I have a couple of open source projects on the go, one of which is a simple Delphi wrapper for SQLite. Lukas Gebauer has now added experimental support for user defined functions. This lets you in effect extend the SQL understood by SQLite to include your own custom functions, written in Delphi.
To try out the feature, download the wrapper and have a look at the file sqlite3udf.pas.
A few links:
Announcement of CTP on Somasegar’s blog
MSDN article on Task Parallel Library
Joe Duffy – concurrent programming blog
Interesting note in Duffy’s blog about PDC 07 (the one that never was):
Note: some might wonder why we released the articles before the CTP was actually online. When we originally put the articles in the magazine’s pipeline, our intent was that they would in fact line up. And both were meant to align with PDC’07. But when PDC was canceled, we also delayed our CTP so that we had more time to make progress on things that would have otherwise been cut.
Concurrency is a big deal; it’s good to see more library support.
Excellent article from ARC on mobile data trends, mentioning that Vodafone has just reported a 50% jump in data traffic:
It is the culmination of attractive data pricing, improved usability and mobile demand for Web 2.0 services which is brewing to form the prefect data storm. As data pricing erodes along the same path travelled by voice, operators must now identify ways to tap into revenues from web services or else be left exposed when the data hurricane arrives.
Personally I don’t need convincing; I’ve been a heavy mobile data user for years; I can’t wait for the data price erosion mentioned. Now we are seeing great little apps from Google (Maps, GMail) and others, better mobile web browsers (iPhone etc) and faster data speeds, and the mass market is waking up to the potential for mobile Internet access. It is taking longer than anyone thought it would, but the trend is unmistakable.
Postscript: see Mobile Web- So Close Yet So Far for a more enigmatic view from Michael Fitzgerald at the New York Times:
For now, widespread use of the mobile Web remains both far off and inevitable.
Note that the piece criticizes the iPhone for not supporting Flash.
The dynamics of this are interesting. Flash is sufficiently entrenched that you can say iPhone is bad for not rendering Flash, not that Flash is bad because it does not work on an iPhone.
Technorati tags: mobile data, vodafone, iphone, google
Red Five Labs has announced its implementation of the .NET Compact Framework for Symbian, the operating system used (most of the time) by Nokia.
It is currently in beta, and implements only version 1.0 of the Compact Framework, but nevertheless it is an intriguing development.
With this, Mono, Moonlight, and Microsoft’s own implementation of Silverlight 1.1 for Mac OS, .NET is getting more interesting for cross-platform development.
Microsoft’s Jason Zander comments on my piece on the early history of ASP.NET:
He says these are corrections though they seem more like supplementary information to me. I don’t have any inside knowledge of this history other than what people who should know say to me (though I do also have my own recollections of what was said publicly). He may be reacting to the idea that the CLR came out of the VB team, which Mark Anders kind-of implied.
One of the reasons I love blogging is that multiple authors can have a crack at getting the facts right. A great personal example is when I asked the question Who invented the wizard; and a good candidate came forward over a year later. If you see something inaccurate or misleadingly incomplete on this site, please do comment or let me know by email.
My comment is here.
Note this debate is not only about the merits of different versions of Javascript/ECMAScript. It is also about power and responsibility. However you spin it, and however far Adobe and/or Microsoft succeed with Flash/Silverlight/AIR, I think we can agree that the browser has an important role for the foreseeable future. It is also likely (though less certain, I guess) that Internet Explorer will continue to have a large market share. The company does have a responsibility not to hold back the Web, and that surely includes not obstructing the evolution of a high-performance Javascript runtime.
It is disappointing that Microsoft says so little about IE8, presuming it exists. If the company sticks by its undertaking to leave no more than two years between IE releases, we should expect it no later than October 2008, less than one year away. It would help web developers to know more about what will be in it.
CodeRage 2007 starts next week. It’s a technical conference covering CodeGear’s products, including Dephi, JBuilder, C++ Builder and 3rdRail, the new Ruby on Rails IDE.
The conference is both free and virtual.
A virtual conference is no substitute for human contact. I’ve learnt this paradox over many years: even if the same content is freely available on the Web, there is substantial benefit in physical attendance. You are more focused, you learn more, you can easily ask questions, and you pick up all those indefinable signals from others who are attending.
Equally, the global fuel crisis and concern about the environmental cost of travel surely means that virtual conferencing is an idea whose time has come. Another benefit is that it includes an array of people for whom a typical tech conference is just not feasible, for financial or other reasons.
I’d like to see more of these.
Thought-provoking paper [PDF] from Daniel J Bernstein, the author of qmail, covering software security and addressing topics such as premature optimization and bug reduction along the way.
In March 1997, I took the unusual step of publicly offering $500 to the first person to publish a verifiable security hole in the latest version of qmail: for example, a way for a user to exploit qmail to take over another account. My offer still stands. Nobody has found any security holes in qmail. I hereby increase the offer to $1000.
He attributes his success to minimizing the amount of trusted code, in contrast to running code with least privilege which he says is ineffective.
(from Schneier on Security).
Yesterday I downloaded the hot new release in the Microsoft development community: Visual Studio 2008.
At least I thought I did. I used the MSDN “Top Downloads” feature, which promises:
… a more direct way to initiate a download of a limited set of selected products
The service uses a plug-in called the Akamai Download Manager. This guy is annoying, especially if you use Windows Vista. First, it doesn’t seem to work at all. Then you realize that you have to disable the IE pop-up blocker. Next, you try to select a download location but it will not let you. It respects some setting in IE that restricts downloads to “safe” locations. You had better have lots of space in your user directory, otherwise this is not going to work.
Fortunately, I do have lots of space, so even the 6GB or so I was downloading should have been OK. I gave in and let it choose the location it wanted. The next thing you see is curious – see here for a screenshot. A message appears telling you the file has been saved (note past tense, though the download is just starting) to the Temporary Internet Files folder, and invites you to open it. I knew the file could not be downloaded yet, but opened it anyway. You get an Explorer window onto a weird location that claims to be in the Windows folder (it isn’t) and shows a single folder labeled C. If you are like me, you shrug, and close it. Don’t do that.
Why not? Well, after several hours or perhaps overnight, the download completes and you look for your files. Where are they?
I looked in Documents, the supposed location. Not there.
I looked in IE’s Temporary Internet Files folder. Not there.
I looked in my Virtual Store, a feature of Vista that supports legacy software which tries to write to locations like the Windows folder. Not there.
I performed a search of my entire User folder, set to show hidden files and folders. Not there.
Before giving up, I opened an administrative command prompt, navigated to the root folder, and typed:
dir *.iso /s
Ah! There they are, in (wait for it):
C:\Users\Username\AppData\Local\Microsoft\Windows\Temporary Internet Files\Virtualized\C\Users\Username\Documents
where “Username” is the current user.
Why didn’t the Explorer search find it? The problem is, you have to have the option:
Hide protected operating system files (Recommended)
set to unchecked in folder options, passing the dire warning that tells you not to do it.
Why do I normally have this checked? The dire warning doesn’t bother me, but I do mind that having this unchecked shows files like desktop.ini on the Vista desktop. Ugly. So I normally have this checked.
Hey, wouldn’t it be good if Microsoft had a single checkbox in its “Advanced” search: to just search everywhere?
So I found the downloads. But honestly, what is this nonsense? The truth is, Akamai Download Manager is not really Vista-compatible; why is Microsoft using it on its premier developer site, for its premier developer product? Ironically, this is the community most likely to be running Microsoft’s latest and [possibly] greatest.
Further, what it is the message here? That Vista adoption is so modest that Akamai can’t be bothered to fix its utility? Or that Microsoft’s own in-house developers can’t build a decent download manager? Or offer to fix the Akamai one?
Excuse my temper. It is no fun to complete a long download and then lose the files.
Update: I also sent a comment and query to the email address given for feedback. It was msdnreply(at)eu.subservices.com. Guess what? Bounced with “User unknown”.
The REST vs WS* wars get ever more interesting, with Dare Obasanjo from the Windows Live team announcing (or confirming) his conversion to RESTful ways, and Project Astoria demonstrating that Microsoft is now building REST services into ASP.NET and ADO.NET (see here for further comment).
Microsoft is still committed to WS*, but equally seems to recognize that much of the world wants to do REST. I’m glad that pragmatism is winning over dogged determination to stand by technology choices.