forked from ericsink/SQLitePCL.raw
-
Notifications
You must be signed in to change notification settings - Fork 0
/
todo.txt
213 lines (141 loc) · 5.9 KB
/
todo.txt
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
61
62
63
64
65
66
67
68
69
70
71
72
73
74
75
76
77
78
79
80
81
82
83
84
85
86
87
88
89
90
91
92
93
94
95
96
97
98
99
100
101
102
103
104
105
106
107
108
109
110
111
112
113
114
115
116
117
118
119
120
121
122
123
124
125
126
127
128
129
130
131
132
133
134
135
136
137
138
139
140
141
142
143
144
145
146
147
148
149
150
151
152
153
154
155
156
157
158
159
160
161
162
163
164
165
166
167
168
169
170
171
172
173
174
175
176
177
178
179
180
181
182
183
184
185
186
187
188
189
190
191
192
193
194
195
196
197
198
199
200
201
202
203
204
205
206
207
208
209
210
211
212
worry about Batteries_V2 bait and switch across
bundle types with project.json or new csproj, where
the "better" support for transitive dependencies
might break things.
in the readme, explain why iOS lib has static in name.
and why internal. and why SQLitePCL.lib.embedded.Init()
need the ability to disconnect:
1. the env used to compile an assembly
2. the lib/env dir in nuget
for bundle_e_sqlite3, we want to build two things for netstandard11:
1. e_sqlite3, into netcoreapp
2. none, into netstandard11
enable_shared_cache. is this present? does akavache need it?
it's kinda wrong to always build batteries.cs with provider_none
for portable and netstandard.
is the v140 e_sqlite build compiled for app container only?
the batteries assemblies for wp80 don't need to
go in build, do they?
config_esqlite3 should get reworked, because it only
generates a nuspec, not a csproj. in fact, it's just
a string toolset.
maybe the subdirs x86/x64 for loading with anycpu
should be the rid. but this would break things,
so probably not.
several of the targets files are using the name
InjectReference when really they're just copying
the native DLL into the bin dir
change all the lib packages to put the binaries
in runtimes/rid/native, and change the .targets files
to use that path, and also change the targets files
to not do the copy when packages.config is present.
oren says kill the (equiv) platform builds and just use
netstandard. and yes, keep NetStandard.Library in
the deps, even though it drags in tons of transitives.
understand runtime.json and rid-specific deps, even
if they are not officially supported
cppinterop is now the only assembly which has the
old name. grumble.
problem with uwp test apps pfx file
problem with uwp test app, deploy, id ?
test runner for net45
rm cruft for old sqlcipher builds
in order to get back to netstandard for the tests,
we would need to switch xunit to the pre
--------
I wish the json generation code was using structures
instead of text writelines
how many cases are there where we could ditch a
platform-specific build and just let it use
netstandard? but then NetStandard.Library gets
dragged in, which is a lot of weird stuff.
problem with current testapps sln is that the
framework used to resolve is predetermined.
it would be better to gen the projects from
scratch and somehow add the nuget packages.
problem is specific to packages.config?
consider creating a new build directory on
every build
why does the error msg say ios project.json needs runtimes : win ?
is CopyNuGetImplementations causing trouble?
with project.json : runtimes?
ios build machine needs sdk 9.3
should MonoAndroid version be 2.3 or 6.0 ?
general-purpose gen_nuspec function
combine net45 libs into one pkg?
do we need a copy of the main raw assembly compiled specifically
for every tfm? like android? or is it sufficient to have
it compiled for each set of compile options, and each platform
use the one that is compatible with it. problem with netstandard
is that it brings in lots of little deps.
need to go to nuget.org and mark some of the older
packages deprecated
maybe the "first pinvoke set that works" code should
go somewhere else, higher up?
currently wp8-SL support is limited to e_sqlite3. no
sqlcipher. no extension SDK. although we could in
theory build multiple versions of this, each linked
against a different library name.
--------
can we do a dylib ios version?
version numbers on SQLitePCL.custom stuff should
not need to match SQLitePCL.raw, and might want to
match the underlying SQLite version number.
does the native code stuff in nuget 3.x get this
specific? can we provide a package which just looks
at the environment and then has a dependency on
The Right One?
xamarin.mac shouldn't even need to be built on the Mac.
just use a compatible build of SQLitePCL.raw.dll
----
clean out the rest of this file
----
Do the collation hook fix thing for config_log
consider whether maybe the sqlite3 intptr dispose should check
the version number and call sqlite3_close_v2 when it's supposedly
available.
add visual studio 2013 c++ runtime sdkreference. but injecting an
SDKReference in the .targets file sort of doesn't work, and using
a powershell tool is so inelegant. MESSAGE.
annoying that an injected reference doesn't show up in the
References list in visual studio
error if configuration is not release or debug
but configuration in xamarin.ios is actually stuff like iPhone, etc
why does the sqlitepcl ios platform assembly show
no assembly info when viewed in xamarin studio?
Win32 in the sln file
auto-generate the test projects as well?
docs: need a complete list, in tabular form, of the platform
assemblies and an explanation of the use cases for each one.
auto-generate this.
need at least one test project for each platform assembly.
add an NUnit project on Windows/Net45, so that build breakage
with NUnit can be detected when making changes on windows.
should reference NUnit as a nuget package
sqlite3.Mac? or use platform-installed sqlite on MacOS?
docs on all the methods in raw, especially where the params
differ from the C API
----------------
sqlite API issues:
should sqlite3_commit_hook() return something? the C function returns the
user pointer, which has no useful parallel for us to return. I sort of hate
the fact that it returns void, since we can't check errors on it. but
the underlying C function doesn't really return any useful info for error
checking either.
unprotected sqlite3_value: sqlite3_column_value, sqlite3_bind_value, sqlite3_result_value
sqlite3_create_module
sqlite3_db_config
sqlite3_db_mutex
sqlite3_get_auxdata
sqlite3_set_auxdata
sqlite3_limit
sqlite3_progress_handler
sqlite3_randomness
sqlite3_soft_heap_limit64
sqlite3_wal_autocheckpoint
sqlite3_wal_checkpoint
sqlite3_wal_checkpoint_v2
sqlite3_wal_hook
sqlite3_busy_handler
sqlite3_collation_needed -- difficult because of the be handle passed
to the callback