The writings on this site are representative of my views and opinions (and those who comment) only and in no way reflect those of my employer.
This site is powered by
Movable Type 3.15
Is it just me or is this RSS Reader the coolest ever use of javascript?
(Works in Opera 7.01, I didn't try it in other browsers...)
I won't be at TechEd Australia but ill be looking forward to reading what the bloggers have to say about the event. Is anyone setting up TechEd Bloggers for the Australian TechEd?
Heading off to the Something for Kate concert tonight @ The Rosemount here in Perth. Should be a great alternative rock'n'roll evening :)
Check out Something for Kate! [ed: fixed link]
Hope everyone going to the Silicon Valley webloggers dinner tonight enjoys themselves as much as I do.
[Update: If you have come from Lockergnome you might want to know that v1.0 of mReader is deprecated and version 1.01 has been available as of 23rd August 2003]
Okay, I have put a version of the mReader jar online. To get a copy of it point your J2ME midp enabled devices web browser to http://mobile.markallanson.net, alternatively the direct link to the jar file is http://mobile.markallanson.net/j2me/mReader.jar.
The jar file is 42kb in size, most of this comes from the kXML2 and XmlPull libraries I am using for XML processing. I might try and cut out a bunch of functionality from these libraries in the future to make the jar smaller.
Features so far
1. Processing of RSS 0.91, 1.0, 2.0, RDF. In case you are interested, the processing for each of the above formats is identical.
2. Importing of an OPML outliner file from the web. For SharpReader users (and any others who use a reader that can export to OPML), ftp your exported OPML to a website and point mReader to it to download a list of existing feeds.
3. HTML tags are stripped out of post descriptions. Later versions will have a "link list" functionality for a post so you can see what links the post contains.
Limitations/Annoyances at the moment
1. You can only store up to 50 Feeds in your feed list.
2. mReader will only allow you to view the first 50 items in a downloaded feed.
3. Once a feed is in your list, you can't remove it. (this will be fixed in a later version)
4. mReader doesnt handle some feeds too well due to the way I currently make use of the kXML2 parser. If you try to load a feed and you get a blank list of items, then mReader has most probably had an error trying to download. Email me a copy of the RSS file (or weblog URL) and I will see whats going on.
5. mReader does not display any "I am currently downloading" type messages. Once you have selected a Feed, wait a while, I can assure you it hasn't crashed!
6. mReader currently doesnt look at content:encoded fields for post text. it uses the description tag instead. For memory reasons.
I have currently tested this jar on the default color phone provided with Sun's Wireless toolkit, aswell as the Nokia 7210 emulator (the phone I use).
Note that there is currently minimal error checking in the app, so you may get an exception every now and then.
If you have any problems, please email me the details.
mReader, the J2ME app I am writing to read news feeds from any midp java phone is coming along nicely. You can now add weblogs to your list (that is stored between sessions), and fetch weblogs to get a list of articles (these are not stored between sessions, too much to store on such devices), then view articles.
It is currently supporting RDF and RSS 2.0, but I will be adding the other popular formats soon. All I have to think about now is bandwidth constraints. The app is currently 54kb, which I should be able to shrink down by stripping out bits from the kXML and XMLPull libraries i'm never going to use (like serialization).
The app size is not my biggest bandwidth concern however. Larger feeds could become expensive to download especially if you are viewing frequently. I might try to implement some caching so that you can flick between feeds on a per session basis and not incurr extra download charges.
Ill release a beta version to the web sometime in the next week.
Paul V says in this post that this article contradicts itsself by claiming that VB is not a better OO language than C#, but later in the article goes on to comment that VB is quicker to develop in than C#.
Since when is speed of development the only indicator as to what language is better than another to use for development of a particular product?
Im working on mReader, a RSS Aggregator for J2ME enabled devices (mobile phones etc).
What do people want in an aggregator of this size. It can't store feeds, so it has to download every time, which could be a killer on download charges. You will be able to point the reader at an online OPML file though so you can select which feeds you want to download, for example, save your sharpreader opml to a website, and point the phone at it to get a list of contacts.
I just scored "16.17357% - Geek" on the geek test.
Was that Longhorn I spotted on Scoble's monitor in the background of his photo for this article in the Seattle Post Intelligencer?
Over the past few days at work I have been involved in the testing of an interop module that links our Point of Sale system to a 3rd Party Accounting and Hotel Reservations system. Development for this interface took place on our behalf a number of months ago.
The unique thing about this particular interop process is that the owner of the company we are developing with wouldn't give us first hand access to the development staff of his product. Any liaison between ourselves and his development staff had to be piped through him to "protect" his development staff from distractions. (gee, they must be a pretty distractable mob)
This was unusual, but acceptable, as they provided us with a simple spec of the interface and data they expected from our system. The spec was programmed, with a few questions being piped through our human interface with their development staff. An initial testing phase was planned.
The interop works in duplex mode. We have to retrieve configuration data from their system, and we have to write sales information back to their system. Initial tests seemed fine, our system was retrieving all configuration data correctly. Our first sales export seemed to go okay, with a bit of tweaking.
We are then presented with a 20-30 page document. This new document (which seems like a bit of a novel), documents the rest of the interfaces we have to test. In short, 2 brand new interfaces were presented to us. The document is dated 4th July, 2003. Confusion crosses their faces, as we explain to them that we had not received this new specification. Aparently their development staff were supposed to forward this specification to us at a previous time. (a break in protocol?)
I can only blame us for this situation. We never have met with their development staff, and still never have. We should have insisted, but never did. Whenever interfaces are being developed you need people with intimate knowledge of each system to be involved in the design of the interface. I know that I know every part of every program I work on better than any manager or tech support technician ever could. Even 3 years after development has finished.
If direct interaction between development staff's is not attainable, then specifications must be detailed with ultra-fine granularity and signed off as correct by a number of parties, especially developers. There is no use in just specifying "The amount of the discount is expected in this field in this format", you need to know how subtle changes to the amount will effect the operation of the system you are interoperating with. Does the discount effect tax values, if so, under what circumstances should they be modified.
Lessons Learned
1. Always make sure specifications you create are correct, and have a ultra-fine granuarity.
2. When developing interoperability guidelines between 2 systems, always join all development parties at least once, and decide on interop standards.
3. Never trust other specifications written by a 3rd party, always look for the holes.
4. Always jump on the holes before testing begins, in a small ISV situation most of the time the first sales have been made by the time testing starts. Revenue, and your next pay packet may depend on it.
Is it just me or is everyone having high CPU utilisation issues with FeedDemon?
Other than the CPU utilisation, FeedDemon is not to bad, I like the "Daily Newspaper" view for an overall view of posts, however the 3 vertical panes layout seems more of a hinderance than a help. Back to SharpReader for me.
Luke, please add the awesome browser functionality as used in FeedDemon, but switch to embedded gecko! :)
Thanks to Danny Ayres for helping me get my MovableType RSS1.0 template working better! MovableType! Please ship a better stock standard RSS1.0 template.
On another note, has anyone made a wml template for MovableType?
The Scud (Mark Phillippoussis) is playing Roger Fedarer in the Wimbleton final tonight! Go the Scud! Keep the Aussies on top!
As reported in The Australian IT as well as other major news vendors around the country, email scams advertising fake online banking websites are hitting people hard.
It's a shame that a large proportion of Internet users are not knowledgable enough to weed out the scams from the authentic. Hopefully no one will be hit to hard. The fake websites seems to be a front for collecting personal information.
What is the solution to this problem?
Had a nice breakfast this morning at 'Bien', located on King Street in the heart of Perth. A great little cafe, and an indulgence in french toast and blueberry pancakes. Talk about sugar high!
Highly recommended for breakfast.
Blogchatter looks interesting, but who _really_ wants to sit staring at their web browser as pings flow in?
Before I head out, here this is certainly an interesting bit of information regarding how far information can spread in MS Word documents. I wonder how long it will be before MS Word Revision History information appears as evidence in a court case.
Thank god I really have no need to use MS Word :)
Dinner tonight at Sarah's. Her brother is in town for a few days from Melbourne, so its time for a good feed. I don't know what we're having, but im sure it will be delectable. Damn she is a good cook! Have a good Saturday night everyone.
Thanks to scoble I am inspired to actually start blogging on a more frequent basis.
Let me introduce myself: I am a 24 year old Software Developer who lives in Perth, Australia. I work for PalmTEQ Limited, a designer of Hospitality Point of Sale software and hardware, centered on Wireless Ordering (ie Waiters use handheld devices to take orders at tables and fire them off via Radio Frequency to kitchens etc). My responsibility at PalmTEQ is centered around the development of our wireless ordering software (based on PalmOS) and some of our backoffice applications. I have a lovely girlfriend, Sarah who has been by my side for over 6 years now, but with no intention of marriage any time soon :)
So, now that my main homepage is this weblog, let me show you around my home. Above is links to general parts of my website that are not to do with the blog, pictures (mainly vacation) but I will start adding more stuff interesting stuff as time goes by, maybe even a photoblog there, but I need a phonecam before I can start with that business. Can anyone recommend a good photocam? The rest of the site needs a bit of a spit-shine as I haven't touched it in a while.
Anyway, look out for more activity here in the future. I have 2 categories, Me and Tech. I should imagine Tech will be more widely used in the short term.