kmxiv-blog

[kmxiv-blog]

Südwegen giraffes logo.

[kmxiv-blog]

[kmxiv-blog]

The Premier League is not as good as it used to be.

It's true.

Despite having some of the most exiting finishes to the season in recent years, featuring a last-kick winner in one year and 3rd bringing down 1st to let 2nd win the trophy at the death - there is no lack of drama.

It is, by far, the most interesting league in the world in terms of storylines, heroes, villains, drama, press conferences, words, tweets, transfers and so on, but when it comes to football, the actual quality isn't there.  It's much lower than it should be, considering the amount of money in the English game.

Manchester City, who have no limit to their budget other than Financial Fair Play, should be way better than this.  It seems that they have built a team that can grind out results week after week against poor opposition, which is great for winning leagues, but is useless for the Champions League, where it's a sprint, not a marathon.

Chelsea, for all their smart spending and transfers, for building a great team and having one of the best managers in history, have won nothing last year.  They look like title favorites this year, but that's only because the only challenger, Manchester City, have failed to put up much of a fight.  Chelsea are grinding out results at home and in Europe, but they should be better than that.   This Chelsea team should score more than Ancelotti's Chelsea team in 2009-10, which has the record for most goals scored in the league.  Mourinho's Madrid has the record for most goals scored as well.  How does bringing that manager to this team result in this kind of Chelsea?   Even if Mourinho wants to turn Chelsea into a team that wins points, not goals, leagues, not praise, the team is still not as good as his earlier Chelsea team - they're not grinding out results as well as they used to.

Arsenal have clearly regressed.  Having gone from a team that just needed a bit more money to challenge or even win a title to one that needs to spend a lot of money to stay 4th.

Liverpool's spending does not seem to even guarantee them 4th, much less the title.  Tottenham's post-Bale spending has yet to return them to the heights of the Bale-Modric era.  Manchester United are not the force they used to be, despite the spending, and play a really boring brand of football.

The top of the league is interesting, the race for 4th has five teams separated by 4 points, even though the quality isn't that great, but the rest is pretty boring.  Aston Villa are painful to watch, Sunderland are quite poor, and I believe that if there was a points or goals threshold for survival from relegation, some years would have 6 or 7 teams getting relegated.

There are way too many boring games with average players not trying hard enough.  It's not that they can't, but many teams have only survival as their season objective - they don't have to do well, they just have to not do as poorly as the worst 3.  You could say that that's just how good they are.  Some players, some teams are just the 14th best team in the country.   But they don't play like the 14th best team in the highest tier of football in the country.  They don't play like they're the 14th best team in the richest, most popular league in the world.  They play like crap, especially once their safety is guaranteed.  

Partly is transfers - some of these players are not good enough.  That's their level and that's all they can do.  Mediocre players playing for mediocre teams playing against other mediocre players in mediocre teams produce boring matches.  But the moment that match is a derby, then it becomes interesting.  Suddenly, they play better as if they have a point to prove.  The question is, why don't they play like this all the time?  Why only do this twice a year when trying to one-up their rivals?

These are the problems facing the Premier League today:

1. Poor spending - overpaying for promising or established English talent, overpaying for mediocre, established foreign talent, no small risks taken on unproved players.  Poor scouting networks that can't find value for money in South America, Eastern Europe.

2. Poor management and coaching.  Even average teams can play well if managed well.  Lack of ambition beyond survival and one-upping rivals.  No tactical discipline.  Average team-work.  Teams don't improve that much over time.  Teams play as a collection of individuals.  Lack of individual training for passing, shooting, dribbling, ball control, etc.  Overemphasis on fitness and athleticism.  Managers try to spend their way out of problems - buying a better defender than improving the defense; a striker that can score from 1 or 2 chances than trying to improve chance-creation.  Reliance on individual moments of brilliance to win games - great saves, off-the-line clearances, incredible goals.  Games become boring 89-minute affairs with 1 moment of magic, luck or stupidity.  

Teams should emulate Germany's 4-0 against Argentina and 7-1 against Brazil as an example of team-work and routine chance-creation and finishing without individual moments of brilliance.  Argentina and Brazil are Germany's peers in international football.  All three are in the very top-tier.  I don't expect teams in the Premier League to play like Germany's national squad, but team's that are 8th and 9th should be able to put their peers to the sword like that.  And even if it's too much to ask, then teams that are 8th and 9th should be able to do that to teams that are 19th and 20th.   It's difficult to imagine that Klose, Müller and Podolski and Götze playing for different clubs have a better understanding of each other and play better together for 6 - 10 games a year than Premier League teams that play 40+ competitive games a year.

Chelsea drawing Burnley 1-1 could be described as giant killing, sure.  But Hull and Sunderland drawing Liverpool 0-0 is less so.  And then when we get to Crystal Palace 0-0 Aston Villa, Everton 0-0 WBA, Aston Villa 0-0 Newcastle, Burnley 0-0 Sunderland,  Southampton 0-0 West Ham... then you start to wonder why some teams even bother showing up.  There have been 22 0-0 draws this season and 29 1–1's and this does not describe the times when one team scores 1 goal and tries to close the game - we have 52 1-0's.  

  To compare this with the German league, the Bundesliga has 17 0-0's to the PL's 22,  37 1-0's to the PL's 52, 20 1-1's to the PL's 29.  

To put these stats to a picture - only Köln and Hamburg have fewer goals than matches played.   The Premier League has 7 teams with a ratio of under 1 per game.   At the end of the year, 3 German teams will be relegated - one or two of those teams would be ones with a ratio of more than 1 goal per game.  The Premier League will relegate 3, meaning at least 4 teams with goal ratios of less than 1 will continue to play in the Premier League, with an injection of 3 more teams from the Championship.  Last year there were 4 teams that scored 37 or fewer - only one survived.

With little to play for and little incentive to change, things are unlikely to change.

This is great news for new teams with a new kind of thinking to break into the Premier League.   

[kmxiv-blog]

The Chromoscope is not a camera made by Südwegen Germany (also called the Mothership).  It is a camera made by several teams across the world.

A camera is more than just a sensitive sensor in a box, which you expose for a moment to capture a picture.  There’s a lot more to it than that.

So let’s start with the sensor.  Oci.  You’ve heard of them before?  Oci makes the sensors for all Südwegen cameras, phone cameras, scanners and medical imagers.  This is not their only product, but this is a big project for them that will take a lot of their time and resources.  The sensor they will use for the Chromoscope is another one named Prokudin-Gorskii.  It is a new sensor, but the design of the silicon is based on one already in use in some scanners.  Scanners are different from photographs in that they are not time critical.  It’s okay for them to be slow, as long as the scan is clear and the colors are correct.    In real-time photography, you don’t have that luxury.   The world won’t stand still for you while you take a picture, so we need to work on speed.

Next is Usi, pronounced Ooshi.  Usi means ears in Russian and if you see their brand identity, it’s filled with rabbits.  Usi makes microphones and sound equipment, some of which I have used in recording my album.  They will be making sure that the sound that is captured by the camera itself is great.

Next up is Okna. Okna make displays for Südwegen and will be making the back LCD for the camera.

Xenith, often called Xenith Russia, will be making the computer behind the sensor.   The Chromoscope will be an Oci sensor connected to a Xenith computer, a computer with RAM, a CPU, an SSD, a file system and so on.  Considering that we have a giant sensor taking big images, we need a computer that can process these fast and write them fast to disk so that you can keep shooting.

First Touch, my baby, will be playing a small part, providing a touch screen digitizer so that the camera controls are easily accessible.

Our recently founded sister company Ionix, the entity that holds many patents and technologies that we need, will also be helping us.  Ionix will be working on the lossless audio and video codecs, lossy wireless streaming from the camera to monitors and lossy codecs for run-and-gun scenarios, easy workflows and easy sharing.

And last, but not least, our friends in the Czech Republic and Slovakia, Kremik. Kremik will make hardware cores from the software that Ionix will develop.  So instead of using the main processing unit to transcode video, stream video, display it on the Okna screen or view it on another monitor, they will make dedicated IP cores that do this with the least amount of power possible.  They will make all the peripheral - WiFi, HDMI-out, Micro-SD and so on.

There is only one maybe in all this and it is the involvement of Voodoo, a small Spanish lab that makes high quality, long-life, very expensive battery packs.  It would be nice if they could make us a battery so that the camera could be used in situations where it’s difficult to have a power source and cabling.

[kmxiv-blog]

I’m still sad about the league title and nothing you say could make it better. :’(

[kmxiv-blog]

I told my friend, “I’m going down to the store…  want something?

He said “Get some sheebs"

“Sheebs?"

“Yeah, Sheebs!"

“What the fuck is sheebs?"

“It’s potatoes, sliced, you know!!! Like Lays or whatever"

“You mean chips?"

“Yeah! That’s what I meant!"

“What the hell would you say sheebs?"

“For you to understand"

“How the fuck would saying sheebs help me understand?"

[kmxiv-blog]

I love you, Masha.

[kmxiv-blog]

Kirche, um Mitternacht

[kmxiv-blog]

No time to brood. Swim or sink. Fight or die.

Silicon Mali

[kmxiv-blog]

So, new camera sensor in the works in parallel with the camera module, DSP and actual camera.  Yes, in parallel.

The final camera will have:

 LCD/OLED(?) monitor from Okna

Touchscreen controls from First Touch

Built-in microphone from Ushi

DSP from Xenith

HDMI-out and all that from Kremik

Camera body designed by Christian Adler

Lens from Südwegen 

Can’t tell you anything about the sensor yet.

AS you can guess, I’ll be designing all the user interface.  I’m looking at a block of aluminum and a sheet of glass with a sensor inside and I must make a camera out of it.

[kmxiv-blog]

Surprise!  That’s what’s waiting for us and the Iceberg team.  Some secrets are just badly kept.  After today’s Proof of Concept cancellation, we had decided to put the whole development of the Iceberg phone on ice (no pun intended) and just focus on beating the benchmarks.  After beating the benchmarks, we can get the green light to get back to work on the phone.

So someone must’ve leaked that bit of information even though it’s supposed to be a secret.

The good thing is that news leaks both ways:

After finding out that we planned to “cheat” to beat the benchmarks, that we were building a benchmark-beating machine rather than a phone that performs really well, they decided, in “secret,” to not give us a date for the postponned POC.  It’ll be a surprise, so we should be ready at any moment.

So now we know that we should be expecting a surprise visit for the POC.

This puts us in a difficult position.

The plan was, work on one month of the benchmarks, beat them, then work on the phone.  Now, we don’t know if the POC is in two months, three months or even more.  We cannot split the team into two, benchmark team and phone team, because then we’ll have two phones, one that can only beat benchmarks and one that can’t.  It’ll split our resources and waste it in this benchmark project, which brings nothing to the Iceberg.  If we keep the team as one and the benchmark is in a month, we will not get those numbers, especially as we don’t know what these numbers are.

So what do we do now?  Build a benchmarking machine that can also run Honey&Ice?

I wonder if they intentionally leaked the “surprise visit” so that we know what we’re up against and they want to see how we handle a difficult decision like this.  How we manage our time and resources and how we deal with unknown deadlines, the element of surprise and many volatile, “external” factors.

It’s as if they’re training us or me in particular for the madness of the market.

That’s how I like to see it.  I see right through them.

And they know that.  And they know that I know that they know that I… 

I’m not sure what the correct decision is though.  Make a half-assed phone that benches well, or make the best phone in the world that gets canned because it didn’t meet certain numbers.

[kmxiv-blog]

I was sitting there, waiting, playing with my phone.  They make you wait.  They want to make you sweat and panic.  Sometimes they’re busy, sometimes they do play these games.  But they’re not just games, they’re really their modus operandi.  It’s their business.

It’s not a game to make you worry.  It’s not personal.  It’s not them vs you.  It’s not about you at all.  

It’s about money.  Their money.  If you panic, they don’t win.  If you panic, they lose.  They realize that they put their money on the wrong horse and put their trust in the wrong man.

If you panic, it’s because you don’t believe in your product or yourself.  It’s a lose-lose for everyone.  Then they panic.  Do we dump the product or dump the man?

But I did not panic, even though the glass was not ready.  I told them many times that October is too soon.   They said it wasn’t a “hard deadline” but they expected our work to be in “hard real time.”  The guy looked at me while saying that, looking me in the eyes for my reaction.  He wanted to say “We’re not just idiots with cash.  We know what we’re talking about… We’re techincal too,”  He also wanted to see if I got the reference, if I was just an ideas guy or if I understood the concepts.  But ultimtely, he wanted to say “work should be processed as it comes in.  It shouldn’t pile up."

We had an argument about this after he had left.  Some members of my staff believed that he was saying “Whatever features you add, the deadlines won’t change” but then he would have said “hard deadline” which is exactly the thing he said it wasn’t.  I believe that he meant “if you have a new feature that will take a month to implement and test, and you have 5 months of work ahead of you, it should take you 6 months to be done with your workload.”  He does not mean “whatever you add, it should still be 5 months" because that is ridiculous.

There is an incredible amount of hostility from engineers towards managers.  The oft-repeated joke is “a project manager thinks that 9 women can give birth to a baby in 1 month."

The retort to that is that if you ask 9 engineers to bake one pie over one afternoon and let them manage themselves, you’ll get 9 different half-baked pies because they all felt like doing it their own way - one engineer insisting on flying Macintoshes in because it just has to be an apple pie, another refusing to let his pie cool by the window, some only want ingredients where they know everything about how and where they was produced, while the rest want to make all the ingredients from scratch.

Hard real time means just that: 5 + 1 = 6.  It is fair.  Many say 5 + 1 stays 5, while engineers say 5 + 1 = 7, because 1 is complex to implement.  If it is complex, then say 2.   Don’t say 1 and mean 2.  Be honest with your estimates.  Don’t say “it takes me a month to develop” so 1 and the rest i the QA team’s fault.  How much time does it need for the company, not for you personally… that’s the question.

But the senior guys did not come today.  One of them was there, but on other business.  They called and said that they’re impressed that I showed up.  They said that they knew the product wasn’t ready, but admired my courage to stand behind a half-baked product.  I was the one being benchmarked today.

[kmxiv-blog]

I was cool as ice - my chilled, calm heart was pumping slush through my veins. But they did not come.

Silicon Mali

[kmxiv-blog]

[kmxiv-blog]

Tomorrow is a very, very important day for Xenith.  It might be a sink or swim day for the Xyrus microprocessor, for the Iceberg phone, many other products and for Xenith itself.

We really have to show that this works and we didn’t spend millions developing a clock that produces heat.

I could lose my job tomorrow.   These guys really mean business.  Wish me luck.

[kmxiv-blog]

Some instruction words are just way too fucking long.

Silicon Mali

[kmxiv-blog]

My phone crashed when I needed it the most and took more than 2 minutes to boot.  Sweet.