domachine opened a new pull request #25: Make secret configurable via
environment
URL: https://github.com/apache/couchdb-docker/pull/25
## Overview
This applies the changes mentioned in #9. Hopefully this makes the
Dockerfile ready to be released on docker-hub.
## Testing
willholley commented on a change in pull request #818: Rename selector to
partialfilterselector in indexes
URL: https://github.com/apache/couchdb/pull/818#discussion_r139371027
##
File path: src/mango/src/mango_cursor.erl
##
@@ -92,7 +92,7 @@
willholley commented on a change in pull request #818: Rename selector to
partialfilterselector in indexes
URL: https://github.com/apache/couchdb/pull/818#discussion_r139371966
##
File path: src/mango/src/mango_idx.erl
##
@@ -139,7 +139,25 @@ new(Db, Opts) ->
willholley commented on a change in pull request #818: Rename selector to
partialfilterselector in indexes
URL: https://github.com/apache/couchdb/pull/818#discussion_r139372354
##
File path: src/mango/test/03-operator-test.py
##
@@ -185,7 +185,7 @@ def
willholley commented on a change in pull request #818: Rename selector to
partialfilterselector in indexes
URL: https://github.com/apache/couchdb/pull/818#discussion_r139371653
##
File path: src/mango/src/mango_idx.erl
##
@@ -44,7 +44,7 @@
to_json/1,
delete/4,
eiri commented on issue #825: Allow urlencoded '+' characters in
_scheduler/jobs/ field
URL: https://github.com/apache/couchdb/issues/825#issuecomment-330279604
You really got me confused here with what exactly you are proposing. At
first I've read this as a suggestion to take `+` in URL
nickva commented on issue #825: Allow urlencoded '+' characters in
_scheduler/jobs/ field
URL: https://github.com/apache/couchdb/issues/825#issuecomment-330239672
@eiri how does this break the URI RFC?
This is an automated
eiri commented on issue #825: Allow urlencoded '+' characters in
_scheduler/jobs/ field
URL: https://github.com/apache/couchdb/issues/825#issuecomment-330239359
I'm against of breaking with [URI
RFC](https://tools.ietf.org/html/rfc3986#section-2.2) for convenience and in my
experience
garrensmith commented on issue #971: Refactor documents/mango addon to use Redux
URL: https://github.com/apache/couchdb-fauxton/pull/971#issuecomment-330206130
@Antonio-Maranhao nice work. I found another bug. If I set the limit it
doesn't obey that limit.
So if set a `limit: 5` and set
garrensmith commented on issue #818: Rename selector to partialfilterselector
in indexes
URL: https://github.com/apache/couchdb/pull/818#issuecomment-330206424
@davisp I've made all the required changes. Could you take a look when you
get a chance.
willholley commented on issue #816: Fix mango json index selection
URL: https://github.com/apache/couchdb/pull/816#issuecomment-330250765
@davisp I went with your second suggestion and implemented a
`has_required_fields` function
nickva opened a new issue #825: Allow urlencoded '+' characters in
_scheduler/jobs/ field
URL: https://github.com/apache/couchdb/issues/825
Since replication IDs could have `+` characters in them (they might look
like `2abc+create_target+continuous` for example), the ID has to be URL
nickva commented on issue #825: Allow urlencoded '+' characters in
_scheduler/jobs/ field
URL: https://github.com/apache/couchdb/issues/825#issuecomment-330248434
`+` can be considered a sub-delim because it separates the base replication
ID from the extensions. For example replication
eiri commented on issue #825: Allow urlencoded '+' characters in
_scheduler/jobs/ field
URL: https://github.com/apache/couchdb/issues/825#issuecomment-330242519
Plus is sub-delim reserved character. RFC's stating:
> If data for a URI component would conflict with a reserved character's
nickva commented on issue #825: Allow urlencoded '+' characters in
_scheduler/jobs/ field
URL: https://github.com/apache/couchdb/issues/825#issuecomment-330248434
`+` can be considered sub-delim because it separates the base replication ID
from the extensions. For example replication
nickva commented on issue #825: Allow urlencoded '+' characters in
_scheduler/jobs/ field
URL: https://github.com/apache/couchdb/issues/825#issuecomment-330248434
+ can be considered sub-delim because it separates the base replication ID
from the extensions. For example replication
wohali commented on issue #25: Make secret configurable via environment
URL: https://github.com/apache/couchdb-docker/pull/25#issuecomment-330299844
A problem with this PR is that because the file you're creating
alphabetically sorts later than `docker.ini`, the admin user will be written
wohali commented on issue #25: Make secret configurable via environment
URL: https://github.com/apache/couchdb-docker/pull/25#issuecomment-330299844
A problem with this PR is that because the file you're creating
alphabetically sorts later than docker.ini, the admin user will be written
nickva commented on issue #825: Allow urlencoded '+' characters in
_scheduler/jobs/ field
URL: https://github.com/apache/couchdb/issues/825#issuecomment-330311639
It would be a different encoding of the same resource. Is encoding the same
as representation, I not sure. For example is
nickva commented on issue #825: Allow urlencoded '+' characters in
_scheduler/jobs/ field
URL: https://github.com/apache/couchdb/issues/825#issuecomment-330311639
It would be a encoding of the same resource. Is encoding the same as
representation, I not sure. For example is document `%61`
eiri commented on issue #825: Allow urlencoded '+' characters in
_scheduler/jobs/ field
URL: https://github.com/apache/couchdb/issues/825#issuecomment-330323062
Right. And you are suggesting to fix this by, depending on perspective,
using non-standard encoder which treats `+`, `%2B` and
eiri commented on issue #825: Allow urlencoded '+' characters in
_scheduler/jobs/ field
URL: https://github.com/apache/couchdb/issues/825#issuecomment-330304329
`2bc+create%20target%2Bcontinuous` is not valid replication ID, but valid
_representation_ of replication ID if we'd go with
davisp commented on a change in pull request #816: Fix mango json index
selection
URL: https://github.com/apache/couchdb/pull/816#discussion_r139482203
##
File path: src/mango/src/mango_selector.erl
##
@@ -566,3 +567,111 @@ match({[{Field, Cond}]}, Value, Cmp) ->
davisp commented on a change in pull request #816: Fix mango json index
selection
URL: https://github.com/apache/couchdb/pull/816#discussion_r139482134
##
File path: src/mango/src/mango_selector.erl
##
@@ -566,3 +567,111 @@ match({[{Field, Cond}]}, Value, Cmp) ->
davisp commented on a change in pull request #816: Fix mango json index
selection
URL: https://github.com/apache/couchdb/pull/816#discussion_r139481549
##
File path: src/mango/src/mango_selector.erl
##
@@ -566,3 +567,111 @@ match({[{Field, Cond}]}, Value, Cmp) ->
davisp commented on a change in pull request #816: Fix mango json index
selection
URL: https://github.com/apache/couchdb/pull/816#discussion_r139482559
##
File path: src/mango/test/05-index-selection-test.py
##
@@ -14,7 +14,6 @@
import user_docs
import unittest
-
davisp commented on a change in pull request #816: Fix mango json index
selection
URL: https://github.com/apache/couchdb/pull/816#discussion_r139481344
##
File path: src/mango/src/mango_selector.erl
##
@@ -566,3 +567,111 @@ match({[{Field, Cond}]}, Value, Cmp) ->
davisp commented on a change in pull request #816: Fix mango json index
selection
URL: https://github.com/apache/couchdb/pull/816#discussion_r139481287
##
File path: src/mango/src/mango_idx_view.erl
##
@@ -114,11 +114,11 @@ columns(Idx) ->
is_usable(Idx, Selector) ->
davisp commented on a change in pull request #816: Fix mango json index
selection
URL: https://github.com/apache/couchdb/pull/816#discussion_r139482797
##
File path: src/mango/test/05-index-selection-test.py
##
@@ -92,6 +91,80 @@ def test_invalid_use_index(self):
davisp commented on a change in pull request #816: Fix mango json index
selection
URL: https://github.com/apache/couchdb/pull/816#discussion_r139482441
##
File path: src/mango/src/mango_selector.erl
##
@@ -566,3 +567,111 @@ match({[{Field, Cond}]}, Value, Cmp) ->
davisp commented on a change in pull request #816: Fix mango json index
selection
URL: https://github.com/apache/couchdb/pull/816#discussion_r139481950
##
File path: src/mango/src/mango_selector.erl
##
@@ -566,3 +567,111 @@ match({[{Field, Cond}]}, Value, Cmp) ->
davisp commented on a change in pull request #816: Fix mango json index
selection
URL: https://github.com/apache/couchdb/pull/816#discussion_r139481723
##
File path: src/mango/src/mango_selector.erl
##
@@ -566,3 +567,111 @@ match({[{Field, Cond}]}, Value, Cmp) ->
davisp commented on a change in pull request #816: Fix mango json index
selection
URL: https://github.com/apache/couchdb/pull/816#discussion_r139481449
##
File path: src/mango/src/mango_selector.erl
##
@@ -566,3 +567,111 @@ match({[{Field, Cond}]}, Value, Cmp) ->
nickva commented on issue #825: Allow urlencoded '+' characters in
_scheduler/jobs/ field
URL: https://github.com/apache/couchdb/issues/825#issuecomment-330294792
> ""2abc+create%20target%2Bcontinuous"`
Replication ID format is not arbitrary though. This is possible exactly
nickva commented on issue #825: Allow urlencoded '+' characters in
_scheduler/jobs/ field
URL: https://github.com/apache/couchdb/issues/825#issuecomment-330294792
> ""2abc+create%20target%2Bcontinuous"`
Replication ID format is not arbitrary though. This is possible exactly
nickva commented on issue #825: Allow urlencoded '+' characters in
_scheduler/jobs/ field
URL: https://github.com/apache/couchdb/issues/825#issuecomment-330311639
Yap it would be a different representation of the same replication.
> Also, anecdotical, reading around gave me a notion
wohali opened a new issue #824: Allow access to current's node's local
interface at /_node/local/
URL: https://github.com/apache/couchdb/issues/824
Right now, to access a given node's local interface, there are two choices:
1. Port 5986 (which we intend to remove with v3.0 and up)
nickva commented on issue #806: FEEDBACK ONLY: Compactor optimize emsort
URL: https://github.com/apache/couchdb/pull/806#issuecomment-329851447
Updated couch_file:append_terms() to take `[{compression, none}]` and it
yields some improvement in compaction times.
```
nickva opened a new pull request #826: Skip compression for EMSort data
URL: https://github.com/apache/couchdb/pull/826
EMSort data is either FDIs or `{{Id, Seq}, Loc}` KVs and using the default
snappy
compression on it is wasteful. snappy is a NIF so no matter how small the
data
nickva commented on issue #806: FEEDBACK ONLY: Compactor optimize emsort
URL: https://github.com/apache/couchdb/pull/806#issuecomment-330401767
Since dropping snappy compression for emsort showed some positive results.
Made a PR for this PR
https://github.com/apache/couchdb/pull/826
40 matches
Mail list logo
