Freigeben über

CAT.NET – How Big Do Your Project Files Grow ?

  • Artikel

We are planning a complete re-write of the core CAT.NET engine starting July 1st to get a scalable engine from which we can build on. Building graphs from code is “expensive” hence the OOM issues with the current version. We do it all in memory.. doh! Lot’s of cool features / improvements will be added as well of course.

In order to make sure we are making some right decisions I need to validate our assumptions on the size ranges of web project files. Its far from scientific as the number of methods etc. has a huge impact on the size of the graph generation but we need some basic data to create a range…….drum beat for the “Mini-survey”, please post responses in the comments section.

How big can YOUR web projects grow (in KB’s) ?

What’s an average (mean) size (in KB’s) of projects you want to be able to scan?

Your feedback affects our engineering ;-)

Comments

  • Anonymous
    May 20, 2009
    PingBack from http://microsoft-sharepoint.simplynetdev.com/catnet-%e2%80%93-how-big-do-your-project-files-grow/

  • Anonymous
    June 01, 2009
    What do you want included in the size?  The obvious ones are .cs and .aspx.  What about .resx, .jpg, .css, ... Also, what about referenced projects such as internally built assemblies?  Are those included and if so is the the compiled .dll size or the code size that matters? Some of our dynamically generated data access layer projects are 10+MB in code size (not data).   Our web projects (we have about 20 actively developed) are in the 5MB to 20MB range if I only include .cs and .aspx.

  • Anonymous
    June 01, 2009
    FYI - After testing this on one of our smaller projects that takes up about 7MB of disk space, it caused Out of Memory exceptions on a build machine with 4GB of ram...

  • Anonymous
    June 02, 2009
    Clively Anything that gets compiled really so .resx as well as .cs, .aspx etc For us its the assembly size that is the rough number we are looking at. Our performance is related to code complexity as well as pure code size.