A bug in Vista’s UAC

Vista’s User Account Control (UAC) elevation prompt sometimes appears when it shouldn’t. Here is an example which works every time for me. I have a folder in Documents (within my home directory) called recordings, containing MP3 files. I double-click one of these files and it opens in Windows Media Player. Now let’s say I try to rename the recordings folder. I get a dialog saying I need to confirm this operation, complete with a UAC shield.

I click Continue and get the screen flash and the elvation prompt. It’s not enough though; I now get a dialog that says “Destination Folder Access Denied – You need permission to perform this action.” If I click Try Again, I get the same dialog, for ever.

It’s nonsense of course. I don’t need permission; I need to close the MP3 file. Indeed, if I close the file I can then rename the folder.

Oddly, if I try this with a Word document, Vista correctly reports that the folder in in use by Word. But if I try a second time, I get the inappropriate UAC prompt.

It’s possible that some Windows API call is returning an access denied error, when it should indicate a file or folder in use by another process, or there could be some other explanation. The end result is a poor user experience.

If Microsoft can’t always get this right, it illustrates how hard it is for developers to give users appropriate error messages when working with UAC on Vista.

 

Technorati tags: , ,

Google apps in the real world

Or at least the semi-real world: Wired’s Michael Calore spent a month working (mostly) with Google apps rather than his usual desktop software (on a Mac). I’ve thought of trying this same experiment myself but haven’t yet felt that it is worth the risk.

A few points interested me. First, that he could live with the apps themselves, but ran into interoperability problems. One that surprised me:

One of our copy editors couldn’t open some docs I had exported, so I was forced to copy and paste those articles into Microsoft Word just for her.

I’d have thought the other way (Word to Google) would be more difficult, because of missing features. And Google’s response (product manager Jonathan Rochelle):

It works best when everybody in the group is sharing on the same platform,” he says. “The experience you’d have if you were just sharing stored docs rather than your co-workers asking you to save down to the desktop would be much closer to ‘Wow, this is an incredible product’ instead of ‘Wow, this really stinks’.

Oh, so if we all use Google we will be fine. That’s no better than Microsoft telling us all to use Office. Nevertheless I take the point to some extent – this web collaboration thing really only works if everyone plays. That doesn’t excuse the compatibility issues.

Calore mentions the privacy aspect, but more needs to be said here. I have no problem with internet storage; for example, I’m happy to save stuff to Amazon S3 without worrying that Jeff Bezos will start poking through the data. In fact, I consider S3 more secure than just saving files to a typical Linux box out on the Internet, especially shared hosts. Google troubles me though, because its business model is contextual advertising and you agree to let it mine your data, albeit with self-imposed limitations.

Finally there’s the question of whether the apps themselves are good enough. Again, I’d have liked Calore to have said more about this, though he reports how much he misses drag-and-drop. The impression I get is that the browser-based apps were a bit frustrating, but he says:

Eventually, I learned to accept that the browser had certain performance limitations that I would have to live with in exchange for the convenience of centralized storage and easy access. Rochelle says it’s just the way our brains are wired from decades of using desktop apps.

Calore has missed a point here – you can have centralized storage and easy access without necessarily using browser-based apps. Moving the server to the cloud is spot-on, but the case for solely browser-based apps is weaker. Its main advantage is zero install; but that’s the way desktop apps are going as well.

 

Using WordPress pages

Yesterday I posted an article on Office Open XML which is too long for a blog entry. Rather than creating a separate HTML file I used a WordPress page entry. WordPress pages are authored in the same way as blog posts, but are not part of the blog itself; they “live outside of the normal blog chronology.” You can organize them into a hierarchy of pages and sub-pages; they are important because they make it possible to build an entire web site in WordPress, using it as a simple content management system.

Curiously the page template in many WordPress themes omits comments. This caught me out: I marked the page as enabled for comments, but no comment form appeared. I fixed this by adding the following line to page.php:

<?php comments_template(); ?>

I’m now happy with the result and will probably use WordPress for further longer articles. In fact, I’ve already added a further page, this being my blog archive. When I migrated from bBlog to WordPress, I left the old blog engine in place so as not to break existing incoming links. However, although the old entries were still in place, most were left with no index link; they were effectively invisible. The new archive page fixes this; you can see all the posts since I started blogging in 2003: errors, insights and all.

 

Technorati tags: , ,

Mono’s new GUI library: how is this hype justified?

Mono’s Miguel de Icaza reports on a new GUI library for Mono:

We have been working on a technology that we believe will revolutionize user interfaces. Today we are announcing the response to Microsoft’s WPF/XAML, a response to Flash and WPF/E. A cross-platform GUI toolkit (supports Windows, MacOS and Linux and is easily ported to new platforms) written entirely in managed code and 100% open source.

There is some skeletal documentation for gui.cs.

Maybe this will be fantastic, but for the moment this looks like an early prototype, not a revolution in user interfaces. How is this hype justified?

Personally I reckon Mono will have to implement WPF eventually, if it wants to keep in touch with what is happening with Microsoft .NET.

 

Technorati tags: , ,

WPF/E is now Silverlight

Microsoft’s Flash alternative has a new name: Silverlight. Undoubtedly a radical shift in naming conventions. Back in 2005, Microsoft renamed Avalon to Windows Presentation Foundation, and I noted that:

These new names seem to be deliberately chosen to be forgettable.

Now we have memorable back. Interesting.

The name may be there, but the product is still in preview; the latest release remains the February CTP, with full release promised before July 2007. The emphasis is on video and vector graphics; there’ s no common language runtime implementation yet. You can write Silverlight code in JavaScript. It’s cross-platform, but currently only supports Windows and Mac; no device support yet. See the faq here for more information; and a useful summary from Tim Sneath who says there is big Silverlight news to come at Mix07.

 

Technorati tags: , , ,

Microsoft’s Jean Paoli on Office Open XML

I spoke to Jean Paoli about Office Open XML and its standardisation. I respect Paoli, one of the originators of the XML specification. His major point, apart from complaining about what he calls IBM’s orchestrated campaign against the ISO standardisation of OOXML, is that only Microsoft’s XML format can maintain fidelity with legacy Office documents. Unfortunately the example he gives – borders around a table – is not often a critical feature; but in general I take the point. He seemed not to understand my question about whether there will be a non-MS Office reference implementation.

Leaving aside OOXML vs ODF for a moment, Paoli observes that “The responsibility of migrating 450 million users is huge.” He is talking about the decision to make XML the default format in Office 2007. Undoubtedly a brave move, and painful for users in some cases, but for developers the ability to work with XML (whether it is OOXML or ODF) is a joy compared to the old binary formats, or Word’s Rich Text Format.

Technorati tags: , , , ,

Will CDs become worthless?

I often see classified ads for CDs with a stated reason for sale like this one, plucked from eBay today (emphasis mine):

CD & INLAY EXCELLENT. CASE SCRATCHED BUT NO CRACKS Having a clearout as gone digital. All CDs genuine and in very good condition unless stated.

Now, one can only speculate about the meaning of “Having a clearout as gone digital”. Perhaps it means the person has purchased all the songs they want from iTunes or similar, or has a Rhapsody subscription, and is now selling off CDs they no longer use. Alternatively it might mean these CDs have been ripped to PC or Mac and are now equally redundant, though one can question the legality and/or ethics of ripping a CD, selling it, and continuing to enjoy the music. In one sense it matters little; the question that intrigues me is what effect this activity has on the secondhand market. If enough people follow suit there is going to be a huge excess of supply over demand.

I can think of several reasons why CDs might remain desirable in a music server household:

  • As proof of ownership of a some kind of licence to rip
  • So you can admire the sleeve and read the booklet
  • To obtain that special mix or mastering that isn’t easily found online
  • Because you can’t bear to admit that your expensive CD player no longer has a useful function

Now ask yourself how much any of the above matter to the average person.

Collectors will still collect, of course. But my advice to anyone contemplating the sale of their CD collection is: do it soon.

 

Technorati tags: , , , , ,

Official performance patch for Outlook 2007

Computerworld has drawn my attention to a new performance patch for Outlook 2007, issued on Friday. Here’s what Microsoft says:

This update fixes a problem in which a calendar item that is marked as private is opened if it is found by using the Search Desktop feature. The update also fixes performance issues that occur when you work with items in a large .pst file or .ost file.

The patch is welcome; there’s no doubting that Outlook 2007 has proved horribly slow for many users. But does it fix the problems? If you read through the comments to earlier postings on this subject you’ll notice that there are actually several performance issues. The main ones I’m aware of:

  1. Slow receive from POP3 mail servers. Sometimes caused by conflicts between Vista’s TCP optimization and certain routers – see comment 27 here for a fix.
  2. Add-ins, for example Dell Media Direct, Acrobat PDFMaker, Microsoft’s Business Contact Manager. See Tools – Trust Center – Add-ins and click Go by the “Com Add-ins” dropdown to manage these.
  3. Desktop search indexing. You can disable this (it’s an add-in) but it is a shame to do so, since it is one of the best new features.
  4. Large local mailbox – could be a standalone .PST (Personal Store), or an .OST (Offline Store) that is kept in synch with Exchange.

The published fix appears to address only the problem with large local mailboxes.

Does it work? I’ve applied it, and it seems to help a bit, though I reckon performance remains worse than Outlook 2003. My hunch is that the issues are too deep-rooted for a quick fix, especially if you keep desktop search enabled. I’ll be interested to see whether the patch fixes another Outlook 2007 annoyance: if you close down Windows while Search is still indexing Outlook, you almost always get a message saying “The data file ‘Mailbox …’ was not closed properly. The file is being checked for problems. Then, of course, you wait and wait.

Is it our fault for having large mailboxes? Here’s a comment from Microsoft’s Jessica Arnold, quoted in the Computerworld article referenced above:

Outlook wasn’t designed to be a file dump, it was meant to be a communications tool,” she said. “There is that fine line, but we don’t necessarily want to optimize the software for people that store their e-mail in the same .PST file for ten years.”

A fair point; yet quick, indexed access to email archives is important to many of us. Archiving to a PST is hazardous, especially since by default Outlook archives to the local machine, not to the server; and in many organizations local documents are not backed up. Running a large mailbox may not be a good solution, but what is better?

Perhaps the answer is Gmail, if you are always online and can cope with the privacy issues. Note the first selling point which Google claims for its service:

Fast search
Use Google search to find the exact message you want, no matter when it was sent or received.

Apparently Google understands that users want to be able to find old messages. Surely a desktop application should be at least as good for finding these, as an internet mailbox that might be thousands of miles away?

Update: I still get “The data file ‘Mailbox …’ was not closed properly.” Not fixed.

See also http://blogs.msdn.com/willkennedy/archive/2007/04/17/outlook-performance-update.aspx where a member of the Outlook team further describes the patch.