Archive for October 2006
We went sitting in an agile project management “evening seminar” last Monday and
I took some notes it inspired a little rambling. There’s no much useful stuff here, this is just a diary entry. I merely went there to see who was participating as I’m less keen to this Finnish agility enthusiasm. There’s two reasons for it: 1) if google-kind of self-organization isn’t enough for managing your projects, you’re already screwed; 2) there isn’t produced especially good software products in Finland and I think they should put more effort on defining and refining the product strategies – as long as nobody’s interested in the stuff you’re producing, it doesn’t really matter how crappy the management methodology is.
A Fact: Project management is an activity, which consumes time to make better use of the rest of the time. The best case is if people can manage their own work so well that a more rigid management is inefficient; this is the Google case and probably one of the reasons why Google must focus on hiring only smart people who are capable of managing themselves.
Only if you are in such an unfortunate situation that you must produce a working solution with the wrong people, you must very much think about the management methodologies: the further the project scope is beyond the skills of the project team, the more focus on management. My life is really too short for that kind of shit, so I’m focusing more on operational (programming techniques) and strategic stuff (SaaS) than on managerial ones (agile programming). Btw., even if Google was adjusting their product management, that shouldn’t much effect their project management.
Mentoring, on the other hand, is an activity and style of thinking I like very much; I love working with ‘fresh’ people, they just mustn’t be placed on positions they’re not competent for. Tools are yet another, pretty much orthogonal aspect to management and good tooling is a must.
Another major failure is in the managerial thinking. A study of Finnish management in general revealed that over half of managers regarded their own employees as incompetent. Like, wtf? I believe the kind of management style that leads into should be illegal! No professions were specified in the research, but as it was paid by Accenture, it’s pretty safe to assume IT staff was included. So, I guess that explains a little why IT people are so stressed (in general.)
The basic principles of management should be focusing on the strengths (vs. weaknesses, which apparently is the case with over half of the managers) and guiding the employees in using their strengths as a basis for building their personal competitiveness. There is no such thing as an ‘incompetent employee’, there are only incompetent managers who have failed at positioning the people; of course, there are no incompetent managers either, also they are just placed in a wrong position. (Well, ok, people may be incompetent for their positions, but it should be thought of bad casting only.)
Then there was also Tom Gilb giving his presentation and it was pretty interesting – as opposed to our local offering. It had nothing to do with agile programming, but this was probably just a good thing. The quantification stuff, related to measurement, was pretty nice. Although, he stated that a high-level feedback-loop should occur weekly and I believe the situation is pretty bad if significant weekly adjustments are justified; you must adjust if you’re using time on going through tactics weekly, mustn’t you? Otherwise it’s just a waste of time; don’t get hypocritical with this. Actually, that is only a project management issue which were to be ignored anyway. The principal, methodology-agnostic idea of his was to precisely quantify the traditionally immeasurable quality requirements, like maintainability, usability etc. But besides motivation and some good results, he had not much magic to represent; just plain old divide-and-conquer.
The mug, t-shirt and just the single sticker on the cover.
But yay, check out the new much improved look! (Left above, vlog 10/26.) Now we’re talking!! Such a transformation in just three days! :) Tantek must’ve been very busy collecting stickers… (And no, I’m not beeing sarcastic, this is genuine admiration; the picture above is much better!)
Yet here’s the authoritative reference by Robert.
Technorati: Technorati, Vlog, Swag
(PS. What a tuning with the layout!! But now it’s perfect! ‘ ’ still has some use…)
Update 10/29: Corrected the links to vlogs as they we’re mixed up.
Jean-Louis brought up the issue of the limits of HTTP, which has been also on my mind since I realized the X in XMPP. I very much agree that HTTP is highly overused; it’s the ‘hammer and nails’ syndrome once again. Of course, it’s easy to waste a lot of time by constantly jumping between all the different technology-bandwagons, but evolution needs to happen eventually. My hopes were high on that Flex/OpenLaszlo would have surpassed AJAX, as HTML is not a user interface language! MXML (of Flex) is, and there’s a world of difference in that. Similarly, HTTP is not a two-way/push protocol, what some folks are trying very hard to work around. XMPP is two-way and, thus, complicated callback protocols wouldn’t be needed if XMPP was used as the SOAP transport [think about it].
It could be interesting to see some statistics, but I guess RSS polling isn’t really taking up significant server time nor bandwidth yet, with caches and all. However, I believe this can change quickly if the polling interval is reduced from (sub-)hours into minutes – and with the upcoming RSS popularization by IE7. HTTP really is not viable for such scenarios – very inefficient use of resources, indeed.
The ‘long tail of technology’ is definitely a smart metaphor. I like it.
The blogging phenomenon and feed technologies have really increased the quantity of both the fresh information and the speed it can be consumed at. Feed readers help to skip the laborious browsing and searching phases by bringing the new content directly into your attention. But, as this regards only somewhat ‘communicative’ content (blogs), also the amount of ‘documentative’ kind of content (wikis) is increasing. Are you suffering from information overload? You should, as it’s just like exercise: you get better only by beating your limits. Has anything ever slowed down in the history of mankind? (No.) It’s time to adapt! As with everything else, we’ve seen pretty much nothing yet. If you’re not scared of the future just keep on reading.
This blog entry was inspired by yesterday’s Technorati vlog episode where they promoted their new search engine for finding more recent stuff compared to Google. And, it just might be so that a significant proportion of the information flowing through the Technorati plumbing really is stale after two weeks. (Btw., one should have this kind of labels also for short-living web pages.) Of course, the wiki-type of content is more permanent in nature, but Technorati is clearly focusing on the latest buzz.
This brings us to the duality of web 2.0 evolution: on the other hand there is this communicative usage, which is transient, and then there’s the ‘documentative’ usage for content that has more long-term value. The latter is the more traditional way to use the web and the former describes the hip MySpace-phenomenon. There was also this inspiring slideshow by Sam Ruby [via] about the future of web where he stated that “Teenagers see web as transient”, which clearly is accurate. The initial web 2.0 definition includes the wiki-kind of collaborative editing but seems to miss the communicative aspect. Blogs can be and are being used against you for both purposes. Transience naturally discards huge amounts of information (and just might lead into The Great Forgetting of the 21st century) but when has personal communication become worth permanent storage? Let’s just leave Google and some out.
Is web becoming a media for real-time communications? Is it replacing or merging with some other communications media? Now, I strongly suggest checking out this video. (And noting especially the web page editing app.) How about collaborative blogging? Your readers can communicate with you in real-time and give you instantaneous feedback while you’re just writing the entry. A very early aggregation is assumed to get the feed readers notified of the blogging activity. We’re talking about some fast communications! Even though I’d probably hate it right at this very moment, I can see that there’s potential.
Btw., just as a reminder, XMPP is the real-time communications protocol. Could it be that the communicative content goes over XMPP? And the web gets back to it’s roots as a storage of documentative stuff? An unlikely scenario as such, but something like it, perhaps.
Ok, the current reality is that much of the tube stuff (feeds) is mere communications, which has little permanent value. How to manage the information overload? Recognize and separate communication from knowledge valuable for you – e.g. by categorizing the feeds accordingly. It’s not a trivial task as the feeds contain mixed information and you need human intellect to extract the bits with long-term value. And sometimes the discussion is more valuable than source information; wikis are just superb in this sense as they cumulate information.
Conclusion: Web is happening as a communications platform and that role is developing very strongly. The communication is only getting faster and one must adapt to it for being able to act. Now, I must leave the rest of thinking up to you as I’ve very much exceeded my bed time. Take care!
P.S. Once again, I failed to be conversational with you guys by picking up my own topic. I guess this entry goes into the ‘Documentative’ bucket, then. :)
P.P.S. I’m sooo going to send some fan-mail to that hottie Aaron. Oh, and if only Google Talk would support video calls he could give me a private vlog session! ♥♥ (And now I really hope you watched that episode.;)
I finally did it! Yay! Now I go sleep! Although, I’m not totally happy with the agressive JPEG compression – again. The funny part is that the original image is always stored on the server (available through link on the right image) but you can’t get it to header. Well, I can’t do much about it – except to buy the premium service, but I’m not that desperate with the image quality…
PS. Ah, of course I had to put my face on a badge there. Jia!
PPS. Gimp all the way! It rocks!
Update 10/02: They’ve done something, but definitely doesn’t look very good still! (Regarding image quality only, I know they can do very little about the face…)
The software design industry is finally developing from crafting somewhat tailored solutions into a more industrial production; the component oriented architecture took a little longer to materialize than anticipated. The software industry is transforming into a Fordistic industry where final software products are assembled from components, for real. “Modularity, and interchangeable modular components are a key component of the modernist approach in software, as in architecture, marketing, production, and elsewhere.” […] Btw., I also cherish the crafting metaphor; if the industrialization is done in a Google-kind of inspiring way, it should be ok, but otherwise I welcome you managers to put the Indian people on the line.
(Just some of my even-less-intelligible ranting: I’m excited to see if the BRIC gets its quality together in software development, while I’m not the biggest believer in the
Indian outshored software quality. I’ve seen some very skilled and passionate people but most of them have been slow, sloppy and noisy; and it feels like a cultural thing. The worst part is that the people doesn’t seem to understand it themselves as it would naturally be ok if they at least would want to get better.. Give them a hint, you fellow countrywo/men.)
This new factory line has a qualitative difference in that the assembly is done at the design level (different level of abstraction), obviously due to the immaterial nature of software :) – i.e., the old story about the cost-free copying of products while design still being manual work. Thus, the pace has not been bad considering that the last shift from the ‘farming Kool-Aid’ into industrialism took three centuries.
This paradigm shift reminded me of this already older development where IT is developing from the support role of the core operations (banking, design, car sales, clothes manufacture etc.) into the core business; which makes programmers the primary workforce. For example, Amazon’s former core business, logistics, is becoming a commodity and the IT services as itself has become the area of most attention already a long time ago. Though, eBay – another example of the most modern ‘systems factories’ – continues to use it’s IT services to support the core business, peer-to-peer sales. Which is actually pretty self-evident: Amazon charges for using it’s services compared to eBay’s web services API that is free to consume.
Naturally, the eBay platform can be – and probably is being – utilized for purposes which don’t directly drive eBay’s revenue, but that small amount of debris is easy to tolerate. Especially as $73/user/yr can be considered a pretty good revenue having most of the servicing immaterial and fully automated. ($1.39b revenue in Q1 with 75.4m users.)
So, I guess that was about it for now… I’ll go grab some lunch.
Disclaimer: I’m just a programmer, I know jack shit.