl committed Oct 30, 2011
  4. @kocolosk

    Fix retrieval of headers larger than 4k

    Our headers start with a <<1>> and then four bytes indicating the length
    of the header and its checksum. When the header is larger than 4090
    bytes it will be split across multiple blocks in the file and will need
    to be reassembled on read. The reassembly consists of stripping out
    <<0>> from the beginning of each subsequent block in the
    remove_block_prefixes/2 function. The bug here is that we tell
    remove_block_prefixes that we're starting 1 byte into the current block
    instead of 5, so it ends up removing one good byte from the header and
    injecting one or more random <<0>>s.
    Headers larger than 4k are very rare and generally require a view group
    with a huge number of indexes or indexes with fairly large reductions,
    which explains why this bug has gone undetected until now.
    Closes COUCHDB-1319.
    kocolosk committed Oct 26, 2011
  5. @trondn @janl

    Try to determine the location of erl_driver.h at configure time

    Use the existence of erl in path to try to figure out where the
    file should be located unless the user specified it.
    Patch by Trond Norbye.
    Closes COUCHDB-162
    trondn committed with janl Sep 21, 2011
  6. @janl

    Allow POST to _log.

    POST /_log {"level":"info|debug|error", "message":"your message here"}
    Patch by Robert Newson.
    Closes COUCHDB-464
    janl committed Oct 30, 2011
  7. @janl

    Prefer stringFun() over toString() and add binary test to update func…

    Closes COUCHDB-626
    Patch by Brian Candler.
    janl committed Oct 30, 2011
  8. @janl

    Fix change password dialog for admin users.

    Fix change password for admin users that have a config entry as well
    as a _users doc.
    The password change flow is now:
     - if the current user has the _admin role set
       - if it has a corresponding config entry
         - update that entry
     - if the current user has a user doc
       - update that user doc
    janl committed Oct 30, 2011
Something went wrong with that request. Please try again.