Feeling less ill today but still coughy and stuff, with the added bonus that since I've not been sleeping well and been coughing a lot, my back now hurts - so every cough sends pain down my back which is just great.

Trying to debug my TMDB fetcher this evening, I discovered some odd headers in some of the messages I was seeing. The HTTP 'Connection' header seemed to be coming back mangled - as either 'nnCoection' or 'Cneonction'. It seems that this is an intentional feature of a loadbalancer that wants to force the client to ignore the connection response. The reason, it appears, is that by swapping the characters in this way, the TCP checksum remains the same. Which variant is used depends on whether the 'C' starts on an odd or even byte offset for the 16 bit checksum to trick to work. Its a bit tacky.

Really, really didn't sleep well last night. I think I got about 2 hours sleep until 8, and then slept through to 11. And then I was out of any cold-stuff, so I had to pop out to the shops.

We live out lives in private shells Ignore our senses and fool ourselves To thinking that out there there's someone else cares Someone to answer all our prayers, all our prayers... [ The Last Straw, from Clutching At Straws, by Marillion ]

I've been trying to do useful things today but to be honest it's not gone well. Concentrating on anything for more than a few minutes at a time just isn't working. But I've got music on, and I'm trying to stay warm with the heater on and a blanket over me. It also turns out that bouncing to music when you're ill makes your head hurt. So don't do that.

Didn't sleep at all well last night and my cough's got worse all day. So I'm definitely ill . I don't like being ill.

Normally nobody calls me, but someone did this evening around 10ish - just a minute after I'd rushed to the bathroom. Doh. Oh well. If it was important, they'll call back, and they didn't leave a message so I guess it's not that important.

I'm too tired to think now, so sleepy byes for me.

I'm slowly working my way through Ally McBeal (1997, FOX)Comedy/DramaAlly McBeal is an American comedy-drama series which aired on the Fox network from 1997 to 2002. The series was created by David E. Kelley, who also served as the executive producer, along with Bill D'Elia. The series stars Calista Flockhart in the title role as a young lawyer working in the fictional Boston law firm Cage and Fish with other young lawyers whose lives and loves were eccentric, humorous and dramatic. Ally McBeal in evenings, rather than watching anything more modern. It's funny and makes me happy sometimes - I'd like to think that some of Ally's whimsy is still with me, and it reminds me that there's something to try to look forward to. Maybe I hope that there's a hope for me and my little bit of crazy.

Tonight it's "Ally McBeal2x11
"In Dreams"Upon discovering her favorite high school teacher is terminally ill, Ally seeks a court order to allow the women to spend her remaining days in a drug-induced coma, happily dreaming of her imaginary lover. Cage must gather his courage to confront Nelle after Fish tells him she thinks Cage has not had sex with her because he's gay.In Dreams", which is one of my favourites from the ones that I watched - Ally's arguing that a patient would be better off living in their dreams, which I can relate to. Certainly seems better than the real world sometimes.

Coo... "Ally McBeal2x12
"Love Unlimited"Ally defends a woman coming out of a nine-year marriage whose husband, claiming mental incompetence, wants to have the marriage annulled. Nelle and Cage are having rocky times in their relationship. Ally and Dr. Butters come to a new understanding after he hears her impassioned defense of a client in court.Love Unlimited" has Bruce Willis in it - I didn't remember that!

I was just thinking... isn't there meant to be a Fringe (2008, FOX)Action and Adventure/Drama/Science-FictionTeleportation. Mind control. Invisibility. Astral projection. Mutation. Reanimation. Phenomena that exist on the Fringe of science unleash their strange powers in this thrilling series, co-created by J.J. Abrams (Lost, Alias), combining the grit of the police procedural with the excitement of the unknown. The story revolves around three unlikely colleagues - a beautiful young FBI agent, a brilliant scientist who's spent the last 17 years in a mental institution and the scientist's sardonic son - who investigate a series of bizarre deaths and disasters known as "the pattern." Someone is using our world as an experimental lab. And all clues lead to Massive Dynamic, a shadowy global corporation that may be more powerful than any nation.Fringe this weekend ? It's my little treat at the end of Sunday if I've been good and done stuff on the Rambles... only it looks like it's on a break until January. January! That's just cruel, especially as there have only been 7 episodes. What can be my special treat now ?

My throat's sore, my face and neck ache, I've got sniffles and I'm not thinking very straight - not that I think that straight normally. So I think I might have caught that cold from work. Damn.

I'm very tired today, mostly 'cos I was up till the late hours of last this morning trying to get some stuff to work - successfully, I might add. But it's left me very tired today. And it seems that I've managed to catch this cold from people at work, if the sore throat is anything to go by.

I wanted to knock something small off my list this evening before I went home, so I looked for a reviewboard RSS feed... it turns out that the version we skipped is the version that they were removed in. RSS feed support was removed from it. That's really quite baffling to me.

Oh, and I watched "Doom (2005)Action, Adventure, Horror, Sci-Fi, ThrillerSomething has gone wrong at a remote scientific research station on Mars. All research has ceased. Communication has failed. And the messages that do get through are less than comforting. It's a level 5 quarantine and the only souls allowed in or out are the Rapid Response Tactical Squad - hardened Marines armed to the teeth with enough firepower to neutralize the enemy...or so they think.Doom" this evening, too. Which was about as bad as I'd expected it to be. Maybe a bit worse. Oh well. It was easier than getting up from the chair.

Still not sleeping great. Woke up today aching a lot and hadn't really slept all night. I got home this evening and just got into bed and slept for a bit. I've had a little tea now but I'm still tired.

I often just questioning what I really want to do. I keep hoping that I'll find something, but to be honest I've been pretty deflated these last couple of years. Maybe it's just 'cos I've ended up where I don't want to be, without the things I wanted and that just makes me sad.

I think I want things to be a certain way and I'm sad when they're not. I pretty much accept that I'm going to be sad. I worry a little that if I finish writing stuff, I'll just completely lose any sort of impetus to do anything and die.

Oh well.

I'd still like to think that there's more than work and then sleep. Especially when the former is frustrating and the latter is... well, also frustrating. I blame me entirely, though. Hey ho.

I should probably try to sleep, though. Maybe tonight it'll work.

Maybe it's feeling guilty. When people say 'how do you sleep at night' to people who do things they don't approve of, that's all they mean. And I feel guilty all the time, regretting things, and wanting to do something better but never really doing so. I apologised to whole bunch of people over the years to try to change that. Some were more surprised than others. No amount of sorry ever seems to make up for the many things that I think were wrong. (of course there's a whole lot I'm not sorry for, where I know I made the right decision, etc, etc, but that doesn't make me feel as guilty).

Oooh... the words went all swirly... I should go sleep. Maybe.

I have a fear of edges. Just a mild one - 'cos if you're at the edge of the road, it's scary. What if you were to jump out into a car. Or at the edge of a high place without a railing. Or even with one. What if you were to jump off, or slide under it ? Would you bounce a few times as you fell ? Would it be feel fun until you hit the jagged rocks ? If you stepped in front of the lorry, how far would you travel and could you make sure that you landed on your arms so that your head didn't break. Would your shirt look like a pretty cape as you flew through the air ? Would you get to think anything other than 'oops' ? If you were lying in a heap what's the coolest thing you could say, that people would remember ? And, having planned the coolest thing you could say, would you really remember it after having bounced do the street/cliff ?

So, I stay away from edges.

Hmm... I feel dizzy again, so I'm going to give up being incoherent and try some of that sleep, tempered by telly. The telly often helps quell any stray thoughts that might be distracting. Mostly Caroline, but other stuff too. And there's a whole world of guilt in itself.

I hate the English language at times. There's so many special cases and words that just don't make sense when you try to work with them. But, that said, I've managed to address all the typos raised in my rambles and I can now continue writing real stuff. It vaguely amuses me that a few of the things that I've written about came up on Usenet recently. More amusing is 'fixed that years ago'.

I tried adding little Facebook 'Like' buttons to the site earlier today, as well. It seemed to work, but it appears that I have to hand over control of part of my site's content to them - trusting that they don't inject anything into my site that I don't like. Firstly, I don't trust Facebook. Secondly, I don't trust any other site to host their content on gerph.org. Thirdly, I don't see why it should be necessary to do so, but they don't offer any local hosted solution. Fourthly, it endorses Facebook over other solutions, and because of the 'firstly' I don't see why I should. So I've disabled the code that adds it.

Much as I like the idea of being able to flag such things as liked, the principles of openness and choice which the Internet should embrace mean that I don't really think I can do that. Interesting and useful to know, but ultimately not going to be used.

I was thinking this morning about how things were going. 'cos I felt a little better than normal. Maybe 'cos I've got the lawn cut. There's no real change, though. It's still the same kind of struggle to know why to get up. Actually that's not quite true. The 'why' is just fear and commitment - fear that otherwise what else do you have, and commitment because that's what you do when you work.

I still don't know why I should do it. Wouldn't it be easier to just get a small flat and pack myself into it - I don't need everything I have right now - and not care about anything else for a couple of years. I'd go mad pretty quickly, but what's wrong with that ? Maybe I'd sleep better, which can't be a bad thing.

I've given up on trying to fit in to the world. I'd like to win the lottery and just move away to a small island. Again, I'd go mad, but it's something I'd want, and as there's very few things I know that I want maybe it'd make me happy.

I sort of remember happy. It comes in waves when you relive things. I often wonder if other people have to put up with constantly remembering and reliving things from their past. Maybe if you have more in your life than I have, you don't have that. It's often a little distracting.

I also wonder what other people do when they don't work. How do they fit in their lives when there's so little time ? I guess they make it. I guess I set aside a lot of time for my stuff because there's nothing else. I gave up trying to find anything else, really. It bothers me that other people manage to be a lot better about this stuff, but I don't have any energy to try. Sometimes it seems like I'll just slip more into a little shut-in world of my own, when even the simple things are just scary. Fortunately I don't think about that so much.

Occasionally, I wonder what I'll do once I finish the rambles, and whether that'll matter. That said, it's now mid-November and whilst I'm still getting stuff done on it, the number of FIXMEs is still very high. There's not a lot else really that I'm caring about - work is work and whilst I care in the sense of wanting to do stuff right, it's not the way that I cared about stuff in the past. It's just not worth the investment really.

Julian had an idea, a while back, of keeping a record of his mood through the day and plotting it in a graph. I keep enough graphs and other records that I'm not sure that another would help.

Thinking about that, I'm not sure where this random need to quantify stuff comes from. Collecting datasets and producing details from them is interesting, despite not always producing interesting details. The graphs I've got from the news sites have been interesting in a sense, but there's not much us any great use that I can draw from them.

The news stuff was interesting, but it got put on the back-burner because once I had some details about other publishers (albeit not like mine), I had what I needed. I wanted to include some other sources, but many of them are behind pay walls so you cannot trawl them in the same way. And The Independant decided to redesign their entire site so that none of it was parseable with my scraper.

I've stuck a snapshot output from the processor as a page linked from the Diary. It's not particularly special, but it does show some interesting bits. I love the fact that the distributions look like proper statistical distributions. Shows that it's actually working, really. There's a whole bunch of other statistics collected about the sites which aren't graphed, and the links to the sites (which I've disabled) go to pages which describe the content of every article processed.

It can be a little overwhelming to go through all the data, but seeing the graphs makes it worthwhile. It says that there was some useful information to be collected from it. I do wish it was a little easier to interpret!

Some bits you can see... 20 words/sentence is common, but CNN tend to go for fewer, however their distribution is less clear - possibly because the number of articles in the sample is smaller so it's harder to tell. The Guardian has more articles with longer words (20-26) than the BBC, which seems to aim for (14-20). The Register has a greater spread, with 18-28 words being its main range - maybe because of their greater use of freelance writers, different styles of articles, or a looser style guide than the others.

The BBC's use of video-only articles, or video+single paragraph skews their results for the number of words per article. The Guardian's use of 'blogs' as articles skew its numbers of words the other way. The Register also has some very long articles, but mostly focuses on the 200-600 range. CNN seems to favour the 200-400 range, but again maybe the limited number of articles in the source data limits this. Both Drobe and IconBar heavily hit the 200 word-article, with a smattering of longer articles as well.

The Short sentence distribution is interesting, as generally it implies that about half (usually 40%) the sentences are classed as 'short', and this seems consistent across the board. Probably this is a factor of the calculation used. Similarly long sentences fall into the 20% category in almost all cases. Don't think there's much you can draw from that.

Now Kincaid readability is more interesting. This is measured in school grade levels. BBC hits grades 8-10, whereas The Guardian hits 9-12 - clearly skewed to the right on the graphs there, which is quite interesting as it implies that The Guardian aims for readers at a higher level, whereas the BBC aims to hit more readers by being of a lower grade level. It's not completely accurate, but these statistics have been taken from some reasonably trusted algorithms (AFAICT).

The Register hits 10-12, but it - like The Guardian - is skewed towards the higher end. CNN focuses on 9-10, which you might think is American dumbing down (if you're a Brit with a typical attitude), but the spread looks more skewed to the right than the BBC graph, to me. Iconbar is 6-10, Drobe is 8-11.

SMOG grading is a similar school-grade level, but calculated a different way. The results are pretty similar to Kincaid, although slightly higher.

Sentences per paragraph is fun, mainly because it shows the BBC for their notorious habit of having single sentences as a paragraph. (although I'm not entirely sure how there are some that have an average of 0.75 sentences per paragraph - that just confuses me)

The Guardian, because of its blogs, gets highly skewed, but generally has more sentences per paragraph, with 1.5 being a common average. The Register and CNN are similar, but never stray too far. This is average, remember, not the maximum or minimum - that would also be an interesting statistic, as would range, but that's not shown here (the data is collected, but not graphed).

Iconbar and Drobe are hard to read due to their extremes. Maybe that's due to a bad parser - they do have some awful markup in their articles - but if we take the results as given, Drobe appears to err on the longer side, but both fit the general patterns.

I think the far wider spreads (in general) on The Iconbar and Drobe indicate a lesser degree of editorial/stylistic control. That may mean that they're more free-flowing, but also mean that there's less of a site-image being put across. Which is precisely what you'd expect from an enthusiast site, really.

From the look of the results, you really need a dataset of at least 200 articles to get useful statistical results out of the sites. Depending on the feeds I've chosen, they might not be available. The CNN feeds I selected obviously weren't as large as some of the others. For IconBar and Drobe the client parsed every article present (where it could understand them) up to mid-October. That's why there are significantly more articles there than with the other sources.

Anyhow, that's just my interpretation of the data. If I had the time and energy, I'd collect more articles over a longer period and analyse them more carefully. It's not that I'm not interested, but that I wanted to get back to Rambles.

I kinda wonder, with this in mind, whether I'm just doing the wrong job. But then what is it that I would enjoy which involves that kind of thing and which wouldn't make it just seem like work ? I'm sure I'd get bored if all I was doing was playin with figures.

Dave Thomas suggested a few days ago that I should get something to check spelling mistakes for me, as he'd spotted one and couldn't remember where. There are a lot in my rambles, it seems. Actually there are a few and there's many words I've invented, or the dictionary doesn't know. I've essentially got a simple stream processor that can take either text or HTML and produce highlighted corrections to it.

This makes it easy to spot the words it thinks are wrong, but tedious to fix them. There's a few regular ones, like 'startup' (which should be 'start up') or 'plugin' (which should be 'plug-in'), but there's also many hyphenated things that seemed right at the time, but don't necessarily exist in any real dictionary. 'off-by-one' is a simple example. In many cases, I can remove the hyphens and it makes sense, but if the hyphens are there to make a compound term that should be thought of as a single entity, that's harder. And it seems that I make up a lot of these compound words . There's also common words that I get wrong, like 'useable' (should be 'usable'), and similar things.

At the start there were about 1880 'typos' - words that the dictionary didn't know. I've managed to get that down, over the last 4 days, to just 330. As it's quite tiring, I've done no writing in the meantime. But also there are some things I spotted that are written badly and can be reworked (there's another - most 're*' words don't seem to be hyphenated, but I liked to do so, which means I've learnt something too).

I've got one rather large ramble which still has 20 'FIXME' marks in it that need to be addressed. I need to split it up, but I'm not sure where to make the break. My solution up to now has been to address the 'FIXME's in the hope that the break will become obvious. So far it's not really worked out, and the ramble's getting longer.

What do you get when you fall in love? A girl with a pin to burst your bubble. [ I'll Never Fall In Love Again, from Our Town- The Greatest Hits, by Deacon Blue ]

I managed to get a few thousand words written today, and now I'm really tired, and a bit sad. But having done that I get my reward to watch Fringe (2008, FOX)Action and Adventure/Drama/Science-FictionTeleportation. Mind control. Invisibility. Astral projection. Mutation. Reanimation. Phenomena that exist on the Fringe of science unleash their strange powers in this thrilling series, co-created by J.J. Abrams (Lost, Alias), combining the grit of the police procedural with the excitement of the unknown. The story revolves around three unlikely colleagues - a beautiful young FBI agent, a brilliant scientist who's spent the last 17 years in a mental institution and the scientist's sardonic son - who investigate a series of bizarre deaths and disasters known as "the pattern." Someone is using our world as an experimental lab. And all clues lead to Massive Dynamic, a shadowy global corporation that may be more powerful than any nation.Fringe.

I keep getting job adverts from Jobsite by email. They stop mailing you more than once a month if you don't visit the site in a bit, but as soon as you do, they start up again. I did a while back to see what was around and they haven't stopped again yet . Anyhow, I was clearing out the ones from my inbox and glanced at the ones from around Cambridge - first one I looked at (ok there were only 2!) - was for Velocix. Or if it's not for Velocix, it's for an incredibly similar place with a similar history.

Having said that it does list 'Windows' as a 'Desireable skill', which seems a little wrong for here. But still it amused me.

Day off today - felt very tired and slept until about 11, which was nice but then I never got to wish Julian well on his driving test. I rang him just after 1:30 though, and found that he'd passed, which was great .

Wrote one of the harder rambles this evening - trying to go back and check some facts was a little more difficult than usual. Some of the bits are interesting and amusing. Other bits are just painful. Actually one of the harder things is trying to create a reasonable linear narrative. Generally development wasn't linear - it would cause other offshoots to be done where problems were found or there was something that needed attention before the current work could progress.

Sometimes those offshoots were completely unrelated - whether they be because a bug report came in which needed looking at, or because whilst trying to test or debug the code something else was noticed to be wrong and had to be investigated there and then whilst it was obvious. Other times, those offshoots lasted quite a while whilst root causes were determined or when it was found that the problems went far deeper than was thought and that blocked continuing on the previous task.

However if you're talking about things you either need to take a linear look at the task, and remove all the distractions and stacked items that went in to making it. Or you look at the linear timeframe and talk about things in the order they happened. The latter is a harder read, unless you're trying to stress the non-linearity of development (which is often obvious). The former tends to condense the apparent timeframe, but if expressing the time isn't a big deal, that doesn't matter.

Actually, there's at least a third option which is to widen the scope of 'task' to a 'goal' based narrative, where you need to reorder the discussion of tasks within it such that they no longer take place in a linear timeframe, but they connect by their interactions (which may have been developed alongside one another, but just cannot be described in such a way - well, not if you want to keep your readers).

I'm working my way through the first season of Ally McBeal (1997, FOX)Comedy/DramaAlly McBeal is an American comedy-drama series which aired on the Fox network from 1997 to 2002. The series was created by David E. Kelley, who also served as the executive producer, along with Bill D'Elia. The series stars Calista Flockhart in the title role as a young lawyer working in the fictional Boston law firm Cage and Fish with other young lawyers whose lives and loves were eccentric, humorous and dramatic. Ally McBeal and it's quite amusing seeing Jane Krakowski in it playing Elaine. Obviously I remember her from there, but since I know her now from 30 Rock (2006, NBC)ComedyEmmy Award Winner Tina Fey writes, executive produces and stars as Liz Lemon, the head writer of a live variety programme in New York City. Liz's life is turned upside down when brash new network executive Jack Donaghy (Alec Baldwin in his Golden Globe winning role) interferes with her show, bringing the wildly unpredictable Tracy Jordan (Tracy Morgan) into the cast. Now its up to Liz to manage the mayhem and still try to have a life. 30 Rock. What I'd not noticed was that the joke about 'my sexuality' had been carried straight over - it seems, certainly in a few episodes of Ally, to have been a running line for her. It was used a few times in 30 Rock as well.

Feeling a little better today, although I was woken up by someone yelling outside at about 20 to 2 this morning. No idea what it was but it was loud...

I've gone home from work today, feeling icky. I think I've got a bug or something - I got home, slept for a few hours and I've been crazily hot ever since, despite the house being about 19 degrees. I don't usually feel this hot in the office, and there it's 26 degrees a lot of the time.

So I've slept a bit, and when I've woken up I've written little bits of words and tried to let my mind settle. When I feel ill, or very tired, it tends to go a little crazy. On Friday night it kept me awake trying to work out the right place to put the commas because they were all wrong.

I've been writing a bit more on the rambles today. It's still really hard to write much, but I'm trying to be good and keep it up. I've got lots of planned bits to write but starting them seems to be very hard. I'm not really expecting anyone to be that interested, but I'd like to finish it.

I've been splitting some bits up where it feels like there's a natural break. It's really hard to read large amounts of text, so I'm trying to keep the numbers to manageable levels. To that end, I've added to the munin graphs a cumulative word count graph, and a 'range' graph which tells me what the average number is, the maximum, minimum and the 1st standard deviation. Ideally (!) they should converge as I reorganise the rambles and finish off those that are part written. I'm not sure if that will actually work out or not, but it'll be interesting to see.

I think I've managed to chip a tooth, or maybe lost a filling. Either way, I'm a little sore at the moment. I may actually have to bite the bullet and see a dentist .

I've been watching Ally McBeal (1997, FOX)Comedy/DramaAlly McBeal is an American comedy-drama series which aired on the Fox network from 1997 to 2002. The series was created by David E. Kelley, who also served as the executive producer, along with Bill D'Elia. The series stars Calista Flockhart in the title role as a young lawyer working in the fictional Boston law firm Cage and Fish with other young lawyers whose lives and loves were eccentric, humorous and dramatic. Ally McBeal today, which has been quite funny and I've really enjoyed it so far.

Whilst I was going through my stories and poems to add the OpenGraph stuff to them, it was pretty obvious that most are focused on either regrets or children. I think that's pretty accurate. I found a few little paragraphs that I'd left in the root of my drive which were ... similar.