We vastly underestimated the Slashdot effect. There were 85k hits in the first hour since we went live, and then the machine collapsed under the weight and has remained in that state despite repeated attempts to get some data out of it.
None of the betas really had this problem, we got some mild load, but today the load on the machine was insane.
Mike Kestner has posted an update on Gtk# in the release.
I had a great afternoon at Usenix, met lots of good friends again.
Tomorrow: Rob Pike's keynote at 9am.
Posted on 30 Jun 2004
The OMG is one step ahead: they have a RFP for the Minimal CORBA. Beautiful.
Nat's tool was demostrated yestearday at GUADEC, here is a photo from the talk.
Posted on 29 Jun 2004
I have been asked recently whether we should fork the C# language and add extensions and improvements to the language and the class libraries to deviate from Microsoft (there are plenty of little small things that can be done with relatively no effort).
I do not believe that making the Mono C# compiler incompatible with the Microsoft and the ECMA standard is a good idea. A big benefit of having a compatible language is that source code can be interchanged between the two platforms with ease. If we were to modify the language to have features not found in the Microsoft compiler, we would make the life of our end users (software developers) harder for no good reason.
On the other hand, I would encourage folks to prototype their ideas on the compiler: if your idea is great, we could help drafting a proposal and bring that forward to ECMA and Microsoft, and who knows, it could even become part of the standard.
I personally did not like the C# language support for nullable types I felt they had no place in the language, and I think a lot of people feel the same way.
I changed my opinion on its usefulness when I explained them to a couple of friends: they immediately saw the benefits to them. It was clear that my programmer background was not the same programmer background that nullable types addressed.
Posted on 28 Jun 2004
A good friend was asking me about why I had abandoned CORBA (I did not). Here is my response:
On the CORBA question: I have not particularly abandoned the idea of CORBA. It was just that CORBA as a platform to solve the multi-lingual issue was too hard (too huge of an investment, steep learning curve). In fact, I would like to use more of IIOP as a replacement for remoting and web services in key places.
I think that *some* elements of CORBA are extremely useful and are better than many of the Microsoft proposed APIs/protocols, but the problem with CORBA is that you must move carefully: it is too easy to get trapped in that world of standards and try to implement and use everything.
Some bits of CORBA are extraordinary, but teaching it is hard given the large scope of it: what pieces of it must be ignored and what pieces are followed is a tricky part.
I think the Next Generation Web Services (WS-*) world has reached the same level of complexity that CORBA achieved in an effort to satisfy the users requests. It seems that WS-* is a victim of its own success: once SOAP web services could be produced and consumed easily with `netcat', but today, realistically to integrate with any of the advanced features you need a stack as complete and as complex as the various CORBA implementations.
Web Services were pushed in the same direction that CORBA was used, and it started to get used in places where CORBA was used, and they ended up with pretty much the same thing CORBA had, and sadly without some of the benefits of it. Not a particularly fascinating subject.
I first heard about Ice from Vladimir, who wrote his own implementation of the protocol in C#. It is available from the Mono CVS repository, and is a subset the features offered by ZeroC's product. ZeroC's product is dual licensed: GPL and Commercial and today they support Java and PHP with a C# version in the works.
Posted on 27 Jun 2004
Having being subject to three years of propaganda, this movie feels like an Oasis.
Too short, seems like a lot of interesting material was left out.
Posted on 26 Jun 2004
Various reports have arrived that Microsoft was demoing Mono at the LinuxDays. I do not have more data other than a Windows machine was used to build a web application, copied to Linux and ran there and that they also showed Monodoc.
Got tickets for the worst possible time for Farenheit's 9/11 showing tonight. I underestimated the demand, all the other showings were sold out.
The Mono Freeze has continuted in preparation for the 1.0 release, everything is looking good.
Posted on 25 Jun 2004
My friend Jon Perr was running a political commentary site, called Perrspectives and using Google adwords to advertise it.
Here is the tale of the adwords censoring his site.
Posted on 23 Jun 2004
One of the best Daily Shows today. Jon Stewart was amazing.
Lame Blog now supports Atom using Atom.NET. It supports Mono out of the box (yup, nice Makefile and all), and comes with awesome docs for it. Am still an Atom newbie, so I do not know if I got everything right, drop me a note if my Atom stinks.
It took 10 minutes, so do not expect Enterprise-enabled, transactional based atom feed just yet.
Posted on 21 Jun 2004
Cory from the EFF went to Microsoft to talk about DRM, here is a transcript which should be mandatory reading.
Joel keeps pumping out the gems.
A key to the proposal from Joel is that some high-level features must be incorporated into HTML in some form, and not be limited to server-side controls implemented with a particular technology (ASP.NET or Struts or PHP): it should be something that can be placed on a static HTML page.
Posted on 19 Jun 2004
My favorite conference is coming to town in a couple of weeks. Hope to see a lot of folks
Martin has been doing quite a lot of work on the Mono Debugger, getting nicer every day. And it has docs ;-)
Posted on 18 Jun 2004