Let's see... ntop runs fine under windows - that's what the #ifdef WIN32 stuff is for. In fact, for a nominal convenience fee Luca delivers a native executable, built under MC V++ .Net through shop.ntop.org.
The projects you are listing (including those you THINK are built via MinGW such as Ethereal (see: http://anonsvn.ethereal.com/ethereal/trunk/README.win32)) - like ntop - all have binary packages available which are built by and run as real Windows executables, vs. the shim layer of MinGW. Anyone using MinGW is - almost by definition - in the hobbiest category. RRD: I've patched ntop for rrd 1.2.x, but at the time RRD itself was going through 1-3 releases per week and it simply wasn't stable enough for ntop to use. In fact, there's data in the FAQ: Q. What about the multi-threaded development version? A. Stay away. (UPDATED) I was able to patch ntop to work with 1.2.7+ and experimented with it a little bit. The binary .rrd file formats are different, so if you try 1.2.x, any new .rrd files which are created are incompatible. The new version doesn't use freetype, it uses libart, which introduces a different (not better, nor worse) chain of dependencies. -----Burton -----Original Message----- From: [EMAIL PROTECTED] [mailto:[EMAIL PROTECTED] On Behalf Of Georger Araujo Sent: Tuesday, August 30, 2005 7:15 AM To: [email protected] Subject: RE: [Ntop-dev] Patches and documentation for building ntop on MingW I'll comment your answers. --- Burton Strauss <[EMAIL PROTECTED]> escreveu: > Georger I appreciate the work you have put in, but it's not a > 'contribution' > to ntop. You have certainly highlighted some issues. But if I were > to more forward with this, all you have done is to make work for me - > and in a platform I don't typically use and thus am not setup to test > in. Hmm, fine. But I think that in the end you'll like the work I'm making for you, because one of the ways to make open source software more popular is to make it run on as many platforms as possible. Like it or not, Windows is more widespread than Linux. I'm running Apache, MySQL, PHP and several other Linux-native apps on Windows, with GREAT results. > > WRT to patches for ntop, there are two ways you can go about this - > > Stage One is to do just enough to make ntop work FOR YOU in YOUR > specific environment. That's great, more power to you, but it doesn't > help anyone else. I think it would be selfish to do so. I contributed back all patches all made for other projects: WINventory, Nagios, TikiWiki, etc. I like to help. > People publish this type of work in the hope - as you indicated - that > somebody else grabs it and continues forward. I hope that happens. I myself am already trying to integrate RRDtool 1.2.11, the graphs generated by 1.0.49 are just plain UGLY. Even though my C/C++ skills are nothing to write home about, I can make a good hack or two. > Stage Two is to then take the next step(s) and to provide usable > patches for baseline ntop. > > Through the lists, I'll work with people (who are capable and > interested in doing the work at stage #2). I'll help point them in > the right directions, tell them 'the ntop way' and provide comments on > interim patches, etc. But that's about the limit. > > The emphasis of stage #2 is on YOU "doing the work". > I don't typically want > to take somebody's effort and do the rework. In general, a > contribution that makes more work for me isn't useful (to me). > > What I want to do is simply take your patch file, apply it and commit, > with credit to the submitter. I always run an eye over ANY > contribution before I commit it. Sometimes if I'm not sure from > looking at the code, I'll lightly test a contributed patch in the > environment(s) I do use. But that's about it. > > You think you are doing #2, but you really are doing #1. > > Your changes work for you and that's great. > > My comments assumed you were interested and capable of stage #2. I'm > sorry only that I misread your interest/capabilities/intentions. "I'm > not a C/C++ coder by trade, I just know enough to fix things I > want/need." pretty well > sets the expectations. Next time I'll ask up > front. > > WRT: "I'm sure that having ntop run on MingW will attract SKILLED > Windows programmers," that's doubtful. There have been a few people > over the years with interest in MinGW, but very few and far between. > MinGW does a creditable job for small codes and for proof-of-concept, > but it's not a production strength tool. The primary audience is the > Unix coder who has to > make 'something' 'work under Windows'. Most > skilled Windows programmers > aren't interested in wasting their time - they would rather be coding > real programs FOR Windows not using a compatibility layer. I don't really agree with you in this. ntop has helped me enormously in finding anomalies in my LAN (and I bet many other people found stunning things as well), and several other programs that employ the compatibility layer - Ethereal, WinDump, Snort, nmap, Cain & Abel - are must-have tools for my day-to-day job. I think you're underestimating ntop. It's a wonderful tool, and deserves more attention and publicity than it currently gets. Regards, Georger > -----Burton _______________________________________________ Ntop-dev mailing list [email protected] http://listgateway.unipi.it/mailman/listinfo/ntop-dev
