Aehso’s Output

Google do free lunches for their staff, and allegedly some Yahoo! staff have been taking the piss.

On a slightly related note, I’ve noticed that fuckedgoogle.com seems to have disappeared – well the domain is there but the content is gone. They had been making allegations of click-fraud and then “poof!”.

aehso google

Doogle.org, a simple but excellent tech-parody of our Google overlords. Try a search and see echos of Father Jack magically appear in your query…

(From the kids at the much-more-serious DSG at TCD no less.)

aehso irish, junk

ComReg have determined a formula for the price that Eircom charge other licensed operators (like UTVinternet, BT) for their wholesale bistream product – the formula links the wholesale price to Eircom’s retail broadband product prices. The intent is to stop Eircom from reducing the cost and increasing the speed of their retail broadband offerings while maintaining a static price for their wholesale bitstream product, thereby undercutting their customers/rivals based on cost/performance.

A couple of comments:

• Where is this magic formula? Is it a secret? The one page ComReg press release gives no specifics.
• Eircom are apparently happy with this regulation. Eircom are never happy with ComReg or any regulation – they obviously see a hole in it somewhere.
• It does nothing to address the exorbitant wholesale line rental charge Eircom continue to charge.
• Was it a co-incidence that Eircom announced a pending speed bump to up to 5Mb downstream/512Kb upstream (and murmurs of future ADSL2+ trials) in the same week that ComReg release this report?

aehso irish, net

A French woman nearly disappeared in a real cloud of smoke when she tried to open the door of an airliner so she could pop out for a mid-flight fag. Read the rest here…

aehso junk

The Forgent JPEG patent I blogged about before is being challenged by the Public Patent Foundation.

If you want to see what type of company Forgent really are, all you have to do is go to their home page – 6 of their 7 last press releases relate to this JPEG patent. Obviously not much software development or innovation going on there then.

Ars Technica are carrying a good write up on what is happening.

aehso patents

Just back from the Saving Europe from Software Patents: Why and How seminar, held in the European Parliament Office in Dublin this morning. I thought I’d give you all a quick report.

• Ciaran O’Riordan’s summary of what happened in Brussels when the CCID was being debated was interesting. He talked mostly of how little MEPs understand about both software and patenting in general. it would appear they spent a significant amount of their time explaining what the directive meant.
• The next critical EU directive in this area is IPRED2 – The second Intellectual Property Rights Enforcement Directive (PDF) (FSF Europe’s analysis here). The main goal of this directive is to criminalise intellectual property infringement. The impact on the software development is nicely summarized by this ZDNet article:
  

  There are two main implications of this directive to free and open source software. Firstly, if a free software developer is found to have infringed intellectual property in the code he has written, he could be subject to criminal charges and may not be granted legal aid to defend himself. Developers of proprietary software in commercial settings would be less affected, as the company they work for would generally be held accountable rather than the individual developer.

  I don’t think I need to explain any more.

• The second speaker, Gareth Bowker, illustrated how software patents would inhibit network security. If the network security guys cannot use network analysis tools (nmap etc) because of software patent restrictions then their ability to counter threats from crackers who do illegally use them will be severely limited. Insteresting, I hadn’t thought of that.
• An ISA representative made an attempt to explain why the ISA had supported the CIID and then reiterated their belief that patents were a good thing. To a less than convinced audience I might add, but sure God loves a trier. The panel were asked (by me) if they were surprised that an “Irish Software Association” was supporting the CIID, given the grassroots and SME opposition to the directive in Ireland. They indicated that at the time they didn’t have the time to look into the ISA’ s position but that they might in the future look at ISA’s membership more closely.
• A Microsoft attendee questioned the future of the European software industry without software patents, citing the apparent imminent arrival of hoards of Chinese software companies who will destroy us all if we (and our multinational overlords) have no patents to protect ourselves. The panel politely pointed out to him that Chinese software companies are just as entitled to obtain European software patents as the rest of us and they could use them to against the European software industry (if they were enforceable).
• An attendee from Google Ireland commented that copyright protection provided more protection to software development than is available in many other creative industries. He strongly asserted that software patents are not required. You know, it is very to like Google more when they articluate positions like this, especially when put side-by-side with the common Microsoft/IBM position.
• Two representatives of the Department of Enterprise, Trade and Employment made some comments in an attempt to try clarify the role of the European Patent Office in all of this. (The European Patent Office, despite it’s name, is not a European Union institution – it was established by the European Patent Convention) I’m not sure how much they clarified things but they did raise the interesting issue of how the EU and EPO related to each other.
• Poor Proinsias didn’t make it – he was stuck in Brussels somewhere.

Many thanks to the Irish Free Software Organisation and the European Parliament for arranging a very interesting seminar.

aehso patents

[Update: It just gets worse and worse for Sony/F4I - now they've found GPL (not LGPL) code in the rootkit]

Credit where credit is due, this post’s title was taken from a comment made by Stewart Baker, recently appointed by President Bush as the Department of Homeland Security’s assistant secretary for policy. (I’m not sure about his insinuations that this might make a bird flu pandemic worse but that’s another discussion).

This statement sums up many peoples sentiments on this whole issue. If buying bread and sticking it in your toaster caused every door and window in your house to fall off their hinges you’d be a bit annoyed. You might even think “Hmmm, I don’t like bread that much”.

Unfortunately, many software users will realise that that in may cases the shrink-wrap software licenses (or EULAs) that they casually accept when installing downloaded software gives that software the right to do it wants on the users system. This can include removing files, killing processes, changing type associations (effectively disabling competitors products) phoning home with personal and usage information (hint: RealPlayer) etc. I know people are vaguely aware of these actions, but I’ve the feeling this while Sony XCP story will raise general awareness of just how draconian most EULA’s are.

Sony got caught fair and square and they are paying the price now. However, it does concern me that some are looking at what is happening to Sony right now and they are seeing opportunities. Brand sabotage, fuelled by the blogsphere, could easily be inflicted on a competing brand by paying someone to do some digging and then release the results in some vague, threatening (but still non-defamatory) language.

With the exposed misbehaviour of an industry giant we can all expect more false alarms and scare stories (and the occasional truth) over the coming weeks and months. Just don’t assume all of these alerts will be started by lone tech blogger who notices some new registry entries on their machine.

I’m afraid the days of casually downloading and installing applications from unknown vendors are drawing to a close. Meanwhile, I’ll be consulting tools like EULAlyzer a little more often. The growth of web based applications (or Web2.0 in hype-speak) is perhaps timely in this regard but the browsers and plugins (Flash etc.) will have to make sure they are so bullet-proof or they’ll suffer a similar backlash if they are ever caught in a similar situation.

aehso drm

Since the introduction of the new OSGi core in Eclipse 3.0 I’ve noticed a couple of quirks that manifest themselves when I am developing and testing plugins via a runtime workbench. A couple of days ago I ended up summarizing some tips in an email to a friend but I think they’ll be useful to others too so here you all go:

• While working on plugin source code, do not leave old binary builds of the plugins that you are working into the Eclipse installation’s plugins/ folder. Having the plugin project in your development workspace should mean that the runtime workspace will load the plugin from your project but this doesn’t completely work. From my experience, the runtime workbench will pick up the new classes from your project but the OSGI implementation does not recognise changes made to the MANIFEST.MF file (for example, new bundle dependencies introduced etc). I believe the problem is related to the fact that Eclipse holds it’s cache of OSGI manifest information in <eclipse>\configuration\org.eclipse.osgi\bundles which is being shared by both your development and runtime workspaces (and therefore the runtime workbench “sees” OSGI info it shouldn’t).

  To fix this:

  1. Delete the installed plugins. Let the dev workbench projects supply the code to the runtime workbench.
  2. Delete the contents of <eclipse>\configuration\. Leave the config.ini file there! Alternatively you can run eclipse -initialize from a shell in the Eclipse installation directory. The process should appear busy for a second (as it cleans the configuration caches) and then exit quietly.
  3. Restart Eclipse. The first restart will be a bit slower as it needs to rebuild the OSGi caches.

• The workspace metadata can get out of sync with reality if you manually tinker with the plugin.xml and/or the MANIFEST.MF files in the back-end <eclipse>\plugins\ folder. If you see build errors when attempting to run the Export Plugin wizard that do not show themselves when building inside the workbench then this is your problem. Again, it’s slightly related to the above – I believe the <workspace-dir>\.metadata\.plugins\org.eclipse.pde.core folder contains dependency info that is out of sync with what exists on the file system.

  To Fix, deleting everything in the <workspace-dir>\.metadata\ folder usually works. Deleting the metadata effectively cleans the workspace. Be warned though – the workbench will have forgotten what projects you had in the workspace and what preferences you had changed. Here’s a hint: either get used to the default preferences (they are appropriate for “good” Java development) or get used to exporting/re-importing your preferred preferences set (inc. team repository, JDK info etc.) whenever you clean/create a workspace (see File->Export/Import->Preferences

• Import your projects into a workspace from an “external” folder (a CVS root folder for example). Generally it’s not a good idea to store project source in the workspace folder unless they are temp/demo projects that you can afford to loose at any time.
  Keeping your project resources external to the workspace means there is a cleaner separation between local workspace junk (scrap projects) & state (.metadata) and the more important shared/versioned product source code.
  It also means you can have several workspaces that are looking at the same code base. For example, I have six workspaces importing different sets of projects that are stored in just two CVS snapshots. (I only ever use one or two concurrently!)
• Rather than copying packaged plugins into the \plugins dir, create a <eclipse>\links\my-drivers.link file (A driver is common eclipse.org developer speak for a binary distro of plugins). In that file, add a single name/value ‘path’ property – something link
  
path=C:/build/_output

  Note the mechanism assumes that “_output” above contains an “eclipse\plugins” sub directory that in turn contain the extra plugins it should load (typically your build output).
  NOTE: if you use this you have to be careful – any time your build output changes, you must remember to run “eclipse -initialize” or the OSGI caches won’t be in sync with changes.

aehso eclipse

This is an interesting development. Irish broadband (in cahoots with Intel) are going to roll out WiMAX networks in Athlone, Arklow, Carlow, Cork, Dublin, Ennis, Galway, Kilkenny, Killaloe, Letterkenny, Limerick, Newbridge, Portlaoise and Wexford.

aehso irish, net

A sad collection of letters from partners of MMORPG addicts. Some of the behavior described is disturbing to say the least:

I am sitting here with tears running down my face. I have just printed the basic papers required to file for divorce.

Divorce, my 2nd. First from an abusive raging alcoholic. 2nd? From a once incredibly sweet, kind loving devoted husband who now spends every waking minute playing EQ. I try to think back on when it all went wrong… 2 years ago? 3 years ago? Probably even before that when he was playing other games where they had weekends away for their Quake tournaments.

That’s sad in that she wasted three years of her life hoping the guy would grown out of it. Hindsight is wonderful sometimes. On the other hand, the following excerpt is, well, a different type of sad:

My boyfriend has tried to suck me into his everquest hole before; sometimes it’s funny but other times it really hurts. He would bring this big everquest manual with him everywhere, even on our dates, I guess he thought he could get me to read it.

Yikes!

aehso games