I post this here, because it probably isn't an issue for the MeAggregator not Ajax Extensions, however a glitch on my computer, which could have importance for others.
Yesterday I had problems with the code he had committed showing a multitude of errors. I notified him from home, and he told he that the client view, that hehas developed, is using the Ajax Extensions for ASP.NET. This shouldn't be a problem as Mono has an implemented version, and MS provides it in .NET3.5 and has installers available for older versions.
On my home computer this library was already installed, and I jumped up to test it. It worked – celebrations!
This morning I began installing the library on my work computer, but it turned out I already had it installed!! Checking the GAC, for sanities sake, it was there. But VS2005 couldn't find it!!?! I unstalled the library and re-installed – no luck!
So in desparation I went into the references in Rob's code, removed the reference and re-added it. Now it is found!!
So it works – I think – but this is a bit of a scare, and I've written this for future reference…
I just had a spam comment on one of my blog entries. We probably ought to look at tightening up the captcha
So I attended an un-conference for the first time in my life. I’ve always been intriged about the concept, both because it seems odd that it should work (no agendas! Can academics really do that?!), and because I would like the concept to work.
I’m pretty sure now that the concept of un-conferences can work. It certainly did for us. We had a very free flow of interesting ideas and discussion of them in the two small groups (at least in ours, and from the sound of it also in the other) I think the reason it did work are manyfold, but especially because we were all interested in the same area of digital identity, and reasonably open in sharing experiences. I liked that the fast exchange and flow almost made it into a brainstorm session.
This fast pace was hampered a bit in the afternoon, because we decided to go with only one group. It did work, however I found it a bit frustrating at times, because it was difficult to air one’s own opinions and ideas. Some of the intimacy of a small group was gone too. For instance when sharing something personal (and we did, afterall we were using examples from our own identities) in a small group, it feels easier to do, because you have more time to elaborate and discuss it, whereas in the bigger group it seemed “digested” faster, thus perhaps scewed a little bit from the intention of the “sharer”.
My brain was full after the day was over, which is a very good sign, and I want to have more of these kind of sessions!
I think I've found an error in the Gtk library. I'm using a Gtk.TreeStore to store information, and as this could be many different kinds of objects I set the type to object. But nothing would work properly. I'd spent the better of 6 hours toying with all kinds of setting, before I tried changing the type to string and use the tostring on the different objects. Voila now it stores everything…
This is bizarre, because the documentation states that any object can be stored in the storage… At the moment I can't be fussed to look into the library to find the cause, but I write this merely so that I might remember 😉
Not sure if this potentially is an issue:
JabberIDs are case in-sensitive, whereas URIs are case sensitive!
I tried a testcase where a user was called systemUser (guess what he had rights to do? 😉 ), and when using that URI in couldn’t authenticate. It has taken me at least 45 minutes to find the when calling XMPPconnection.getUser() in the smack library the username is converted to lowercase. I use this method to add the user to the list of users with system rights, so when authenticating the JabberID systemUser, then this wasn’t on the list!!
It is obviously easy to hack this problem by checking lowercase versions of IDs,, but this is an awful hack, which might break, unless we specify that MeaUserID == jabberID and must be unique in lowercase versions.
Is this a fair assumption? I don’t know…
(EDIT: I’ve added this to the issues list at www.meaggregator.org)
I thought I finished a big part of the MeAggregator FFS authentication scheme. Just to end up in an exception being thrown. I was puzzled, how could this happen, as I had been testing all the way so that each bit I changed didn’t change any behaviour in the FFS?
The answer: Well, although I kind of knew this was the case I’d hope the jabber account adresses could work as URI. According to the URI spec (RFC3986) URIs start with a scheme + “://”, such as “http://” or “mailto://”. Therefore I’d testedthat Jena could handle jabber account without scheme and it does. But when I serialize the model into a string (where I stream it), which is the last thing that happens when exporting a model the exeption is thrown.
Strange that Jena allows usage of URI all the way until the model is streamed! But hi, I can add a scheme to the jabber accounts, just a bit annoying now that I thought it worked… Luckily this was an easy bug to find 8)
From the outset this library (Soapbox from Coversant http://developers.coversant.net/) looked like the most promissing .NET library for the XMPP functionality of the MeAggregator. Unfortunately my beliefe in it has fallen to a low point. Although it is very extensive in its supported functionalities, there is one that is really not that useful. The Jabber ID (JID) is tightly locked to the URL of the Jabber server that you use, hence you can't login to one Jabber server using a JID from another server. This is a functionality that is used (I use it often myself) thus this library can't be used for the MeA work!
I've spent 1/2 day now looking at the possibilities of changingthe behaviour, but there seems to be way too many references to the fact that server is part of jid, and that the login server is that server.
Garry's Mod is a Half Life 2 (HL2) modification, that turns HL2 into something that I can best describe as a big box of lego. Suddenly you stand with the tools to sprawn new object, attaches rules of physics, engines, balloons, well almost anything to the objects and people from the HL2 games.
There are absolutely no rules, objectives or statistics, exept for the ones you create yourself. It is only your own imagination, scripting cababilities and artistic flair that set the boundaries. You can create even multiplayer scenarios.
Now, I've only tried this game for the first time this lunchtime, which turned into a lunchtime and a bit. At this point in time I'm facinated and can see a lot of posibilities. However I'm not sure if this facination will keep me building stuff in the long run. There are plenty of games and puzzles created by the user community, but I'd like to create something myself. Will I do that, or will this just turn out like the longtime want to create a "proper" computer game, that just keeps being taken over by other commitments like job and family life?
Pages
Blogroll
Tags
.NET Android Android gamedevel Android game development Android tutorial anti-google Blackboard BoB Bug Christmas competency maps export facebook Firefox gamedevel game development Google java Linux challenge MeAggregator Microsoft Mono MS New computer odinlab ontology patent proxy rant review rss rss-feeds Rune Escape Scoreloop search desktop secondlife security spam trace VS2005 widget Windows challenge Windows game development XMPP XNASearch Me!
Archive
Site admin