Boost logo

Boost :

Subject: [boost] [ITL][Review] A late mini review
From: Thomas Klimpel (Thomas.Klimpel_at_[hidden])
Date: 2010-02-28 13:05:12


> Please always state in your review, whether you think the
> library should be accepted as a Boost library.

ITL should be accepted as a Boost library.

> - What is your evaluation of the design?

I haven't spend sufficient time reviewing the library to be able to judge the design. At least, I haven't seen any obvious flaws during my short review.

> - What is your evaluation of the implementation?

I took a cursory look at all source (header) files, and they look nicely written.
I took a cursory look at the directory structure, to verify that it follows boost conventions. It does so in general, but I was surprised to find a "libs/itl/build" folder, since I "know" that ITL is a headers only library. The content of the build folder also didn't make me happy (it's extremely "unportable"), but this is a completely different topic that has nothing to do with this specific library.

The template meta programming in boost/itl/type_traits/* could follow MPL conventions more closely.

> - What is your evaluation of the documentation?

The documentation is nicely written. However, please don't invent new words like "aggrovering" or "unon". Please also try to avoid changing the meaning of existing words like "neutron".

> - What is your evaluation of the potential usefulness of the library?

The need to work with "interval sets" arises sufficiently frequently in "time based" and "geometry based" problem domains to warrant a library dedicated especially to this problem domain. (In my review of Boost.Polygon, I pointed out that interval set concepts were missing.) I have to admit that I haven't spend sufficient time evaluating ITL to be able to judge whether ITL is able to handle all these needs.

> - Did you try to use the library? With what compiler? Did you have any problems?

No, I didn't try to use the library.

> - How much effort did you put into your evaluation?
> A glance? A quick reading? In-depth study?

Some month ago, I read through the documentation just for the pleasure it gave me, pondered a bit about the concept and looked at the BoostCon2009 slide show. I was certainly planning to do a proper review during the review period, but I was (and still am) occupied with other things. What I've done now while writing this review was just a glance.

> - Are you knowledgeable about the problem domain?

This is not important in this case, since I haven't spend enough time for the review anyway.

Regards,
Thomas

P.S.:
As I guessed from recent experience, the review period has just been extended. However, I decided to submit this mini review nevertheless. I may have a look at "I have to admit that I haven't spend sufficient time evaluating ITL to be able to judge whether ITL is able to handle all these needs." with respect to the issue raised by Phil Endecott in
http://lists.boost.org/Archives/boost/2010/02/162198.php
(and write about it in case I can make up my own opinion), because this seems to be an interesting question. Even so I want to thank Phil Endecott for his courage to bring up this issue, and I guess that he is right (in a certain sense), I think it is wiser to base my review only on my personal experiences and not on my judgment of whether Phil is right or wrong with respect to this issue.


Boost list run by bdawes at acm.org, gregod at cs.rpi.edu, cpdaniel at pacbell.net, john at johnmaddock.co.uk