Sure does. Most of us love the stuff, but you mentioned the word packet in your struct, so I wasn't sure if you were doing something really low level. If so, be prepared for latency and lots of slow down if you're expecting high throughput traffic on the order of a device driver, but if you're not, then by all means, totally go xml. This has an added benefit for you Ty. This allows you to then transparently use SSL as the encryption mechanism if you so choose. Take care, Sina -----Original Message----- From: programmingblind-bounce@xxxxxxxxxxxxx [mailto:programmingblind-bounce@xxxxxxxxxxxxx] On Behalf Of Tyler Littlefield Sent: Wednesday, June 03, 2009 9:09 PM To: programmingblind@xxxxxxxxxxxxx Subject: Re: network programming xml makes sense, as it's easier to manage and manipulate, and as mentioned it's flexable. Thanks, Tyler Littlefield Web: tysdomain.com email: tyler@xxxxxxxxxxxxx My programs don't have bugs, they're called randomly added features. ----- Original Message ----- From: "Ken Perry" <whistler@xxxxxxxxxxxxx> To: <programmingblind@xxxxxxxxxxxxx> Sent: Wednesday, June 03, 2009 7:06 PM Subject: RE: network programming > > That is how I do it with my commercial game server however I am shifting > to > using xml as the structure where you can still use the sockets and the > write > and read but the packets are xml for better error checking. The current > way > I do it leaves a mess at the receiving end and you can't make its o > someone > can easily write a client for your game if you want that ability with pure > binary. I mean sure you can tell the person read 8 bites for this and 9 > for > that but it's much easier if they can use an xml parser to get your data. > Not only that but it's easier for you to up keep in the long run which is > why xml structures are becoming the norm. All the boost library does that > I > suggested is packet he data into a xml structure from a class or a > structure > with no work then you send it by horse, space ship or pipe to the distant > end and use the same boost to unpack it. Its fast and takes no work from > the coder. Not only that but as Tialor wants its open source for that > matter I think it's all c++ headers at least most of the library is just > headers. > > Ken > > Ken > > > Ken > > -----Original Message----- > From: programmingblind-bounce@xxxxxxxxxxxxx > [mailto:programmingblind-bounce@xxxxxxxxxxxxx] On Behalf Of Sina Bahram > Sent: Wednesday, June 03, 2009 8:39 PM > To: programmingblind@xxxxxxxxxxxxx > Subject: RE: network programming > > On the other hand, if you're dealing with actual packets, you might want > to > just use the write command and write the binary structure to the socket. > > Take care, > Sina > > > -----Original Message----- > From: programmingblind-bounce@xxxxxxxxxxxxx > [mailto:programmingblind-bounce@xxxxxxxxxxxxx] On Behalf Of Ken Perry > Sent: Wednesday, June 03, 2009 7:14 PM > To: programmingblind@xxxxxxxxxxxxx > Subject: RE: network programming > > > Me I would use boost library and the serialize class that makes it simple > to > turn a structure or class into an xml structure and extract it on the > other > end. > > Ken > > -----Original Message----- > From: programmingblind-bounce@xxxxxxxxxxxxx > [mailto:programmingblind-bounce@xxxxxxxxxxxxx] On Behalf Of Tyler > Littlefield > Sent: Wednesday, June 03, 2009 12:37 PM > To: programmingblind@xxxxxxxxxxxxx > Subject: network programming > > Hello list, > I've got a question. I've been working on a program, and I have a document > that explains the protocol and how things work. > It's about 20 pages so far, but it'll probably get revised to be shorter > once I start doing the actual work and seeing whether things will work. > My question is this: > I'm working with a structure that looks something like: > struct packet > { > MessageType mtype; > int DataLength; > char* data; > }; > How would I go about serializing that so that it'll go across the stream? > I'll need to memcpy the data I'm thinking, no? > I'm also looking at encryption; I need to find some sort of way to let the > client encrypt with the host, without a predefined key; the only way I can > think of is for the host to build a random key, but then it'd have to send > it cleartext to the client, which is rather counterproductive. > > > > Thanks, > Tyler Littlefield > Web: tysdomain.com > email: tyler@xxxxxxxxxxxxx > My programs don't have bugs, they're called randomly added features. > > __________ > View the list's information and change your settings at > //www.freelists.org/list/programmingblind > > __________ > View the list's information and change your settings at > //www.freelists.org/list/programmingblind > > __________ > View the list's information and change your settings at > //www.freelists.org/list/programmingblind > > __________ > View the list's information and change your settings at > //www.freelists.org/list/programmingblind > __________ View the list's information and change your settings at //www.freelists.org/list/programmingblind __________ View the list's information and change your settings at //www.freelists.org/list/programmingblind