on 6/1/00 9:48 PM, Steven Fines at itzamna at sunflower.com wrote: >> > > Ian, > It looks like there is no Freenet.crypt.ciphers.Rijndael class. This is > causing > the cipher_ctx=Util.getCipherByName(cipher) (method generator_init(String > cipher)) to return null. If you build the Rijndael*.java files, it starts up > Ok. >
That will allow freeloader to start but it does not support crypto and the document concept. The cli clients got a bit mish mashed and SimpleClient was forgotten when we added the concept of documents and crypto. Brandon was talking about fixing the that up .... on 5/24/00 1:46 PM, Brandon at blanu at uts.cc.utexas.edu wrote: > > I'm going to spend a couple of days reworking the cli clients, so if > anyone plans on messing with them, please let me know now or wait until > next week. It will be messy to merge changes because I'm going to > restructure the code. > but we have not heard from him. I was thinking about fixing up SimpleClient after Brandon restructured. Unfortunately, I will be going on a trip for a few weeks, so if this SimpleClient part wants to get done soon someone else will need to take it on when we see Brandon's work. Here was my recommendations for Brandon's part: on 5/26/00 1:50 PM, Paul Kappler at kapplepc at rose-hulman.edu wrote: > We need to move the code that pertains to constructing and reading the > encrypted metadata file out of the main function of the clients. I think > the encryption algorithm and content type should be a parameters of insert. > > We need to move of the Transport concept ,metadata file concept, Document > concept, and the cypher key concept inside the insert and request functions. > > I would like to establish a design principle that will make it easier to > write clients for Freenet: It should be possible to search for files, insert > files into, and request files out of Freenet with a single function call (or > as few calls as possible). One other issue I see with the cli clients now is that we are using try catch exit too many times. I think in most cases that should be handled by a single big try block. The whole idea of try is to allow you to put all the error handling at the end. I hate to be all talk and no action. Sorry to disappoint, Paul _______________________________________________ Freenet-dev mailing list Freenet-dev at lists.sourceforge.net http://lists.sourceforge.net/mailman/listinfo/freenet-dev
