Basant Kukreja wrote: > On Sun, Aug 24, 2008 at 05:18:55PM +0100, Peter Tribble wrote: >> On Thu, Aug 21, 2008 at 9:23 AM, rahul <Rahul.G.Nair at sun.com> wrote: >>> Hi, >>> >>> Please find attached the DRAFT ARC case for mod_sed. >>> Do let us (Basant and me) know your comments on this, >> [changed website-discuss to webstack-discuss as I believe that was >> what was intended] >> >> My first question would be - why integrate this? >> >> I've never heard of it; it's clearly marked as being alpha. >> >> Note that I'm not saying that it shouldn't be integrated, just that I would >> expect the case to provide a little more detail on why it's a good thing >> to have and what consumers (if any) of the technology might exist. >> >> My second comment is based on the relative immaturity of the code. >> Why integrate some early unversioned alpha release rather than wait >> for the code to stabilise? > Regarding immaturity of the code, I would like to say that > * This code has been there for around a year now. > * The code has been bundled by apachelounge on Windows and they have done some > testing and reported issues which I fixed. > * I took the test cases from GNU sed and did the basic testing. > * I have also done stress testing and watched for memory leaks and crashes. It > seems to work fine.
I don't think the code maturity is a problem. As the above outlines, (and given the history of a similar module in the Sun webserver) mod_sed is not immature by the standards of newly-released code. Regarding demand, I can say from personal knowledge that there's a substantial demand for this functionality. mod_sed is now the state-of-the-art amongst "sed-like" text filters. My own module mod_line_edit[1] occupied that position for a couple of years, and user feedback (both general feedback and paid consultancy) demonstrated a high level of interest. [1] http://apache.webthing.com/mod_line_edit/ -- Nick Kew
