Hyvä ystäväni Heimo on Apple-fani, ja vinkkasi Applen jouluhintapäivästä.

Päätin katsoa millaisia tarjouksia sieltä löytyisi, erityisesti suhteessa hankinnan alla olevaan uuteen työkoneeseeni.

Price: Eur 2.376,23
PROMOTION SAVINGS: -Eur 74,59
VAT: Eur 506,36
Subtotal: Eur 2.808,00

* 2.8GHz Intel Core 2 Extreme
* 4GB 667MHz DDR2 SDRAM - 2 x 2GB
* 500GB Serial ATA Drive
* SuperDrive 8x (DVD±R DL/DVD±RW/CD-RW)
* Apple Mighty Mouse
* Keyboard (Finnish) + Mac OS X (Finnish)
* ATI Radeon HD 2600 PRO with 256MB memory
* 24-inch glossy widescreen LCD
* AirPort Extreme
* Bluetooth 2.0 + EDR

Vertaa tätä Mikromafian Challenger2-laskurilla laskettuun tuota lähimpänä olevaan, mutta silti huomattavasti parempaan kokoonpanoon:

Kotelo: Antec P182 ATX musta + 400W Fortron
Emolevy: Intel DG33TL, 4xDDR, vga/dvi, gblan, hd-audio, firew, raid, mAtx
Prosessori: Intel Core 2 Quad Q6600 2,4 GHz 8 Mt S775
Muisti: Kingston 4096Mt DDR2 667 MHz CL5 2x2048 Mt Kit
Kiintolevy: Seagate Barracuda 7200.10 500GB SATA2 32MB
Optinen asema: Samsung DVD+/-RW Dual Layer 20x, SATA, musta
Näytönohjain: Geforce 8800GT PCI-E 512 Mt (tulossa)
Käyttöjärjestelmä: Windows XP Professional UK (OEM)

Paketin hinta: 1239 eur (sis. alv 22%)

Lisäksi tietysti näyttö, Samsungin 24" 245B hintaan 520 EUR.

Vielä tätä parempi kokoonpano, jota suunnittelen pohjaksi seuraavalle työkoneelleni, on:

Kotelo: Antec P182 ATX musta + 400W Fortron
Emolevy: Intel DG33TL, 4xDDR, vga/dvi, gblan, hd-audio, firew, raid, mAtx
Prosessori: Intel Core 2 Quad Q6600 2,4 GHz 8 Mt S775
Muisti: Kingston 8192Mt DDR2 667 MHz CL5 4x2048 Mt Kit
Kiintolevy: Seagate Barracuda 7200.11 1TB SATA2 32MB
Toinen kiintolevy: Seagate Barracuda 7200.11 1TB SATA2 32MB
Optinen asema: Samsung DVD+/-RW Dual Layer 20x, SATA, musta
Näytönohjain: Geforce 8800GT PCI-E 512 Mt (tulossa)

Paketin hinta: 1587 eur (sis. alv 22%)

Järeän kokoonpanon tarkoituksena on toimia alustana usealle samanaikaiselle virtuaalikoneelle, joissa voin pyörittää kulloisenkin tarpeen mukaan eri J2EE-sovelluspalvelimia ja tietokantoja kuten DB2 Express-C ja Oracle XE. Isot levyt ovat tarpeen, sillä aion antaa VMwarelle käyttöön suoraan raa'an levypartition, siinä uskossa että se parantaa virtuaalikoneen suorituskykyä.

En voi siis löytää minkäänlaista perustetta harkita Applen tuotteita, edes tuon 70 EUR alennuksen kanssa, sillä vähintään 1000 EUR hintaero huonomman Applen ja paremman standardituotteen välillä on aivan liikaa.
Mary Poppendieck gave a really great talk at QCon on the role of leadership in software management. It's freely accessible, as well as the first two chapters of her book Implementing Lean Software Development: From Concept to Cash.

Both the talk and the book are very exciting.
It's on Safari now!

Implementation Patterns really ties common sense software development concepts into a neat, digestible package that creates new common ground among your developers.

It's divided into well named themed sections, so that you can easily look up the solution you need.

I very much recommend Kent Beck's Implementation Patterns to both junior and senior developers, as well as their managers.
Olen julkisen sektorin työntekijä, ammatiltani ohjelmistoarkkitehti.

En ole koskaan pitänyt itseäni erityisen isänmaallisena.

Siksi minua häiritsi kun huomasin ajattelevani Helsingin Sanomien lakkoartikkeleita lukiessani että se mitä Kokoomus ja Keskusta ovat tekemässä hoitajille on epäsuomalaista.

Miksi ajattelin että se on epäsuomalaista? Eikö tälläinen epäisänmaallisuus ole enemmän amerikkalainen käsite? Enkö ole aina ollut sitä mieltä, että on asiatonta arvostella toisten sitoutumista yhteiskuntasopimukseen, jos he eivät suorastaan heittele polttopulloja?

Pidän suomalaisesta yhteiskunnasta. Olen ajatellut, että se, että täällä on niin kylmää niin suuren osan vuodesta, on tarkoittanut lukemattomille sukupolville sitä, että paikat on pidettävä kunnossa, ja rakennukset ja työkalut rakennettava kunnolla, jotta talvella ei tule kaikille kylmä ja nälkä.

Ihmisten välisiä suhteita on muokannut se tosiasia, että jos jätät apua tarvitsevan hankeen yöksi, hän kuolee takuulla, ja se on hankeenjättäjän vastuu. Siksi aina on autettava. Samalla tavalla jokaisen on saatava ja syötävä oma reilu osuutensa tai kaikki nälkiintyvät tai kuolevat.

Nämä periaatteet ovat iskostuneet suomalaiseen kulttuuriin sukupolvien kokemuksen kautta, ja näitä ajattelutapoja itse kukin tiedostamatta soveltaa myös moderniin elämään niinkuin parhaiten osaa.

Se mitä Kokoomus ja Keskusta tekevät nyt rikkoo tätä vastaan.

Se on epäsuomalaista.

Esi-isäni häpeäisivät elämistä samassa tuvassa näin käyttäytyvien ihmisten kanssa.

Minä häpeän sitä, että tuen heidän toimintaansa työskentelemällä julkisella sektorilla tälläisen hallinnon alaisuudessa.

Olen jo jonkin aikaa arvioinut julkisella sektorilla työskentelemisen todellista hyödyllisyyttä yhteiskunnalle kriittisesti, mutta tämä heinänkorsi katkaisi nyt karhun selän.

Sinä päivänä kun käskykirjeet lähtevät irtisanoutuneille hoitajille, minä eroan julkisen sektorin työstäni. Vaikka vien tietysti myös henkilökohtaisen paperin, tätä voi tarvittaessa myös pitää irtisanoutumisilmoituksena.

Tuolla tavalla ei vain saa toimia. Se on väärin.
Viime aikana uudet tuotekokeilut ovat menneet yllättävän hyvin, eikä Blå Band Mexican patakaan pettänyt.

400g HK:n naudan jauhelihaa ja Mexican-pussi maksoivat yhteensä noin 5€, mutta en huomannut ostaa standardireseptiin kuuluvaa papupurkkia, ja korvaan chilipaprikan sitruunapippurilla ja chilijauheella.

Tätä suomalais-meksikolaista kokemusta täydentää pussi nachoja ja purkki salsaa.

Hyvää oli, mutta seuraavalla kerralla lisään myös pavut ja juustoa päälle.

Päivitys: myös Italian oli hyvää.
Game over, check please.

http://code.google.com/android/what-is-android.html

Android is to Nokia's development environments what a developer is to a bureucrat.

Getting the SDK, installing it, and creating an application is SO much easier with Android.

It's as if Google had figured out some way to make money without trying to extort every available penny out of everyone who even looks your platform's way, and without trying to control every possible aspect of their developers' lives.
There are two basic modes of action in business: making money and saving money.

Well, make it three: the black sheep is, of course, losing money.

Making money is much more interesting than saving money.

Making money means that you pour money into a business process, and get more money out from the other end of the process than you originally invested.

Your boss's boss has a lot of money to invest. He's only interested in hearing about business processes that are both profitable and, as importantly, can accomodate the kind of amount of money he has available to invest.

If a business process is profitable, but can only take 500 EUR and turn it to 1000 EUR, he won't be interested. Yes, even if a 50% profit is guaranteed, it's not worth it unless you can put in at least 1,000,000 EUR, or in some large companies, 100,000,000 EUR, and get a 20% profit on that.

Similarly, saving money is interesting if the way to save money is applicable to a large business process that goes through a lot of money to provide a profit.

When the business is turning a loss, however, your boss and your boss's boss will be eager to find savings even in medium and small processes, especially in those parts of the process that are not directly turning existing inventory or other assets into revenue. They are eager, not because of any direct business effects of those savings, but because in times of loss, there is a lot of blame to go around, and they will be eager to receive as little as possible. One way to duck blame is to be able to make an argument, or even show, that your department is cutting costs. The blame pertains both to them, personally, and to their whole department. If you don't help them duck it, it might be your position that gets cut.

OK, are you ready to sell technology?

First, start with a self-assessment. Do you know enough about your department's and the company's business, objectives, strategy and budget so that you can talk about them with your boss in such a way that your boss will find you credible and want to continue the conversation?

The answer is probably no. You've spent all your life learning to be really good at what you do, while your boss has spent his life trying to do the same. There's not a lot of natural common ground, which is a shame.

If you can't immediately sound credible, don't start selling before you've had a chance to prepare.

The good news is that if your boss has a high opinion of you as a person, expert and employee, she'll probably want to help you to become able to discuss those subjects credibly.

If not, or if she's just too busy, or there's some other reason, you can try to move laterally, and start by expanding your knowledge of, for example, sales and marketing, by finding a trustworthy contact and expressing the wish to learn. People are generally very eager to talk about themselves and what they do well.

You might want to think through why you're embarking on this quest for knowledge. People will be interested in your motivations and objectives. Learn to express your sincere desire to be able to work with different kinds of people better, and your awakening to the fact that other professions are imporant, too. They will be curious to know whether you have a specific political motivation for your questions. Best not to mention political matters at this point. If you're known as a firm proponent of some idea that's losing traction, try to remember that it's very important to try to look at that matter, and the whole, from an alternative point of view.

With a little help from your friends, you'll be able to find out what larger objectives and constraints your business and your department has, how the political landscape looks, and who you must win on your side to be able to sell your ideas. Be careful to find out what your boss and boss's boss are interested in, and are in a position to sell upwards.

Now, knowing what you do about technology, and with your newfound knowledge, draw a grid of all of the matters you find important at this point, and think of how the technology you want to sell can make a little money in a lot of places, a lot of money in few places, save money, and so on like we've discussed.

Then pull back and think of what in that technology creates that particular opportunity. Generalize to the level of a fairly well known phenomenon. If you'd want to sell VMware, you'd think virtualization at this point, or maybe even consolidation. OK, take those broader concepts and think things through. What other things associate to the concept? What kind of a history does that concept have, what kind of a future do credible or popular writers predict, and would you personally bet ten thousand euros on that technology. What are the alternative ways to approach that business problem? Now that you think about it, is the technology you want to sell really an appropriate solution to this problem? Go through your grid and write these things down.

Sleep on it, and let your thoughts simmer for a couple of days.

Now it's time to cull the herd. Cut out the ideas that you don't have enough expertise to discuss with confidence. Cut out ideas that don't apply to your department's current situation. Cut it all down, leave three to five ideas. Cutting doesn't mean that you abandon the idea entirely, it just means that you'll have to think about it more before going forward with it.

Research the three to five things you picked. Who else is doing this, or similar things? What kinds of resources would you really need to do this? How many buyers are there on the market? Has your boss or your boss's boss said something that could be interpreted as interest in this problem?

You want to do two things that are difficult to do at once: prepare yourself very well for a discussion about your ideas, but also keep your mind and opinions on the matter fluid enough that you can see the problem from the viewpoint of others, adapt to new or conflicting information, or change your opinion entirely if need be.

Then it's time to start selling your ideas. If you're unsuccesful, you'll get the blame. If you're successful, someone else will take the credit. But you don't care about that, right? Right?
Now to head out to the solaris-indiana mailing list... I've been a subscriber for a few months, but haven't really read the list so far, just archived it into a GMail folder.

What a bunch whiners I found on that list, going on endlessly about irrelevant details like obscure default shells, distribution naming, or why nobody has consulted their endlessly bickering subcommittee about some urgent detail.

Someone there might want to watch this video.

To be continued...
After a quick trip to EasyVMX! to create a suitable empty image for VMware Player...

I created a 10GB disk image to install Indiana on, it should be enough. Gave it 1GB of memory. The install is running pretty slowly.

For sure, I'll get 8GB of memory for my next work-from-home computer, as well as an extra 500GB disk I'll slice up so that I can give my virtual machines direct disk access for faster action.

The "Loading smf(5) service descriptions" part of the first bootup takes a while, wouldn't it be possible to precompute whatever this does?

First boot done, now to log in as the username I provided to the installer.

Accept the license without reading it, just like everybody else. Wow, the fonts are screwed up even worse. Change the resolution to 1024x768. Set the font antialiasing again, and this time go to the Advanced tab and set the DPI setting manually to 96.

Start up a terminal. The home directory is in /export/home/$USER. The default shell is bash, thank goodness. The root file system is ZFS.

How to get the network adapter working under VMware? The solaris-indiana mailing list to the rescue, this is a known defect, http://defect.opensolaris.org/bz/show_bug.cgi?id=66.

Shutting down the system from the Gnome button doesn't work (another known defect) so I had to run "pfexec gnome-sys-suspend" from a terminal.

I chose the "e1000" workaround for the networking defect, had to disable the automagic networking option as described at http://opensolaris.org/os/project/nwam/phase0/nwamd_1m/ and then configure the virtual network adapter through the System -> Administration -> Network utility. A reboot, and the network connection is up and working.

Created a ZFS snapshot of my home directory by running the command "pfexec zfs shapshot zpl_slim/export/home@now". Have to remind myself that "pfexec" means "sudo" in Solaris-land. I was lucky enough to find out about the pfexec through a Google search, this would have been a very short test drive if I hadn't. Hope others are as lucky.

I don't think I'll try zones yet. They probably won't work yet.

At least Firefox works after I've set its fonts to look reasonable.

To be continued... part 3.
Oh, the Indiana preview is now available, say the RSS feeds of blogs.sun.com in my Google Reader.

The download page is pretty simple. Wow, I just clicked on the ISO image link, and it started downloading immediately, instead of the usual Sun rigamarole with five logins and checkbox clickery.

Download speed is around 600kBps...

Boots up nicely in VMware... the first window it opens contains a huge, huge license file, which everyone with a brain knows nobody will read through.

Xorg and Gnome are configured with a DPI value that results the fonts being really tiny.

System -> Preferences -> Appearance -> Fonts, set subpixel smoothing on.

System -> Preferences -> Screen Resolution, set to 1024x768.

Applications -> System Tools -> Terminal.

Wow, the default shell just works like I expect it to work. What a huge step forward for Solaris.

Sudo doesn't work. Networking doesn't work. "su -" does work, the password is "opensolaris".

Open Firefox. The default font is terrible. Firefox -> Edit -> Preferences -> Content -> Fonts and Colors -> Advanced: Serif => "Bitstream Vera Serif", Sans-serif => "Bitstream Vera Sans", Monospace => "Bitstream Vera Sans Mono".

Now to create a VMware virtual disk to install this on.

To be continued... part 2.
top