search

Scan-o-Matic / scanomatic

Scanomatic
GNU General Public License v3.0
10 stars 4 forks source link

Converted images in unit tests to compressed tiff #238

Closed skymandr closed 6 years ago

skymandr commented 6 years ago

Used command: convert test_fixture_easy.tiff -compress lzw test_fixture_easy.tiff

Did not touch image in integration test for now, for no apparent reason. Could convert that one as well.

stickler-ci commented 6 years ago

Could not review pull request. It may be too large, or contain no reviewable changes.

stickler-ci commented 6 years ago

Could not review pull request. It may be too large, or contain no reviewable changes.

coveralls commented 6 years ago

Coverage Status

Coverage increased (+0.005%) to 39.995% when pulling 8106f1d1284e93748769180b7c5bb68008cee091 on use_compressed_images_in_unit_tests into 6fdc5223eb420bda3dcf6182ff6557cd1d0a23b8 on master.

gdetrez commented 6 years ago

Is there a point beside saving a few mb (although at this point it won't save anything...)? And are the images generated by the scanner compressed? I'm worried that this makes the inputs for the tests systematically different from the runtime condition without obvious benefits.

skymandr commented 6 years ago

At the moment no, no point. I made a PR mostly to prompt Travis to run the tests for me, since I couldn't get it to work on my machine, and to share the results. =) Experiment successful, I'd say, but I think this can safely, and probably should, wait until we've implemented compression in the image acquisition (should be simple). The image data is identical (the compression is lossless, and I checked to be sure), but different parts of libtiff are of course used, and I think you're absolutely right to not want that kind of inconsistency in the code base just to save a few Mb.

I'll withdraw the PR now that I know you've seen it! ^^

On 8 December 2017 at 10:08, Grégoire Détrez notifications@github.com wrote:

Is there a point beside saving a few mb (although at this point it won't save anything...)? And are the images generated by the scanner compressed? I'm worried that this makes the inputs for the tests systematically different from the runtime condition.

— You are receiving this because you authored the thread. Reply to this email directly, view it on GitHub https://github.com/Scan-o-Matic/scanomatic/pull/238#issuecomment-350212179, or mute the thread https://github.com/notifications/unsubscribe-auth/AAV41suDTLCsKLPeLOkNV0gp5Kw6_dOHks5s-PyWgaJpZM4Q5XxJ .

-- Andreas Skyman, Ph.D. Data scientist

http://www.molflow.com/

Möller Data Workflow Systems AB (Molflow) Kungsgatan 56 411 08 Göteborg SWEDEN

phone: +46 (0)707 - 40 15 92