Skip to content

lightmare

  • Joined on Feb 8, 2012
Apr 25, 2016
lightmare reopened issue mapnik/mapnik#3384
@lightmare
Build error: scons: *** Do not know how to make File target `plugins/input/geojson/geojson_datasource.os'
Apr 25, 2016
lightmare commented on issue mapnik/mapnik#3384
@lightmare

Reopening as the issue persists. Optional objects listed explicitly in Makefile: #3384 (comment)

Apr 25, 2016
lightmare commented on issue mapnik/mapnik#3380
@lightmare

@artemp No. The error comes from this line. Neither the visitor class, nor the operator function are templates, meaning the function will be compil…

Apr 20, 2016
@lightmare

++index doesn't seem right. The warning you were trying to silence is spurious. No static analysis tool can know that it won't go below 0 because r…

Apr 15, 2016
@lightmare
Defective rendering of interfacing multipolygons, leisure=pitch and natural=scrub
Apr 15, 2016
lightmare commented on issue mapnik/mapnik#3388
@lightmare

Closing as OSM style error, not mapnik bug.

Apr 13, 2016
lightmare commented on issue mapnik/mapnik#3400
@lightmare

Isn't that stack trace weird? #3 evaluate operator()(mapnik::attribute const&) (not global_attribute) #1 and #2 value copy ctor (twice?) #0 and the…

Apr 12, 2016
@lightmare

@springmeyer that page specifically mentions ${TRAVIS_PULL_REQUEST} why not use that?

Apr 6, 2016
lightmare commented on issue mapnik/mapnik#3339
@lightmare

585de59 still running out of memory with the above loop. This has nothing to do with GDALRegisterAll, the race condition is in libgeotiff initializ…

Apr 5, 2016
lightmare commented on pull request philsquared/Catch#631
@lightmare

Why is there a "generic tree interface" ITracker when all its nodes have to be SectionTracker? It looks like the whole tree should've been template…

Apr 5, 2016
lightmare commented on issue mapnik/node-mapnik#623
@lightmare

The non-const overload you linked returns a reference to the pointer, no ref-counting there.

Apr 3, 2016
lightmare commented on issue mapnik/mapnik-support#70
@lightmare

That's a warm-up query to figure out what columns there are. You have a type in Layer srs="... +no _defs"

Apr 3, 2016
lightmare commented on issue mapnik/mapnik#3392
@lightmare

Check config.log what compiler was used, by default it's c++. Is your c++ symlinked to g++, or is it a different compiler? You can also try forcing…

Apr 2, 2016
lightmare commented on pull request philsquared/Catch#628
@lightmare

In case anyone wonders why they are needed, it's because some systems (win?) define function-like macro max(a, b), and expanding that after std:: w…

Apr 1, 2016
lightmare commented on pull request mapnik/mapnik#3389
@lightmare

Didn't check the old ustring rule before, it has no_skip which should be in the primary-expr alternative, too: unquoted_ustring %= no_skip[alpha >>…

Apr 1, 2016
lightmare commented on pull request mapnik/mapnik#3389
@lightmare

can you think of a clean way to keep the operator precedence fix but restore the support for unquoted strings? Yes, it's easy. Simply add ((alpha…

Mar 31, 2016
lightmare commented on pull request mapnik/mapnik#3389
@lightmare

Yes, it should've been mentioned in changelog; the pull fixed an operator precedence regression -- and that part is there -- but the retraction of …

Mar 31, 2016
lightmare commented on pull request mapnik/mapnik#3389
@lightmare

It's not a regression, the change was intentional. See #3287 for reference.

Mar 29, 2016
lightmare commented on issue philsquared/Catch#622
@lightmare

Take a look at https://github.com/mapnik/mapnik/blob/master/test/unit/run.cpp (adds options --prefix and --working-directory). For this to work you…

Mar 29, 2016
lightmare commented on pull request philsquared/Catch#623
@lightmare

convertInto(bool, T&) throws, but it's only used if !IsBool<T>. There's also convertInto(bool, bool&) overload which doesn't throw.

Mar 27, 2016
lightmare commented on pull request philsquared/Catch#562
@lightmare

While this may help a bit, it doesn't avoid a significant part of stringification cost -- dynamic allocations -- you need to allocate AnyTypeHolder…

Mar 24, 2016
lightmare commented on issue mapnik/mapnik#3384
@lightmare

Hum, I have 32GB ram on my build system. What can I do to build this? Well then the easiest is probably edit Makefile and remove this dependency …

Mar 24, 2016
lightmare commented on issue mapnik/mapnik#3384
@lightmare

There is Ubuntu Precise build on travis -- but with many upgraded packages that are even newer than Trusty's. This is caused by optional sources in…

Mar 24, 2016
lightmare commented on issue mapnik/mapnik#3380
@lightmare

Same here with Linux Mint 17, Boost 1.55 Looks like there should be one travis build with minimum required boost version (1.47?) and one with highe…

Mar 23, 2016
@lightmare
Mar 23, 2016
lightmare merged pull request mapnik/mapnik#3382
@lightmare
Fix reading nodata value for data rasters
4 commits with 6 additions and 9 deletions
Mar 23, 2016
lightmare commented on pull request mapnik/mapnik#3358
@lightmare

I think the run function should be local, too. Unit tests are linked into a single executable. Either make it static, or move to anonymous namespac…

Mar 23, 2016
lightmare commented on pull request mapnik/mapnik#3382
@lightmare

Some minor notes can you move nodata initialization to the variable definition, instead of separate line? consequently, move val inside the loop, …

Mar 19, 2016
lightmare commented on issue mapnik/mapnik#3369
@lightmare

Yes, they're much lower in master already, we made a bunch of changes 2-3 weeks ago addressing this.

Mar 18, 2016
lightmare commented on issue mapnik/mapnik#3329
@lightmare

It's not a compile time check, just missing overload. In my version it's more complicated, as I used inheritance to wrap the sequence of converters…

Something went wrong with that request. Please try again.