Re: [boost] Formal Review of IO and Toolbox extensions to Boost.GIL starts TOMORROW

Subject: Re: [boost] Formal Review of IO and Toolbox extensions to Boost.GIL starts TOMORROW
From: Christian Henning (chhenning_at_[hidden])
Date: 2010-12-04 11:21:38

Hi Phil, good to hear from you.

On Fri, Dec 3, 2010 at 1:54 PM, Phil Endecott
<spam_from_boost_dev_at_[hidden]> wrote:
> Hi Christian, Mateusz and others
>
> Mateusz Loskot <mateusz_at_[hidden]> wrote:
>>
>> According to the Boost Formal Review Schedule [1],
>> review of Christian Henning's extensions to the Boost Generic Image
>> Library (Boost.GIL, [2]), it is:
>>  * Boost.GIL.IO
>>  * Boost.GIL.Toolbox
>> starts on December 1st and lasts until December 10th, 2010.
>
>> The boost_review.zip is about 20MB due to its extensive collection of
>> test images. They are part of the test suite to make sure that different
>> variations of each image format is read and written correctly.
>
> It looks like these test images are not included in your Google Code
> repository, which is not unreasonable; how do you plan to handle this in the
> Boost repository, if accepted?  Are there any precedents elsewhere in Boost?

Yeah, I separated them for the time being. I don't think it's big deal
to add a bunch of test images to the boost source tree. It isn't
really that big after all.

When running the test suite I write out a lot of pictures which
amounts to 5MB. But this is only done once since the files are being
overridden at the next run.

If there is resistance to add 20MB in test images I would have to
resort to images in memory ( std::stringstream ). I cannot recommend
such approach.

The biggest problem I can see is the configuration of the test suite.
There are some manual steps involved, as they are described in the
readme.txt. I need to talk to the boost build guys to get the
configuration done automatically. But that's for after the review.

>
> Also, can you point out which images and tests exercise the error paths that
> we discussed at length (e.g. setjmp/longjmp) a few months ago?

Every time you read/write a jpeg or a png you use the setjmp/longjmp
combination. I have tested it on my own machine and it works pretty
well. There are no tests that are targeted specifically at
setjmp/longjmp.

Does that answer your question?

Regards,
Christian