[Networkit] All pairs shortest path

Christian Staudt christian.staudt at kit.edu
Wed Dec 14 20:00:35 CET 2016


Matteo, thanks for your feedback sandwich. ;)

> On 14 Dec 2016, at 17:50, Matteo Riondato <matteo at cs.brown.edu> wrote:
> 
> Let me play the devil’s advocate here, and say that NetworKit doesn’t have the technical debt in check.


To discuss this we would need a good definition of bankruptcy. The agile methods expert defined this kind of bankruptcy as “all developers are completely busy fixing bugs so that they cannot add new features, even though they are demanded”. So that is driven by user demand and feedback, of which NetworKit has very little at the moment.*


> The API organization in modules is a bit of a mess.


That’s true, reasons are:
- it’s mostly “organically grown"
- refactoring is a pain (see lack of tool support)
-  there are few incentives to clean up until someone finds a way to publish papers on good library software design.
- ….


* ...being a bit of a collection of solutions in search of problems. My favorite example is an algorithm for layouting a complex network with millions of edges in 2D - just think about it. :) Can you come up with an example of how to use this to solve a real problem? If we are naughty we can say that NetworKit developers are building algorithms the wrong way around and if we are nice we can say that they are pursuing a technology push model…


-------------- next part --------------
A non-text attachment was scrubbed...
Name: signature.asc
Type: application/pgp-signature
Size: 495 bytes
Desc: Message signed with OpenPGP using GPGMail
URL: <http://lists.ira.uni-karlsruhe.de/pipermail/networkit/attachments/20161214/7eddb15a/attachment.sig>


More information about the NetworKit mailing list